Sphinx 2.2.10-release reference manual

Free open-source SQL full-text search engine

Table of Contents

1. Introduction
1.1. About
1.2. Sphinx features
1.3. Where to get Sphinx
1.4. License
1.5. Credits
1.6. History
2. Installation
2.1. Supported systems
2.2. Compiling Sphinx from source
2.2.1. Required tools
2.2.2. Compiling on Linux
2.2.3. Known compilation issues
2.3. Installing Sphinx packages on Debian and Ubuntu
2.4. Installing Sphinx packages on RedHat and CentOS
2.5. Installing Sphinx on Windows
2.6. Sphinx deprecations and changes in default configuration
2.7. Quick Sphinx usage tour
3. Indexing
3.1. Data sources
3.2. Full-text fields
3.3. Attributes
3.4. MVA (multi-valued attributes)
3.5. Indexes
3.6. Restrictions on the source data
3.7. Charsets, case folding, translation tables, and replacement rules
3.8. SQL data sources (MySQL, PostgreSQL)
3.9. xmlpipe2 data source
3.10. tsvpipe\csvpipe (Tab\Comma Separated Values) data source
3.11. Live index updates
3.12. Delta index updates
3.13. Index merging
4. Real-time indexes
4.1. RT indexes overview
4.2. Known caveats with RT indexes
4.3. RT index internals
4.4. Binary logging
5. Searching
5.1. Matching modes
5.2. Boolean query syntax
5.3. Extended query syntax
5.4. Search results ranking
5.4.1. Ranking overview
5.4.2. Available built-in rankers
5.4.3. Expression based ranker (SPH_RANK_EXPR)
5.4.4. Quick summary of the ranking factors
5.4.5. Document-level ranking factors
5.4.6. Field-level ranking factors
5.4.7. Ranking factor aggregation functions
5.4.8. Formula expressions for all the built-in rankers
5.5. Expressions, functions, and operators
5.5.1. Operators
5.5.2. Numeric functions
5.5.3. Date and time functions
5.5.4. Type conversion functions
5.5.5. Comparison functions
5.5.6. Miscellaneous functions
5.6. Sorting modes
5.7. Grouping (clustering) search results
5.8. Distributed searching
5.9. searchd query log formats
5.9.1. Plain log format
5.9.2. SphinxQL log format
5.10. MySQL protocol support and SphinxQL
5.11. Multi-queries
5.12. Collations
6. Extending Sphinx
6.1. Sphinx UDFs (User Defined Functions)
6.2. Sphinx plugins
6.3. Ranker plugins
7. Command line tools reference
7.1. indexer command reference
7.2. searchd command reference
7.3. spelldump command reference
7.4. indextool command reference
7.5. wordbreaker command reference
8. SphinxQL reference
8.1. SELECT syntax
8.2. SELECT @@system_variable syntax
8.3. SHOW META syntax
8.4. SHOW WARNINGS syntax
8.5. SHOW STATUS syntax
8.6. INSERT and REPLACE syntax
8.7. REPLACE syntax
8.8. DELETE syntax
8.9. SET syntax
8.10. SET TRANSACTION syntax
8.11. BEGIN, COMMIT, and ROLLBACK syntax
8.12. BEGIN syntax
8.13. ROLLBACK syntax
8.14. CALL SNIPPETS syntax
8.15. CALL KEYWORDS syntax
8.16. SHOW TABLES syntax
8.17. DESCRIBE syntax
8.18. CREATE FUNCTION syntax
8.19. DROP FUNCTION syntax
8.20. SHOW VARIABLES syntax
8.21. SHOW COLLATION syntax
8.23. UPDATE syntax
8.24. ALTER syntax
8.25. ATTACH INDEX syntax
8.26. FLUSH RTINDEX syntax
8.27. FLUSH RAMCHUNK syntax
8.30. SHOW PROFILE syntax
8.31. SHOW INDEX STATUS syntax
8.33. OPTIMIZE INDEX syntax
8.34. SHOW PLAN syntax
8.35. SHOW DATABASES syntax
8.36. CREATE PLUGIN syntax
8.37. DROP PLUGIN syntax
8.38. SHOW PLUGINS syntax
8.39. SHOW THREADS syntax
8.40. Multi-statement queries
8.41. Comment syntax
8.42. List of SphinxQL reserved keywords
8.43. SphinxQL upgrade notes, version 2.0.1-beta
9. API reference
9.1. General API functions
9.1.1. GetLastError
9.1.2. GetLastWarning
9.1.3. SetServer
9.1.4. SetRetries
9.1.5. SetConnectTimeout
9.1.6. SetArrayResult
9.1.7. IsConnectError
9.2. General query settings
9.2.1. SetLimits
9.2.2. SetMaxQueryTime
9.2.3. SetOverride
9.2.4. SetSelect
9.3. Full-text search query settings
9.3.1. SetMatchMode
9.3.2. SetRankingMode
9.3.3. SetSortMode
9.3.4. SetWeights
9.3.5. SetFieldWeights
9.3.6. SetIndexWeights
9.4. Result set filtering settings
9.4.1. SetIDRange
9.4.2. SetFilter
9.4.3. SetFilterRange
9.4.4. SetFilterFloatRange
9.4.5. SetGeoAnchor
9.4.6. SetFilterString
9.5. GROUP BY settings
9.5.1. SetGroupBy
9.5.2. SetGroupDistinct
9.6. Querying
9.6.1. Query
9.6.2. AddQuery
9.6.3. RunQueries
9.6.4. ResetFilters
9.6.5. ResetGroupBy
9.7. Additional functionality
9.7.1. BuildExcerpts
9.7.2. UpdateAttributes
9.7.3. BuildKeywords
9.7.4. EscapeString
9.7.5. Status
9.7.6. FlushAttributes
9.8. Persistent connections
9.8.1. Open
9.8.2. Close
10. MySQL storage engine (SphinxSE)
10.1. SphinxSE overview
10.2. Installing SphinxSE
10.2.1. Compiling MySQL 5.0.x with SphinxSE
10.2.2. Compiling MySQL 5.1.x with SphinxSE
10.2.3. Checking SphinxSE installation
10.3. Using SphinxSE
10.4. Building snippets (excerpts) via MySQL
11. Reporting bugs
12. sphinx.conf options reference
12.1. Data source configuration options
12.1.1. type
12.1.2. sql_host
12.1.3. sql_port
12.1.4. sql_user
12.1.5. sql_pass
12.1.6. sql_db
12.1.7. sql_sock
12.1.8. mysql_connect_flags
12.1.9. mysql_ssl_cert, mysql_ssl_key, mysql_ssl_ca
12.1.10. odbc_dsn
12.1.11. sql_query_pre
12.1.12. sql_query
12.1.13. sql_joined_field
12.1.14. sql_query_range
12.1.15. sql_range_step
12.1.16. sql_query_killlist
12.1.17. sql_attr_uint
12.1.18. sql_attr_bool
12.1.19. sql_attr_bigint
12.1.20. sql_attr_timestamp
12.1.21. sql_attr_float
12.1.22. sql_attr_multi
12.1.23. sql_attr_string
12.1.24. sql_attr_json
12.1.25. sql_column_buffers
12.1.26. sql_field_string
12.1.27. sql_file_field
12.1.28. sql_query_post
12.1.29. sql_query_post_index
12.1.30. sql_ranged_throttle
12.1.31. xmlpipe_command
12.1.32. xmlpipe_field
12.1.33. xmlpipe_field_string
12.1.34. xmlpipe_attr_uint
12.1.35. xmlpipe_attr_bigint
12.1.36. xmlpipe_attr_bool
12.1.37. xmlpipe_attr_timestamp
12.1.38. xmlpipe_attr_float
12.1.39. xmlpipe_attr_multi
12.1.40. xmlpipe_attr_multi_64
12.1.41. xmlpipe_attr_string
12.1.42. xmlpipe_attr_json
12.1.43. xmlpipe_fixup_utf8
12.1.44. mssql_winauth
12.1.45. unpack_zlib
12.1.46. unpack_mysqlcompress
12.1.47. unpack_mysqlcompress_maxsize
12.1.48. csvpipe_delimiter
12.2. Index configuration options
12.2.1. type
12.2.2. source
12.2.3. path
12.2.4. docinfo
12.2.5. mlock
12.2.6. morphology
12.2.7. dict
12.2.8. index_sp
12.2.9. index_zones
12.2.10. min_stemming_len
12.2.11. stopwords
12.2.12. wordforms
12.2.13. embedded_limit
12.2.14. exceptions
12.2.15. min_word_len
12.2.16. charset_table
12.2.17. ignore_chars
12.2.18. min_prefix_len
12.2.19. min_infix_len
12.2.20. max_substring_len
12.2.21. prefix_fields
12.2.22. infix_fields
12.2.23. ngram_len
12.2.24. ngram_chars
12.2.25. phrase_boundary
12.2.26. phrase_boundary_step
12.2.27. html_strip
12.2.28. html_index_attrs
12.2.29. html_remove_elements
12.2.30. local
12.2.31. agent
12.2.32. agent_persistent
12.2.33. agent_blackhole
12.2.34. agent_connect_timeout
12.2.35. agent_query_timeout
12.2.36. preopen
12.2.37. inplace_enable
12.2.38. inplace_hit_gap
12.2.39. inplace_docinfo_gap
12.2.40. inplace_reloc_factor
12.2.41. inplace_write_factor
12.2.42. index_exact_words
12.2.43. overshort_step
12.2.44. stopword_step
12.2.45. hitless_words
12.2.46. expand_keywords
12.2.47. blend_chars
12.2.48. blend_mode
12.2.49. rt_mem_limit
12.2.50. rt_field
12.2.51. rt_attr_uint
12.2.52. rt_attr_bool
12.2.53. rt_attr_bigint
12.2.54. rt_attr_float
12.2.55. rt_attr_multi
12.2.56. rt_attr_multi_64
12.2.57. rt_attr_timestamp
12.2.58. rt_attr_string
12.2.59. rt_attr_json
12.2.60. ha_strategy
12.2.61. bigram_freq_words
12.2.62. bigram_index
12.2.63. index_field_lengths
12.2.64. regexp_filter
12.2.65. stopwords_unstemmed
12.2.66. global_idf
12.2.67. rlp_context
12.2.68. ondisk_attrs
12.3. indexer program configuration options
12.3.1. mem_limit
12.3.2. max_iops
12.3.3. max_iosize
12.3.4. max_xmlpipe2_field
12.3.5. write_buffer
12.3.6. max_file_field_buffer
12.3.7. on_file_field_error
12.3.8. lemmatizer_cache
12.4. searchd program configuration options
12.4.1. listen
12.4.2. log
12.4.3. query_log
12.4.4. query_log_format
12.4.5. read_timeout
12.4.6. client_timeout
12.4.7. max_children
12.4.8. pid_file
12.4.9. seamless_rotate
12.4.10. preopen_indexes
12.4.11. unlink_old
12.4.12. attr_flush_period
12.4.13. max_packet_size
12.4.14. mva_updates_pool
12.4.15. max_filters
12.4.16. max_filter_values
12.4.17. listen_backlog
12.4.18. read_buffer
12.4.19. read_unhinted
12.4.20. max_batch_queries
12.4.21. subtree_docs_cache
12.4.22. subtree_hits_cache
12.4.23. workers
12.4.24. dist_threads
12.4.25. binlog_path
12.4.26. binlog_flush
12.4.27. binlog_max_log_size
12.4.28. snippets_file_prefix
12.4.29. collation_server
12.4.30. collation_libc_locale
12.4.31. mysql_version_string
12.4.32. rt_flush_period
12.4.33. thread_stack
12.4.34. expansion_limit
12.4.35. watchdog
12.4.36. prefork_rotation_throttle
12.4.37. sphinxql_state
12.4.38. ha_ping_interval
12.4.39. ha_period_karma
12.4.40. persistent_connections_limit
12.4.41. rt_merge_iops
12.4.42. rt_merge_maxiosize
12.4.43. predicted_time_costs
12.4.44. shutdown_timeout
12.4.45. ondisk_attrs_default
12.4.46. query_log_min_msec
12.4.47. agent_connect_timeout
12.4.48. agent_query_timeout
12.4.49. agent_retry_count
12.4.50. agent_retry_delay
12.5. Common section configuration options
12.5.1. lemmatizer_base
12.5.2. on_json_attr_error
12.5.3. json_autoconv_numbers
12.5.4. json_autoconv_keynames
12.5.5. rlp_root
12.5.6. rlp_environment
12.5.7. rlp_max_batch_size
12.5.8. rlp_max_batch_docs
12.5.9. plugin_dir
A. Sphinx revision history
A.1. Version 2.2.10-release, 07 sep 2015
A.2. Version 2.2.9-release, 16 apr 2015
A.3. Version 2.2.8-release, 09 mar 2015
A.4. Version 2.2.7-release, 20 jan 2015
A.5. Version 2.2.6-release, 13 nov 2014
A.6. Version 2.2.5-release, 06 oct 2014
A.7. Version 2.2.4-release, 11 sep 2014
A.8. Version 2.2.3-beta, 13 may 2014
A.9. Version 2.2.2-beta, 11 feb 2014
A.10. Version 2.2.1-beta, 13 nov 2013
A.11. Version 2.1.9-release, 03 jul 2014
A.12. Version 2.1.8-release, 28 apr 2014
A.13. Version 2.1.7-release, 30 mar 2014
A.14. Version 2.1.6-release, 24 feb 2014
A.15. Version 2.1.5-release, 22 jan 2014
A.16. Version 2.1.4-release, 18 dec 2013
A.17. Version 2.1.3-release, 12 nov 2013
A.18. Version 2.1.2-release, 10 oct 2013
A.19. Version 2.1.1-beta, 20 feb 2013
A.20. Version 2.0.11-dev, xx xxx xxxx
A.21. Version 2.0.10-release, 22 jan 2014
A.22. Version 2.0.9-release, 26 aug 2013
A.23. Version 2.0.8-release, 26 apr 2013
A.24. Version 2.0.7-release, 26 mar 2013
A.25. Version 2.0.6-release, 22 oct 2012
A.26. Version 2.0.5-release, 28 jul 2012
A.27. Version 2.0.4-release, 02 mar 2012
A.28. Version 2.0.3-release, 23 dec 2011
A.29. Version 2.0.2-beta, 15 nov 2011
A.30. Version 2.0.1-beta, 22 apr 2011
A.31. Version 1.10-beta, 19 jul 2010
A.32. Version 0.9.9-release, 02 dec 2009
A.33. Version 0.9.9-rc2, 08 apr 2009
A.34. Version 0.9.9-rc1, 17 nov 2008
A.35. Version, 30 oct 2008
A.36. Version 0.9.8, 14 jul 2008
A.37. Version 0.9.7, 02 apr 2007
A.38. Version 0.9.7-rc2, 15 dec 2006
A.39. Version 0.9.7-rc1, 26 oct 2006
A.40. Version 0.9.6, 24 jul 2006
A.41. Version 0.9.6-rc1, 26 jun 2006

List of Tables


Chapter 1. Introduction

1.1. About

Sphinx is a full-text search engine, publicly distributed under GPL version 2. Commercial licensing (eg. for embedded use) is available upon request.

Technically, Sphinx is a standalone software package provides fast and relevant full-text search functionality to client applications. It was specially designed to integrate well with SQL databases storing the data, and to be easily accessed by scripting languages. However, Sphinx does not depend on nor require any specific database to function.

Applications can access Sphinx search daemon (searchd) using any of the three different access methods: a) via Sphinx own implementation of MySQL network protocol (using a small SQL subset called SphinxQL, this is recommended way), b) via native search API (SphinxAPI) or c) via MySQL server with a pluggable storage engine (SphinxSE).

Official native SphinxAPI implementations for PHP, Perl, Python, Ruby and Java are included within the distribution package. API is very lightweight so porting it to a new language is known to take a few hours or days. Third party API ports and plugins exist for Perl, C#, Haskell, Ruby-on-Rails, and possibly other languages and frameworks.

Starting from version 1.10-beta, Sphinx supports two different indexing backends: "disk" index backend, and "realtime" (RT) index backend. Disk indexes support online full-text index rebuilds, but online updates can only be done on non-text (attribute) data. RT indexes additionally allow for online full-text index updates. Previous versions only supported disk indexes.

Data can be loaded into disk indexes using a so-called data source. Built-in sources can fetch data directly from MySQL, PostgreSQL, MSSQL, ODBC compliant database (Oracle, etc) or a pipe in TSV or a custom XML format. Adding new data sources drivers (eg. to natively support other DBMSes) is designed to be as easy as possible. RT indexes, as of 1.10-beta, can only be populated using SphinxQL.

As for the name, Sphinx is an acronym which is officially decoded as SQL Phrase Index. Yes, I know about CMU's Sphinx project.

1.2. Sphinx features

Key Sphinx features are:

  • high indexing and searching performance;

  • advanced indexing and querying tools (flexible and feature-rich text tokenizer, querying language, several different ranking modes, etc);

  • advanced result set post-processing (SELECT with expressions, WHERE, ORDER BY, GROUP BY, HAVING etc over text search results);

  • proven scalability up to billions of documents, terabytes of data, and thousands of queries per second;

  • easy integration with SQL and XML data sources, and SphinxQL, SphinxAPI, or SphinxSE search interfaces;

  • easy scaling with distributed searches.

To expand a bit, Sphinx:

  • has high indexing speed (upto 10-15 MB/sec per core on an internal benchmark);

  • has high search speed (upto 150-250 queries/sec per core against 1,000,000 documents, 1.2 GB of data on an internal benchmark);

  • has high scalability (biggest known cluster indexes over 3,000,000,000 documents, and busiest one peaks over 50,000,000 queries/day);

  • provides good relevance ranking through combination of phrase proximity ranking and statistical (BM25) ranking;

  • provides distributed searching capabilities;

  • provides document excerpts (snippets) generation;

  • provides searching from within application with SphinxQL or SphinxAPI interfaces, and from within MySQL with pluggable SphinxSE storage engine;

  • supports boolean, phrase, word proximity and other types of queries;

  • supports multiple full-text fields per document (upto 32 by default);

  • supports multiple additional attributes per document (ie. groups, timestamps, etc);

  • supports stopwords;

  • supports morphological word forms dictionaries;

  • supports tokenizing exceptions;

  • supports UTF-8 encoding;

  • supports stemming (stemmers for English, Russian, Czech and Arabic are built-in; and stemmers for French, Spanish, Portuguese, Italian, Romanian, German, Dutch, Swedish, Norwegian, Danish, Finnish, Hungarian, are available by building third party libstemmer library);

  • supports MySQL natively (all types of tables, including MyISAM, InnoDB, NDB, Archive, etc are supported);

  • supports PostgreSQL natively;

  • supports ODBC compliant databases (MS SQL, Oracle, etc) natively;

  • ...has 50+ other features not listed here, refer configuration manual!

1.3. Where to get Sphinx

Sphinx is available through its official Web site at http://sphinxsearch.com/.

Currently, Sphinx distribution tarball includes the following software:

  • indexer: an utility which creates fulltext indexes;

  • searchd: a daemon which enables external software (eg. Web applications) to search through fulltext indexes;

  • sphinxapi: a set of searchd client API libraries for popular Web scripting languages (PHP, Python, Perl, Ruby).

  • spelldump: a simple command-line tool to extract the items from an ispell or MySpell (as bundled with OpenOffice) format dictionary to help customize your index, for use with wordforms.

  • indextool: an utility to dump miscellaneous debug information about the index, added in version 0.9.9-rc2.

  • wordbreaker: an utility to break down compound words into separate words, added in version 2.1.1.

1.4. License

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. See COPYING file for details.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

Non-GPL licensing (for OEM/ISV embedded use) can also be arranged, please contact us to discuss commercial licensing possibilities.

1.5. Credits


Sphinx initial author (and a benevolent dictator ever since):


Past and present employees of Sphinx Technologies Inc who should be noted on their work on Sphinx (in alphabetical order):

  • Adam Rice

  • Adrian Nuta

  • Alexander Klimenko

  • Alexey Dvoichenkov

  • Alexey Vinogradov

  • Anton Tsitlionok

  • Eugene Kosov

  • Gloria Vinogradova

  • Ilya Kuznetsov

  • Kirill Shmatov

  • Rich Kelm

  • Stanislav Klinov

  • Steven Barker

  • Vladimir Fedorkov

  • Yuri Schapov


People who contributed to Sphinx and their contributions (in no particular order):

  • Robert "coredev" Bengtsson (Sweden), initial version of PostgreSQL data source

  • Len Kranendonk, Perl API

  • Dmytro Shteflyuk, Ruby API

Many other people have contributed ideas, bug reports, fixes, etc. Thank you!

1.6. History

Sphinx development was started back in 2001, because I didn't manage to find an acceptable search solution (for a database driven Web site) which would meet my requirements. Actually, each and every important aspect was a problem:

  • search quality (ie. good relevance)

    • statistical ranking methods performed rather bad, especially on large collections of small documents (forums, blogs, etc)

  • search speed

    • especially if searching for phrases which contain stopwords, as in "to be or not to be"

  • moderate disk and CPU requirements when indexing

    • important in shared hosting environment, not to mention the indexing speed.

Despite the amount of time passed and numerous improvements made in the other solutions, there's still no solution which I personally would be eager to migrate to.

Considering that and a lot of positive feedback received from Sphinx users during last years, the obvious decision is to continue developing Sphinx (and, eventually, to take over the world).

Chapter 2. Installation

2.1. Supported systems

Sphinx can be compiled either from source or installed using prebuilt packages. Most modern UNIX systems with a C++ compiler should be able to compile and run Sphinx without any modifications.

Currently known systems Sphinx has been successfully running on are:

  • Linux 2.4.x, 2.6.x, 3.x (many various distributions)

  • Windows 2000, XP, 7, 8

  • FreeBSD 4.x, 5.x, 6.x, 7.x, 8.x

  • NetBSD 1.6, 3.0

  • Solaris 9, 11

  • Mac OS X

CPU architectures known to work include i386 (aka x86), amd64 (aka x86_64), SPARC64, and ARM.

Chances are good that Sphinx should work on other Unix platforms and/or CPU architectures just as well. Please report any other platforms that worked for you!

All platforms are production quality. There are no principal functional limitations on any platform.

2.2. Compiling Sphinx from source

2.2.1. Required tools

On UNIX, you will need the following tools to build and install Sphinx:

  • a working C++ compiler. GNU gcc and clang are known to work.

  • a good make program. GNU make is known to work.

On Windows, you will need Microsoft Visual C/C++ Studio .NET 2005 or above. Other compilers/environments will probably work as well, but for the time being, you will have to build makefile (or other environment specific project files) manually.

2.2.2. Compiling on Linux

  1. Extract everything from the distribution tarball (haven't you already?) and go to the sphinx subdirectory. (We are using version 2.2.10-release here for the sake of example only; be sure to change this to a specific version you're using.)

    $ tar xzvf sphinx-2.2.10-release.tar.gz
    $ cd sphinx

  2. Run the configuration program:

    $ ./configure

    There's a number of options to configure. The complete listing may be obtained by using --help switch. The most important ones are:

    • --prefix, which specifies where to install Sphinx; such as --prefix=/usr/local/sphinx (all of the examples use this prefix)

    • --with-mysql, which specifies where to look for MySQL include and library files, if auto-detection fails;

    • --with-static-mysql, which builds Sphinx with statically linked MySQL support;

    • --with-pgsql, which specifies where to look for PostgreSQL include and library files.

    • --with-static-pgsql, which builds Sphinx with statically linked PostgreSQL support;

  3. Build the binaries:

    $ make

  4. Install the binaries in the directory of your choice: (defaults to /usr/local/bin/ on *nix systems, but is overridden with configure --prefix)

    $ make install

2.2.3. Known compilation issues

If configure fails to locate MySQL headers and/or libraries, try checking for and installing mysql-devel package. On some systems, it is not installed by default.

If make fails with a message which look like

/bin/sh: g++: command not found
make[1]: *** [libsphinx_a-sphinx.o] Error 127

try checking for and installing gcc-c++ package.

If you are getting compile-time errors which look like

sphinx.cpp:67: error: invalid application of `sizeof' to
    incomplete type `Private::SizeError<false>'

this means that some compile-time type size check failed. The most probable reason is that off_t type is less than 64-bit on your system. As a quick hack, you can edit sphinx.h and replace off_t with DWORD in a typedef for SphOffset_t, but note that this will prohibit you from using full-text indexes larger than 2 GB. Even if the hack helps, please report such issues, providing the exact error message and compiler/OS details, so I could properly fix them in next releases.

If you keep getting any other error, or the suggestions above do not seem to help you, please don't hesitate to contact me.

2.3. Installing Sphinx packages on Debian and Ubuntu

There are two ways of getting Sphinx for Ubuntu: regular deb packages and the Launchpad PPA repository.

Deb packages:

  1. Sphinx requires a few libraries to be installed on Debian/Ubuntu. Use apt-get to download and install these dependencies:

    $ sudo apt-get install mysql-client unixodbc libpq5
  2. Now you can install Sphinx:

    $ sudo dpkg -i sphinxsearch_2.2.10-release-0ubuntu12~trusty_amd64.deb

PPA repository (Ubuntu only).

Installing Sphinx is much easier from Sphinxsearch PPA repository, because you will get all dependencies and can also update Sphinx to the latest version with the same command.

  1. First, add Sphinxsearch repository and update the list of packages:

    $ sudo add-apt-repository ppa:builds/sphinxsearch-rel22

    $ sudo apt-get update

  2. Install/update sphinxsearch package:

    $ sudo apt-get install sphinxsearch

Sphinx searchd daemon can be started/stopped using service command:

$ sudo service sphinxsearch start

2.4. Installing Sphinx packages on RedHat and CentOS

Currently we distribute Sphinx RPMS and SRPMS on our website for both 5.x and 6.x versions of Red Hat Enterprise Linux, but they can be installed on CentOS as well.

  1. Before installation make sure you have these packages installed:

    $ yum install postgresql-libs unixODBC

  2. Download RedHat RPM from Sphinx website and install it:

    $ rpm -Uhv sphinx-2.2.1-1.rhel6.x86_64.rpm

  3. After preparing configuration file (see Quick tour), you can start searchd daemon:

    $ service searchd start

2.5. Installing Sphinx on Windows

Installing Sphinx on a Windows server is often easier than installing on a Linux environment; unless you are preparing code patches, you can use the pre-compiled binary files from the Downloads area on the website.

  1. Extract everything from the .zip file you have downloaded - sphinx-2.2.10-release-win32.zip, or sphinx-2.2.10-release-win32-pgsql.zip if you need PostgresSQL support as well. (We are using version 2.2.10-release here for the sake of example only; be sure to change this to a specific version you're using.) You can use Windows Explorer in Windows XP and up to extract the files, or a freeware package like 7Zip to open the archive.

    For the remainder of this guide, we will assume that the folders are unzipped into C:\Sphinx, such that searchd.exe can be found in C:\Sphinx\bin\searchd.exe. If you decide to use any different location for the folders or configuration file, please change it accordingly.

  2. Edit the contents of sphinx.conf.in - specifically entries relating to @CONFDIR@ - to paths suitable for your system.

  3. Install the searchd system as a Windows service:

    C:\Sphinx\bin> C:\Sphinx\bin\searchd --install --config C:\Sphinx\sphinx.conf.in --servicename SphinxSearch

  4. The searchd service will now be listed in the Services panel within the Management Console, available from Administrative Tools. It will not have been started, as you will need to configure it and build your indexes with indexer before starting the service. A guide to do this can be found under Quick tour.

    During the next steps of the install (which involve running indexer pretty much as you would on Linux) you may find that you get an error relating to libmysql.dll not being found. If you have MySQL installed, you should find a copy of this library in your Windows directory, or sometimes in Windows\System32, or failing that in the MySQL core directories. If you do receive an error please copy libmysql.dll into the bin directory.

2.6. Sphinx deprecations and changes in default configuration

In 2.2.1-beta version we decided to start removing some old features. All of them was 'unofficially' deprecated for some time. And we're informing you now about it.

Changes are as follows:

  • 32-bit document IDs are now deprecated. Our binary releases are now all built with 64-bit IDs by default. Note that they can still load older indexes with 32-bit IDs, but that support will eventually be removed. In fact, that was deprecated awhile ago, but now we just want to make it clear: we don't see any sense in trying to save your server's RAM this way.

  • dict=crc is now deprecated. It has a bunch of limitations, the most important ones being keyword collisions, and no (good) wildcard matching support. You can read more about those limitations in our documentation.

  • charset_type=sbcs is now deprecated, we're slowly switching to UTF-only. Even if your database is SBCS (likely for legacy reasons too, eh?), this should be absolutely trivial to workaround, just add a pre-query to fetch your data in UTF-8 and you're all set. Also, in fact, our current UTF-8 tokenizer is even faster than the SBCS one.

  • custom sort (@custom) is now removed from Sphinx. This feature was introduced long before sort by expression became a reality and it has been deprecated for a very long time.

  • enable_star is deprecated now. Previous default mode was enable_star=0 which was due to compatibility with a very old Sphinx version. Such implicit star search isn't very intuitive. So, we've decided to eventually remove it and have marked it as deprecated just recently. We plan to totally remove this configuration key in the 2.2.X branch.

  • str2ordinal attributes are deprecated. This feature allows you to perform sorting by a string. But it's also possible to do this with ordinary string attributes, which is much easier to use. str2ordinal only covers a small part of this functionality and is not needed now.

  • str2wordcount attributes are deprecated. index_field_lengths=1 will create an integer attribute with field length set automatically and we recommend to use this configuration key when you need to store field lengths. Also, index_field_lengths=1 allows you to use new ranking formulas like BM25F().

  • hit_format is deprecated. This is a hidden configuration key - it's not mentioned in our documentation. But, it's there and it's possible that someone may use it. And now we're urging you: don't use it. The default value is 'inline' and it's a new standard. 'plain' hit_format is obsolete and will be removed in the near future.

  • docinfo=inline is deprecated. You can now use ondisk_attrs or ondisk_attrs_default instead.

  • workers=threads is a new default for all OS now. We're gonna get rid of other modes in future.

  • mem_limit=128M is a new default.

  • rt_mem_limit=128M is a new default.

  • ondisk_dict is deprecated. No need to save RAM this way.

  • ondisk_dict_default is deprecated. No need to save RAM this way.

  • compat_sphinxql_magics was removed. Now you can't use an old result format and SphinxQL always looks more like ANSI SQL.

  • Completely removed xmlpipe. This was a very old ad hoc solution for a particular customer. xmlpipe2 surpasses it in every single aspect.

None of the different querying methods are deprecated, but as of version 2.2.1-beta, SphinxQL is the most advanced method. We plan to remove SphinxAPI and Sphinx SE someday so it would be a good idea to start using SphinxQL.

  • The SetWeights() API call has been deprecated for a long time and has now been removed from official APIs.

  • The default matching mode for the API is now 'extended'. Actually, all other modes are deprecated. We recommend using the extended query syntax instead.

Changes for 2.2.2-beta:

  • Removed deprecated "address" and "port" directives. Use "listen" instead.

  • Removed str2wordcount attributes. Use index_field_lengths=1 instead.

  • Removed str2ordinal attributes. Use string attributes for sorting.

  • ondisk_dict and ondisk_dict_default was removed.

  • Removed charset_type and mssql_unicode - we now support only UTF-8 encoding.

  • Removed deprecated enable_star. Now always work as with enable_star=1.

  • Removed CLI search which confused people instead of helping them and sql_query_info.

  • Deprecated SetMatchMode() API call.

  • Changed default thread_stack value to 1M.

  • Deprecated SetOverride() API call.

Changes for 2.2.3-beta:

  • Removed unneeded max_matches key from config file.

2.7. Quick Sphinx usage tour

All the example commands below assume that you installed Sphinx in /usr/local/sphinx, so searchd can be found in /usr/local/sphinx/bin/searchd.

To use Sphinx, you will need to:

  1. Create a configuration file.

    Default configuration file name is sphinx.conf. All Sphinx programs look for this file in current working directory by default.

    Sample configuration file, sphinx.conf.dist, which has all the options documented, is created by configure. Copy and edit that sample file to make your own configuration: (assuming Sphinx is installed into /usr/local/sphinx/)

    $ cd /usr/local/sphinx/etc
    $ cp sphinx.conf.dist sphinx.conf
    $ vi sphinx.conf

    Sample configuration file is setup to index documents table from MySQL database test; so there's example.sql sample data file to populate that table with a few documents for testing purposes:

    $ mysql -u test < /usr/local/sphinx/etc/example.sql

  2. Run the indexer to create full-text index from your data:

    $ cd /usr/local/sphinx/etc
    $ /usr/local/sphinx/bin/indexer --all

  3. Query your newly created index!

Now query your indexes!

Connect to server:

$ mysql -h0 -P9306

SELECT * FROM test1 WHERE MATCH('my document');

INSERT INTO rt VALUES (1, 'this is', 'a sample text', 11);

INSERT INTO rt VALUES (2, 'some more', 'text here', 22);

SELECT gid/11 FROM rt WHERE MATCH('text') GROUP BY gid;



SELECT *, WEIGHT() FROM test1 WHERE MATCH('"document one"/1');SHOW META;

SET profiling=1;SELECT * FROM test1 WHERE id IN (1,2,4);SHOW PROFILE;

SELECT id, id%3 idd FROM test1 WHERE MATCH('this is | nothing') GROUP BY idd;SHOW PROFILE;

SELECT id FROM test1 WHERE MATCH('is this a good plan?');SHOW PLAN;

SELECT COUNT(*) c, id%3 idd FROM test1 GROUP BY idd HAVING COUNT(*)>1;


CALL KEYWORDS ('one two three', 'test1');

CALL KEYWORDS ('one two three', 'test1', 1);

Happy searching!

Chapter 3. Indexing

3.1. Data sources

The data to be indexed can generally come from very different sources: SQL databases, plain text files, HTML files, mailboxes, and so on. From Sphinx point of view, the data it indexes is a set of structured documents, each of which has the same set of fields and attributes. This is similar to SQL, where each row would correspond to a document, and each column to either a field or an attribute.

Depending on what source Sphinx should get the data from, different code is required to fetch the data and prepare it for indexing. This code is called data source driver (or simply driver or data source for brevity).

At the time of this writing, there are built-in drivers for MySQL, PostgreSQL, MS SQL (on Windows), and ODBC. There is also a generic driver called xmlpipe2, which runs a specified command and reads the data from its stdout. See Section 3.9, “xmlpipe2 data source” section for the format description. In 2.2.1-beta a tsvpipe (Tab Separated Values) and csvpipe (Comma Separated Values) data source was added. You can get more information here Section 3.10, “tsvpipe\csvpipe (Tab\Comma Separated Values) data source”.

There can be as many sources per index as necessary. They will be sequentially processed in the very same order which was specified in index definition. All the documents coming from those sources will be merged as if they were coming from a single source.

3.2. Full-text fields

Full-text fields (or just fields for brevity) are the textual document contents that get indexed by Sphinx, and can be (quickly) searched for keywords.

Fields are named, and you can limit your searches to a single field (eg. search through "title" only) or a subset of fields (eg. to "title" and "abstract" only). Sphinx index format generally supports up to 256 fields. However, up to version 2.0.1-beta indexes were forcibly limited by 32 fields, because of certain complications in the matching engine. Full support for up to 256 fields was added in version 2.0.2-beta.

Note that the original contents of the fields are not stored in the Sphinx index. The text that you send to Sphinx gets processed, and a full-text index (a special data structure that enables quick searches for a keyword) gets built from that text. But the original text contents are then simply discarded. Sphinx assumes that you store those contents elsewhere anyway.

Moreover, it is impossible to fully reconstruct the original text, because the specific whitespace, capitalization, punctuation, etc will all be lost during indexing. It is theoretically possible to partially reconstruct a given document from the Sphinx full-text index, but that would be a slow process (especially if the CRC dictionary is used, which does not even store the original keywords and works with their hashes instead).

3.3. Attributes

Attributes are additional values associated with each document that can be used to perform additional filtering and sorting during search.

It is often desired to additionally process full-text search results based not only on matching document ID and its rank, but on a number of other per-document values as well. For instance, one might need to sort news search results by date and then relevance, or search through products within specified price range, or limit blog search to posts made by selected users, or group results by month. To do that efficiently, Sphinx allows to attach a number of additional attributes to each document, and store their values in the full-text index. It's then possible to use stored values to filter, sort, or group full-text matches.

Attributes, unlike the fields, are not full-text indexed. They are stored in the index, but it is not possible to search them as full-text, and attempting to do so results in an error.

For example, it is impossible to use the extended matching mode expression @column 1 to match documents where column is 1, if column is an attribute, and this is still true even if the numeric digits are normally indexed.

Attributes can be used for filtering, though, to restrict returned rows, as well as sorting or result grouping; it is entirely possible to sort results purely based on attributes, and ignore the search relevance tools. Additionally, attributes are returned from the search daemon, while the indexed text is not.

A good example for attributes would be a forum posts table. Assume that only title and content fields need to be full-text searchable - but that sometimes it is also required to limit search to a certain author or a sub-forum (ie. search only those rows that have some specific values of author_id or forum_id columns in the SQL table); or to sort matches by post_date column; or to group matching posts by month of the post_date and calculate per-group match counts.

This can be achieved by specifying all the mentioned columns (excluding title and content, that are full-text fields) as attributes, indexing them, and then using API calls to setup filtering, sorting, and grouping. Here as an example.

Example sphinx.conf part:

sql_query = SELECT id, title, content, \
    author_id, forum_id, post_date FROM my_forum_posts
sql_attr_uint = author_id
sql_attr_uint = forum_id
sql_attr_timestamp = post_date

Example application code (in PHP):

// only search posts by author whose ID is 123
$cl->SetFilter ( "author_id", array ( 123 ) );

// only search posts in sub-forums 1, 3 and 7
$cl->SetFilter ( "forum_id", array ( 1,3,7 ) );

// sort found posts by posting date in descending order
$cl->SetSortMode ( SPH_SORT_ATTR_DESC, "post_date" );

Attributes are named. Attribute names are case insensitive. Attributes are not full-text indexed; they are stored in the index as is. Currently supported attribute types are:

  • unsigned integers (1-bit to 32-bit wide);

  • UNIX timestamps;

  • floating point values (32-bit, IEEE 754 single precision);

  • strings (since 1.10-beta);

  • JSON (since 2.1.1-beta);

  • MVA, multi-value attributes (variable-length lists of 32-bit unsigned integers).

The complete set of per-document attribute values is sometimes referred to as docinfo. Docinfos can either be

  • stored separately from the main full-text index data ("extern" storage, in .spa file), or

  • attached to each occurrence of document ID in full-text index data ("inline" storage, in .spd file).

When using extern storage, a copy of .spa file (with all the attribute values for all the documents) is kept in RAM by searchd at all times. This is for performance reasons; random disk I/O would be too slow. On the contrary, inline storage does not require any additional RAM at all, but that comes at the cost of greatly inflating the index size: remember that it copies all attribute value every time when the document ID is mentioned, and that is exactly as many times as there are different keywords in the document. Inline may be the only viable option if you have only a few attributes and need to work with big datasets in limited RAM. However, in most cases extern storage makes both indexing and searching much more efficient.

Search-time memory requirements for extern storage are (1+number_of_attrs)*number_of_docs*4 bytes, ie. 10 million docs with 2 groups and 1 timestamp will take (1+2+1)*10M*4 = 160 MB of RAM. This is PER DAEMON, not per query. searchd will allocate 160 MB on startup, read the data and keep it shared between queries. The children will NOT allocate any additional copies of this data.

3.4. MVA (multi-valued attributes)

MVAs, or multi-valued attributes, are an important special type of per-document attributes in Sphinx. MVAs let you attach sets of numeric values to every document. That is useful to implement article tags, product categories, etc. Filtering and group-by (but not sorting) on MVA attributes is supported.

As of version 2.0.2-beta, MVA values can either be unsigned 32-bit integers (UNSIGNED INTEGER) or signed 64-bit integers (BIGINT). Up to version 2.0.1-beta, only the unsigned 32-bit values were supported.

The set size is not limited, you can have an arbitrary number of values attached to each document as long as RAM permits (.spm file that contains the MVA values will be precached in RAM by searchd). The source data can be taken either from a separate query, or from a document field; see source type in sql_attr_multi. In the first case the query will have to return pairs of document ID and MVA values, in the second one the field will be parsed for integer values. There are absolutely no requirements as to incoming data order; the values will be automatically grouped by document ID (and internally sorted within the same ID) during indexing anyway.

When filtering, a document will match the filter on MVA attribute if any of the values satisfy the filtering condition. (Therefore, documents that pass through exclude filters will not contain any of the forbidden values.) When grouping by MVA attribute, a document will contribute to as many groups as there are different MVA values associated with that document. For instance, if the collection contains exactly 1 document having a 'tag' MVA with values 5, 7, and 11, grouping on 'tag' will produce 3 groups with 'COUNT(*)' equal to 1 and 'GROUPBY()' key values of 5, 7, and 11 respectively. Also note that grouping by MVA might lead to duplicate documents in the result set: because each document can participate in many groups, it can be chosen as the best one in in more than one group, leading to duplicate IDs. PHP API historically uses ordered hash on the document ID for the resulting rows; so you'll also need to use SetArrayResult() in order to employ group-by on MVA with PHP API.

3.5. Indexes

To be able to answer full-text search queries fast, Sphinx needs to build a special data structure optimized for such queries from your text data. This structure is called index; and the process of building index from text is called indexing.

Different index types are well suited for different tasks. For example, a disk-based tree-based index would be easy to update (ie. insert new documents to existing index), but rather slow to search. Sphinx architecture allows internally for different index types, or backends, to be implemented comparatively easily.

Starting with 1.10-beta, Sphinx provides 2 different backends: a disk index backend, and a RT (realtime) index backend.

Disk indexes are designed to provide maximum indexing and searching speed, while keeping the RAM footprint as low as possible. That comes at a cost of text index updates. You can not update an existing document or incrementally add a new document to a disk index. You only can batch rebuild the entire disk index from scratch. (Note that you still can update document's attributes on the fly, even with the disk indexes.)

This "rebuild only" limitation might look as a big constraint at a first glance. But in reality, it can very frequently be worked around rather easily by setting up multiple disk indexes, searching through them all, and only rebuilding the one with a fraction of the most recently changed data. See Section 3.11, “Live index updates” for details.

RT indexes enable you to implement dynamic updates and incremental additions to the full text index. RT stands for Real Time and they are indeed "soft realtime" in terms of writes, meaning that most index changes become available for searching as quick as 1 millisecond or less, but could occasionally stall for seconds. (Searches will still work even during that occasional writing stall.) Refer to Chapter 4, Real-time indexes for details.

Last but not least, Sphinx supports so-called distributed indexes. Compared to disk and RT indexes, those are not a real physical backend, but rather just lists of either local or remote indexes that can be searched transparently to the application, with Sphinx doing all the chores of sending search requests to remote machines in the cluster, aggregating the result sets, retrying the failed requests, and even doing some load balancing. See Section 5.8, “Distributed searching” for a discussion of distributed indexes.

There can be as many indexes per configuration file as necessary. indexer utility can reindex either all of them (if --all option is specified), or a certain explicitly specified subset. searchd utility will serve all the specified indexes, and the clients can specify what indexes to search in run time.

3.6. Restrictions on the source data

There are a few different restrictions imposed on the source data which is going to be indexed by Sphinx, of which the single most important one is:


If this requirement is not met, different bad things can happen. For instance, Sphinx can crash with an internal assertion while indexing; or produce strange results when searching due to conflicting IDs. Also, a 1000-pound gorilla might eventually come out of your display and start throwing barrels at you. You've been warned.

3.7. Charsets, case folding, translation tables, and replacement rules

When indexing some index, Sphinx fetches documents from the specified sources, splits the text into words, and does case folding so that "Abc", "ABC" and "abc" would be treated as the same word (or, to be pedantic, term).

To do that properly, Sphinx needs to know

  • what encoding is the source text in (and this encoding should always be UTF-8);

  • what characters are letters and what are not;

  • what letters should be folded to what letters.

This should be configured on a per-index basis using charset_table option. charset_table specifies the table that maps letter characters to their case folded versions. The characters that are not in the table are considered to be non-letters and will be treated as word separators when indexing or searching through this index.

Default tables currently include English and Russian characters. Please do submit your tables for other languages!

As of version 2.1.1-beta, you can also specify text pattern replacement rules. For example, given the rules

regexp_filter = \b(\d+)\" => \1 inch
regexp_filter = (BLUE|RED) => COLOR

the text 'RED TUBE 5" LONG' would be indexed as 'COLOR TUBE 5 INCH LONG', and 'PLANK 2" x 4"' as 'PLANK 2 INCH x 4 INCH'. Rules are applied in the given order. Text in queries is also replaced; a search for "BLUE TUBE" would actually become a search for "COLOR TUBE". Note that Sphinx must be built with the --with-re2 option to use this feature.

3.8. SQL data sources (MySQL, PostgreSQL)

With all the SQL drivers, indexing generally works as follows.

Most options, such as database user/host/password, are straightforward. However, there are a few subtle things, which are discussed in more detail here.

Ranged queries

Main query, which needs to fetch all the documents, can impose a read lock on the whole table and stall the concurrent queries (eg. INSERTs to MyISAM table), waste a lot of memory for result set, etc. To avoid this, Sphinx supports so-called ranged queries. With ranged queries, Sphinx first fetches min and max document IDs from the table, and then substitutes different ID intervals into main query text and runs the modified query to fetch another chunk of documents. Here's an example.

Example 3.1. Ranged query usage example

# in sphinx.conf

sql_query_range = SELECT MIN(id),MAX(id) FROM documents
sql_range_step = 1000
sql_query = SELECT * FROM documents WHERE id>=$start AND id<=$end

If the table contains document IDs from 1 to, say, 2345, then sql_query would be run three times:

  1. with $start replaced with 1 and $end replaced with 1000;

  2. with $start replaced with 1001 and $end replaced with 2000;

  3. with $start replaced with 2001 and $end replaced with 2345.

Obviously, that's not much of a difference for 2000-row table, but when it comes to indexing 10-million-row MyISAM table, ranged queries might be of some help.

sql_query_post vs. sql_query_post_index

The difference between post-query and post-index query is in that post-query is run immediately when Sphinx received all the documents, but further indexing may still fail for some other reason. On the contrary, by the time the post-index query gets executed, it is guaranteed that the indexing was successful. Database connection is dropped and re-established because sorting phase can be very lengthy and would just timeout otherwise.

3.9. xmlpipe2 data source

xmlpipe2 lets you pass arbitrary full-text and attribute data to Sphinx in yet another custom XML format. It also allows to specify the schema (ie. the set of fields and attributes) either in the XML stream itself, or in the source settings.

When indexing xmlpipe2 source, indexer runs the given command, opens a pipe to its stdout, and expects well-formed XML stream. Here's sample stream data:

Example 3.2. xmlpipe2 document stream

<?xml version="1.0" encoding="utf-8"?>

<sphinx:field name="subject"/>
<sphinx:field name="content"/>
<sphinx:attr name="published" type="timestamp"/>
<sphinx:attr name="author_id" type="int" bits="16" default="1"/>

<sphinx:document id="1234">
<content>this is the main content <![CDATA[[and this <cdata> entry
must be handled properly by xml parser lib]]></content>
<subject>note how field/attr tags can be
in <b class="red">randomized</b> order</subject>
<misc>some undeclared element</misc>

<sphinx:document id="1235">
<subject>another subject</subject>
<content>here comes another document, and i am given to understand,
that in-document field order must not matter, sir</content>

<!-- ... even more sphinx:document entries here ... -->



Arbitrary fields and attributes are allowed. They also can occur in the stream in arbitrary order within each document; the order is ignored. There is a restriction on maximum field length; fields longer than 2 MB will be truncated to 2 MB (this limit can be changed in the source).

The schema, ie. complete fields and attributes list, must be declared before any document could be parsed. This can be done either in the configuration file using xmlpipe_field and xmlpipe_attr_XXX settings, or right in the stream using <sphinx:schema> element. <sphinx:schema> is optional. It is only allowed to occur as the very first sub-element in <sphinx:docset>. If there is no in-stream schema definition, settings from the configuration file will be used. Otherwise, stream settings take precedence.

Unknown tags (which were not declared neither as fields nor as attributes) will be ignored with a warning. In the example above, <misc> will be ignored. All embedded tags and their attributes (such as <b> in <subject> in the example above) will be silently ignored.

Support for incoming stream encodings depends on whether iconv is installed on the system. xmlpipe2 is parsed using libexpat parser that understands US-ASCII, ISO-8859-1, UTF-8 and a few UTF-16 variants natively. Sphinx configure script will also check for libiconv presence, and utilize it to handle other encodings. libexpat also enforces the requirement to use UTF-8 charset on Sphinx side, because the parsed data it returns is always in UTF-8.

XML elements (tags) recognized by xmlpipe2 (and their attributes where applicable) are:


Mandatory top-level element, denotes and contains xmlpipe2 document set.


Optional element, must either occur as the very first child of sphinx:docset, or never occur at all. Declares the document schema. Contains field and attribute declarations. If present, overrides per-source settings from the configuration file.


Optional element, child of sphinx:schema. Declares a full-text field. Known attributes are:

  • "name", specifies the XML element name that will be treated as a full-text field in the subsequent documents.

  • "attr", specifies whether to also index this field as a string. Possible value is "string". Introduced in version 1.10-beta.


Optional element, child of sphinx:schema. Declares an attribute. Known attributes are:

  • "name", specifies the element name that should be treated as an attribute in the subsequent documents.

  • "type", specifies the attribute type. Possible values are "int", "bigint", "timestamp", "bool", "float", "multi" and "json".

  • "bits", specifies the bit size for "int" attribute type. Valid values are 1 to 32.

  • "default", specifies the default value for this attribute that should be used if the attribute's element is not present in the document.


Mandatory element, must be a child of sphinx:docset. Contains arbitrary other elements with field and attribute values to be indexed, as declared either using sphinx:field and sphinx:attr elements or in the configuration file. The only known attribute is "id" that must contain the unique integer document ID.


Optional element, child of sphinx:docset. Contains a number of "id" elements whose contents are document IDs to be put into a kill-list for this index.

3.10. tsvpipe\csvpipe (Tab\Comma Separated Values) data source

This is the simplest way to pass data to the indexer. It was created due to xmlpipe2 limitations. Namely, indexer must map each attribute and field tag in XML file to corresponding schema element. This mapping requires some time. And time increases with increasing the number of fields and attributes in schema. There is no such issue in tsvpipe because each field and attribute is a particular column in TSV file. So, in some cases tsvpipe could work slightly faster than xmlpipe2. Added in 2.2.1-beta.

The first column in TSV\CSV file must be a document ID. The rest ones must mirror the declaration of fields and attributes in schema definition.

The difference between tsvpipe and csvpipe is delimiter and quoting rules. tsvpipe has tab character as hardcoded delimiter and has no quoting rules. csvpipe has option csvpipe_delimiter for delimiter with default value ',' and also has quoting rules, such as:

  • any field may be quoted

  • fields containing a line-break, double-quote or commas should be quoted

  • a double quote character in a field must be represented by two double quote characters

tsvpipe and csvpipe have same field and attrribute declaration derectives as xmlpipe.

tsvpipe declarations:

tsvpipe_command, tsvpipe_field, tsvpipe_field_string, tsvpipe_attr_uint, tsvpipe_attr_timestamp, tsvpipe_attr_bool, tsvpipe_attr_float, tsvpipe_attr_bigint, tsvpipe_attr_multi, tsvpipe_attr_multi_64, tsvpipe_attr_string, tsvpipe_attr_json

csvpipe declarations:

csvpipe_command, csvpipe_field, csvpipe_field_string, csvpipe_attr_uint, csvpipe_attr_timestamp, csvpipe_attr_bool, csvpipe_attr_float, csvpipe_attr_bigint, csvpipe_attr_multi, csvpipe_attr_multi_64, csvpipe_attr_string, csvpipe_attr_json

source tsv_test
	type = tsvpipe
	tsvpipe_command = cat /tmp/rock_bands.tsv
	tsvpipe_field = name
	tsvpipe_attr_multi = genre_tags
1	Led Zeppelin	35,23,16
2	Deep Purple	35,92
3	Frank Zappa	35,23,16,92,33,24

3.11. Live index updates

There are two major approaches to maintaining the full-text index contents up to date. Note, however, that both these approaches deal with the task of full-text data updates, and not attribute updates. Instant attribute updates are supported since version 0.9.8. Refer to UpdateAttributes() API call description for details.

First, you can use disk-based indexes, partition them manually, and only rebuild the smaller partitions (so-called "deltas") frequently. By minimizing the rebuild size, you can reduce the average indexing lag to something as low as 30-60 seconds. This approach was the the only one available in versions 0.9.x. On huge collections it actually might be the most efficient one. Refer to Section 3.12, “Delta index updates” for details.

Second, versions 1.x (starting with 1.10-beta) add support for so-called real-time indexes (RT indexes for short) that on-the-fly updates of the full-text data. Updates on a RT index can appear in the search results in 1-2 milliseconds, ie. 0.001-0.002 seconds. However, RT index are less efficient for bulk indexing huge amounts of data. Refer to Chapter 4, Real-time indexes for details.

3.12. Delta index updates

There's a frequent situation when the total dataset is too big to be reindexed from scratch often, but the amount of new records is rather small. Example: a forum with a 1,000,000 archived posts, but only 1,000 new posts per day.

In this case, "live" (almost real time) index updates could be implemented using so called "main+delta" scheme.

The idea is to set up two sources and two indexes, with one "main" index for the data which only changes rarely (if ever), and one "delta" for the new documents. In the example above, 1,000,000 archived posts would go to the main index, and newly inserted 1,000 posts/day would go to the delta index. Delta index could then be reindexed very frequently, and the documents can be made available to search in a matter of minutes.

Specifying which documents should go to what index and reindexing main index could also be made fully automatic. One option would be to make a counter table which would track the ID which would split the documents, and update it whenever the main index is reindexed.

Example 3.3. Fully automated live updates

# in MySQL
CREATE TABLE sph_counter
    max_doc_id INTEGER NOT NULL

# in sphinx.conf
source main
    # ...
    sql_query_pre = SET NAMES utf8
    sql_query_pre = REPLACE INTO sph_counter SELECT 1, MAX(id) FROM documents
    sql_query = SELECT id, title, body FROM documents \
        WHERE id<=( SELECT max_doc_id FROM sph_counter WHERE counter_id=1 )

source delta : main
    sql_query_pre = SET NAMES utf8
    sql_query = SELECT id, title, body FROM documents \
        WHERE id>( SELECT max_doc_id FROM sph_counter WHERE counter_id=1 )

index main
    source = main
    path = /path/to/main
    # ... all the other settings

# note how all other settings are copied from main,
# but source and path are overridden (they MUST be)
index delta : main
    source = delta
    path = /path/to/delta

Note how we're overriding sql_query_pre in the delta source. We need to explicitly have that override. Otherwise REPLACE query would be run when indexing delta source too, effectively nullifying it. However, when we issue the directive in the inherited source for the first time, it removes all inherited values, so the encoding setup is also lost. So sql_query_pre in the delta can not just be empty; and we need to issue the encoding setup query explicitly once again.

3.13. Index merging

Merging two existing indexes can be more efficient that indexing the data from scratch, and desired in some cases (such as merging 'main' and 'delta' indexes instead of simply reindexing 'main' in 'main+delta' partitioning scheme). So indexer has an option to do that. Merging the indexes is normally faster than reindexing but still not instant on huge indexes. Basically, it will need to read the contents of both indexes once and write the result once. Merging 100 GB and 1 GB index, for example, will result in 202 GB of IO (but that's still likely less than the indexing from scratch requires).

The basic command syntax is as follows:

indexer --merge DSTINDEX SRCINDEX [--rotate]

Only the DSTINDEX index will be affected: the contents of SRCINDEX will be merged into it. --rotate switch will be required if DSTINDEX is already being served by searchd. The initially devised usage pattern is to merge a smaller update from SRCINDEX into DSTINDEX. Thus, when merging the attributes, values from SRCINDEX will win if duplicate document IDs are encountered. Note, however, that the "old" keywords will not be automatically removed in such cases. For example, if there's a keyword "old" associated with document 123 in DSTINDEX, and a keyword "new" associated with it in SRCINDEX, document 123 will be found by both keywords after the merge. You can supply an explicit condition to remove documents from DSTINDEX to mitigate that; the relevant switch is --merge-dst-range:

indexer --merge main delta --merge-dst-range deleted 0 0

This switch lets you apply filters to the destination index along with merging. There can be several filters; all of their conditions must be met in order to include the document in the resulting merged index. In the example above, the filter passes only those records where 'deleted' is 0, eliminating all records that were flagged as deleted (for instance, using UpdateAttributes() call).

Chapter 4. Real-time indexes

Real-time indexes (or RT indexes for brevity) are a new backend that lets you insert, update, or delete documents (rows) on the fly. RT indexes were added in version 1.10-beta. While querying of RT indexes is possible using any of the SphinxAPI, SphinxQL, or SphinxSE, updating them is only possible via SphinxQL at the moment. Full SphinxQL reference is available in Chapter 8, SphinxQL reference.

4.1. RT indexes overview

RT indexes should be declared in sphinx.conf, just as every other index type. Notable differences from the regular, disk-based indexes are that a) data sources are not required and ignored, and b) you should explicitly enumerate all the text fields, not just attributes. Here's an example:

Example 4.1. RT index declaration

index rt
    type = rt
    path = /usr/local/sphinx/data/rt
    rt_field = title
    rt_field = content
    rt_attr_uint = gid

As of 2.0.1-beta and above, RT indexes are production quality, despite a few missing features.

RT index can be accessed using MySQL protocol. INSERT, REPLACE, DELETE, and SELECT statements against RT index are supported. For instance, this is an example session with the sample index above:

$ mysql -h -P 9306
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 1
Server version: 1.10-dev (r2153)

Type 'help;' or '\h' for help. Type '\c' to clear the buffer.

mysql> INSERT INTO rt VALUES ( 1, 'first record', 'test one', 123 );
Query OK, 1 row affected (0.05 sec)

mysql> INSERT INTO rt VALUES ( 2, 'second record', 'test two', 234 );
Query OK, 1 row affected (0.00 sec)

mysql> SELECT * FROM rt;
| id   | weight | gid  |
|    1 |      1 |  123 |
|    2 |      1 |  234 |
2 rows in set (0.02 sec)

mysql> SELECT * FROM rt WHERE MATCH('test');
| id   | weight | gid  |
|    1 |   1643 |  123 |
|    2 |   1643 |  234 |
2 rows in set (0.01 sec)

mysql> SELECT * FROM rt WHERE MATCH('@title test');
Empty set (0.00 sec)

Both partial and batch INSERT syntaxes are supported, ie. you can specify a subset of columns, and insert several rows at a time. Deletions are also possible using DELETE statement; the only currently supported syntax is DELETE FROM <index> WHERE id=<id>. REPLACE is also supported, enabling you to implement updates.

mysql> INSERT INTO rt ( id, title ) VALUES ( 3, 'third row' ), ( 4, 'fourth entry' );
Query OK, 2 rows affected (0.01 sec)

mysql> SELECT * FROM rt;
| id   | weight | gid  |
|    1 |      1 |  123 |
|    2 |      1 |  234 |
|    3 |      1 |    0 |
|    4 |      1 |    0 |
4 rows in set (0.00 sec)

mysql> DELETE FROM rt WHERE id=2;
Query OK, 0 rows affected (0.00 sec)

mysql> SELECT * FROM rt WHERE MATCH('test');
| id   | weight | gid  |
|    1 |   1500 |  123 |
1 row in set (0.00 sec)

mysql> INSERT INTO rt VALUES ( 1, 'first record on steroids', 'test one', 123 );
ERROR 1064 (42000): duplicate id '1'

mysql> REPLACE INTO rt VALUES ( 1, 'first record on steroids', 'test one', 123 );
Query OK, 1 row affected (0.01 sec)

mysql> SELECT * FROM rt WHERE MATCH('steroids');
| id   | weight | gid  |
|    1 |   1500 |  123 |
1 row in set (0.01 sec)

Data stored in RT index should survive clean shutdown. When binary logging is enabled, it should also survive crash and/or dirty shutdown, and recover on subsequent startup.

4.2. Known caveats with RT indexes

RT indexes are currently quality feature, but there are still a few known usage quirks. Those quirks are listed in this section.

  • Default conservative RAM chunk limit (rt_mem_limit) of 32M can lead to poor performance on bigger indexes, you should raise it to 256..1024M if you're planning to index gigabytes.

  • High DELETE/REPLACE rate can lead to kill-list fragmentation and impact searching performance.

  • No transaction size limits are currently imposed; too many concurrent INSERT/REPLACE transactions might therefore consume a lot of RAM.

  • In case of a damaged binlog, recovery will stop on the first damaged transaction, even though it's technically possible to keep looking further for subsequent undamaged transactions, and recover those. This mid-file damage case (due to flaky HDD/CDD/tape?) is supposed to be extremely rare, though.

  • Multiple INSERTs grouped in a single transaction perform better than equivalent single-row transactions and are recommended for batch loading of data.

4.3. RT index internals

RT index is internally chunked. It keeps a so-called RAM chunk that stores all the most recent changes. RAM chunk memory usage is rather strictly limited with per-index rt_mem_limit directive. Once RAM chunk grows over this limit, a new disk chunk is created from its data, and RAM chunk is reset. Thus, while most changes on the RT index will be performed in RAM only and complete instantly (in milliseconds), those changes that overflow the RAM chunk will stall for the duration of disk chunk creation (a few seconds).

Since version 2.1.1-beta, Sphinx uses double-buffering to avoid INSERT stalls. When data is being dumped to disk, the second buffer is used, so further INSERTs won't be delayed. The second buffer is defined to be 10% the size of the standard buffer, rt_mem_limit, but future versions of Sphinx may allow configuring this further.

Disk chunks are, in fact, just regular disk-based indexes. But they're a part of an RT index and automatically managed by it, so you need not configure nor manage them manually. Because a new disk chunk is created every time RT chunk overflows the limit, and because in-memory chunk format is close to on-disk format, the disk chunks will be approximately rt_mem_limit bytes in size each.

Generally, it is better to set the limit bigger, to minimize both the frequency of flushes, and the index fragmentation (number of disk chunks). For instance, on a dedicated search server that handles a big RT index, it can be advised to set rt_mem_limit to 1-2 GB. A global limit on all indexes is also planned, but not yet implemented yet as of 1.10-beta.

Disk chunk full-text index data can not be actually modified, so the full-text field changes (ie. row deletions and updates) suppress a previous row version from a disk chunk using a kill-list, but do not actually physically purge the data. Therefore, on workloads with high full-text updates ratio index might eventually get polluted by these previous row versions, and searching performance would degrade. Physical index purging that would improve the performance is planned, but not yet implemented as of 1.10-beta.

Data in RAM chunk gets saved to disk on clean daemon shutdown, and then loaded back on startup. However, on daemon or server crash, updates from RAM chunk might be lost. To prevent that, binary logging of transactions can be used; see Section 4.4, “Binary logging” for details.

Full-text changes in RT index are transactional. They are stored in a per-thread accumulator until COMMIT, then applied at once. Bigger batches per single COMMIT should result in faster indexing.

4.4. Binary logging

Binary logs are essentially a recovery mechanism. With binary logs enabled, searchd writes every given transaction to the binlog file, and uses that for recovery after an unclean shutdown. On clean shutdown, RAM chunks are saved to disk, and then all the binlog files are unlinked.

During normal operation, a new binlog file will be opened every time when binlog_max_log_size limit is reached. Older, already closed binlog files are kept until all of the transactions stored in them (from all indexes) are flushed as a disk chunk. Setting the limit to 0 pretty much prevents binlog from being unlinked at all while searchd is running; however, it will still be unlinked on clean shutdown. (This is the default case as of 2.0.3-release, binlog_max_log_size defaults to 0.)

There are 3 different binlog flushing strategies, controlled by binlog_flush directive which takes the values of 0, 1, or 2. 0 means to flush the log to OS and sync it to disk every second; 1 means flush and sync every transaction; and 2 (the default mode) means flush every transaction but sync every second. Sync is relatively slow because it has to perform physical disk writes, so mode 1 is the safest (every committed transaction is guaranteed to be written on disk) but the slowest. Flushing log to OS prevents from data loss on searchd crashes but not system crashes. Mode 2 is the default.

On recovery after an unclean shutdown, binlogs are replayed and all logged transactions since the last good on-disk state are restored. Transactions are checksummed so in case of binlog file corruption garbage data will not be replayed; such a broken transaction will be detected and, currently, will stop replay. Transactions also start with a magic marker and timestamped, so in case of binlog damage in the middle of the file, it's technically possible to skip broken transactions and keep replaying from the next good one, and/or it's possible to replay transactions until a given timestamp (point-in-time recovery), but none of that is implemented yet as of 1.10-beta.

One unwanted side effect of binlogs is that actively updating a small RT index that fully fits into a RAM chunk part will lead to an ever-growing binlog that can never be unlinked until clean shutdown. Binlogs are essentially append-only deltas against the last known good saved state on disk, and unless RAM chunk gets saved, they can not be unlinked. An ever-growing binlog is not very good for disk use and crash recovery time. Starting with 2.0.1-beta you can configure searchd to perform a periodic RAM chunk flush to fix that problem using a rt_flush_period directive. With periodic flushes enabled, searchd will keep a separate thread, checking whether RT indexes RAM chunks need to be written back to disk. Once that happens, the respective binlogs can be (and are) safely unlinked.

Note that rt_flush_period only controls the frequency at which the checks happen. There are no guarantees that the particular RAM chunk will get saved. For instance, it does not make sense to regularly re-save a huge RAM chunk that only gets a few rows worth of updates. The search daemon determine whether to actually perform the flush with a few heuristics.

Chapter 5. Searching

5.1. Matching modes

So-called matching modes are a legacy feature that used to provide (very) limited query syntax and ranking support. Currently, they are deprecated in favor of full-text query language and so-called rankers. Starting with version 0.9.9-release, it is thus strongly recommended to use SPH_MATCH_EXTENDED and proper query syntax rather than any other legacy mode. All those other modes are actually internally converted to extended syntax anyway. SphinxAPI still defaults to SPH_MATCH_ALL but that is for compatibility reasons only.

There are the following matching modes available:

  • SPH_MATCH_ALL, matches all query words;

  • SPH_MATCH_ANY, matches any of the query words;

  • SPH_MATCH_PHRASE, matches query as a phrase, requiring perfect match;

  • SPH_MATCH_BOOLEAN, matches query as a boolean expression (see Section 5.2, “Boolean query syntax”);

  • SPH_MATCH_EXTENDED, matches query as an expression in Sphinx internal query language (see Section 5.3, “Extended query syntax”);

  • SPH_MATCH_EXTENDED2, an alias for SPH_MATCH_EXTENDED (default mode);

  • SPH_MATCH_FULLSCAN, matches query, forcibly using the "full scan" mode as below. NB, any query terms will be ignored, such that filters, filter-ranges and grouping will still be applied, but no text-matching.

SPH_MATCH_EXTENDED2 was used during 0.9.8 and 0.9.9 development cycle, when the internal matching engine was being rewritten (for the sake of additional functionality and better performance). By 0.9.9-release, the older version was removed, and SPH_MATCH_EXTENDED and SPH_MATCH_EXTENDED2 are now just aliases.

The SPH_MATCH_FULLSCAN mode will be automatically activated in place of the specified matching mode when the following conditions are met:

  1. The query string is empty (ie. its length is zero).

  2. docinfo storage is set to extern.

In full scan mode, all the indexed documents will be considered as matching. Such queries will still apply filters, sorting, and group by, but will not perform any full-text searching. This can be useful to unify full-text and non-full-text searching code, or to offload SQL server (there are cases when Sphinx scans will perform better than analogous MySQL queries). An example of using the full scan mode might be to find posts in a forum. By selecting the forum's user ID via SetFilter() but not actually providing any search text, Sphinx will match every document (i.e. every post) where SetFilter() would match - in this case providing every post from that user. By default this will be ordered by relevancy, followed by Sphinx document ID in ascending order (earliest first).

5.2. Boolean query syntax

Boolean queries allow the following special operators to be used:

  • operator OR:

    hello | world
  • operator NOT:

    hello -world
    hello !world

  • grouping:

    ( hello world )

Here's an example query which uses all these operators:

Example 5.1. Boolean query example

( cat -dog ) | ( cat -mouse)

There always is implicit AND operator, so "hello world" query actually means "hello & world".

OR operator precedence is higher than AND, so "looking for cat | dog | mouse" means "looking for ( cat | dog | mouse )" and not "(looking for cat) | dog | mouse".

Since version 2.1.1-beta, queries may be automatically optimized if OPTION boolean_simplify=1 is specified. Some transformations performed by this optimization include:

  • Excess brackets: ((A | B) | C) becomes ( A | B | C ); ((A B) C) becomes ( A B C )

  • Excess AND NOT: ((A !N1) !N2) becomes (A !(N1 | N2))

  • Common NOT: ((A !N) | (B !N)) becomes ((A|B) !N)

  • Common Compound NOT: ((A !(N AA)) | (B !(N BB))) becomes (((A|B) !N) | (A !AA) | (B !BB)) if the cost of evaluating N is greater than the added together costs of evaluating A and B

  • Common subterm: ((A (N | AA)) | (B (N | BB))) becomes (((A|B) N) | (A AA) | (B BB)) if the cost of evaluating N is greater than the added together costs of evaluating A and B

  • Common keywords: (A | "A B"~N) becomes A; ("A B" | "A B C") becomes "A B"; ("A B"~N | "A B C"~N) becomes ("A B"~N)

  • Common phrase: ("X A B" | "Y A B") becomes (("X|Y") "A B")

  • Common AND NOT: ((A !X) | (A !Y) | (A !Z)) becomes (A !(X Y Z))

  • Common OR NOT: ((A !(N | N1)) | (B !(N | N2))) becomes (( (A !N1) | (B !N2) ) !N)

Note that optimizing the queries consumes CPU time, so for simple queries -or for hand-optimized queries- you'll do better with the default boolean_simplify=0 value. Simplifications are often better for complex queries, or algorithmically generated queries.

Queries like "-dog", which implicitly include all documents from the collection, can not be evaluated. This is both for technical and performance reasons. Technically, Sphinx does not always keep a list of all IDs. Performance-wise, when the collection is huge (ie. 10-100M documents), evaluating such queries could take very long.

5.3. Extended query syntax

The following special operators and modifiers can be used when using the extended matching mode:

  • operator OR:

    hello | world
  • operator MAYBE (introduced in verion 2.2.3-beta):

    hello MAYBE world
  • operator NOT:

    hello -world
    hello !world

  • field search operator:

    @title hello @body world
  • field position limit modifier (introduced in version 0.9.9-rc1):

    @body[50] hello
  • multiple-field search operator:

    @(title,body) hello world
  • ignore field search operator (will ignore any matches of 'hello world' from field 'title'):

    @!title hello world
  • ignore multiple-field search operator (if we have fields title, subject and body then @!(title) is equivalent to @(subject,body)):

    @!(title,body) hello world
  • all-field search operator:

    @* hello
  • phrase search operator:

    "hello world"
  • proximity search operator:

    "hello world"~10
  • quorum matching operator:

    "the world is a wonderful place"/3
  • strict order operator (aka operator "before"):

    aaa << bbb << ccc
  • exact form modifier (introduced in version 0.9.9-rc1):

    raining =cats and =dogs
  • field-start and field-end modifier (introduced in version 0.9.9-rc2):

    ^hello world$
  • keyword IDF boost modifier (introduced in version 2.2.3-beta):

    boosted^1.234 boostedfieldend$^1.234
  • NEAR, generalized proximity operator (introduced in version 2.0.1-beta):

    hello NEAR/3 world NEAR/4 "my test"
  • SENTENCE operator (introduced in version 2.0.1-beta):

    all SENTENCE words SENTENCE "in one sentence"
  • PARAGRAPH operator (introduced in version 2.0.1-beta):

    "Bill Gates" PARAGRAPH "Steve Jobs"
  • ZONE limit operator:


    only in these titles

  • ZONESPAN limit operator:


    only in a (single) title

Here's an example query that uses some of these operators:

Example 5.2. Extended matching mode: query example

"hello world" @title "example program"~5 @body python -(php|perl) @* code

The full meaning of this search is:

  • Find the words 'hello' and 'world' adjacently in any field in a document;

  • Additionally, the same document must also contain the words 'example' and 'program' in the title field, with up to, but not including, 5 words between the words in question; (E.g. "example PHP program" would be matched however "example script to introduce outside data into the correct context for your program" would not because two terms have 5 or more words between them)

  • Additionally, the same document must contain the word 'python' in the body field, but not contain either 'php' or 'perl';

  • Additionally, the same document must contain the word 'code' in any field.

There always is implicit AND operator, so "hello world" means that both "hello" and "world" must be present in matching document.

OR operator precedence is higher than AND, so "looking for cat | dog | mouse" means "looking for ( cat | dog | mouse )" and not "(looking for cat) | dog | mouse".

Field limit operator limits subsequent searching to a given field. Normally, query will fail with an error message if given field name does not exist in the searched index. However, that can be suppressed by specifying "@@relaxed" option at the very beginning of the query:

@@relaxed @nosuchfield my query

This can be helpful when searching through heterogeneous indexes with different schemas.

Field position limit, introduced in version 0.9.9-rc1, additionally restricts the searching to first N position within given field (or fields). For example, "@body[50] hello" will not match the documents where the keyword 'hello' occurs at position 51 and below in the body.

Proximity distance is specified in words, adjusted for word count, and applies to all words within quotes. For instance, "cat dog mouse"~5 query means that there must be less than 8-word span which contains all 3 words, ie. "CAT aaa bbb ccc DOG eee fff MOUSE" document will not match this query, because this span is exactly 8 words long.

Quorum matching operator introduces a kind of fuzzy matching. It will only match those documents that pass a given threshold of given words. The example above ("the world is a wonderful place"/3) will match all documents that have at least 3 of the 6 specified words. Operator is limited to 255 keywords. Instead of an absolute number, you can also specify a number between 0.0 and 1.0 (standing for 0% and 100%), and Sphinx will match only documents with at least the specified percentage of given words. The same example above could also have been written "the world is a wonderful place"/0.5 and it would match documents with at least 50% of the 6 words.

Strict order operator (aka operator "before"), introduced in version 0.9.9-rc2, will match the document only if its argument keywords occur in the document exactly in the query order. For instance, "black << cat" query (without quotes) will match the document "black and white cat" but not the "that cat was black" document. Order operator has the lowest priority. It can be applied both to just keywords and more complex expressions, ie. this is a valid query:

(bag of words) << "exact phrase" << red|green|blue

Exact form keyword modifier, introduced in version 0.9.9-rc1, will match the document only if the keyword occurred in exactly the specified form. The default behavior is to match the document if the stemmed keyword matches. For instance, "runs" query will match both the document that contains "runs" and the document that contains "running", because both forms stem to just "run" - while "=runs" query will only match the first document. Exact form operator requires index_exact_words option to be enabled. This is a modifier that affects the keyword and thus can be used within operators such as phrase, proximity, and quorum operators. Starting with 2.2.2-beta, it is possible to apply an exact form modifier to the phrase operator. It's really just syntax sugar - it adds an exact form modifier to all terms contained within the phrase.

="exact phrase"

Field-start and field-end keyword modifiers, introduced in version 0.9.9-rc2, will make the keyword match only if it occurred at the very start or the very end of a fulltext field, respectively. For instance, the query "^hello world$" (with quotes and thus combining phrase operator and start/end modifiers) will only match documents that contain at least one field that has exactly these two keywords.

Starting with 0.9.9-rc1, arbitrarily nested brackets and negations are allowed. However, the query must be possible to compute without involving an implicit list of all documents:

// correct query
aaa -(bbb -(ccc ddd))

// queries that are non-computable
aaa | -bbb

Starting with 2.2.2-beta, the phrase search operator may include a 'match any term' modifier. Terms within the phrase operator are position significant. When the 'match any term' modifier is implemented, the position of the subsequent terms from that phrase query will be shifted. Therefore, 'match any' has no impact on search performance.

"exact * phrase * * for terms"

NEAR operator, added in 2.0.1-beta, is a generalized version of a proximity operator. The syntax is NEAR/N, it is case-sensitive, and no spaces are allowed between the NEAR keyword, the slash sign, and the distance value.

The original proximity operator only worked on sets of keywords. NEAR is more generic and can accept arbitrary subexpressions as its two arguments, matching the document when both subexpressions are found within N words of each other, no matter in which order. NEAR is left associative and has the same (lowest) precedence as BEFORE.

You should also note how a (one NEAR/7 two NEAR/7 three) query using NEAR is not really equivalent to a ("one two three"~7) one using keyword proximity operator. The difference here is that the proximity operator allows for up to 6 non-matching words between all the 3 matching words, but the version with NEAR is less restrictive: it would allow for up to 6 words between 'one' and 'two' and then for up to 6 more between that two-word matching and a 'three' keyword.

SENTENCE and PARAGRAPH operators, added in 2.0.1-beta, matches the document when both its arguments are within the same sentence or the same paragraph of text, respectively. The arguments can be either keywords, or phrases, or the instances of the same operator. Here are a few examples:

one SENTENCE two
one SENTENCE "two three"
one SENTENCE "two three" SENTENCE four

The order of the arguments within the sentence or paragraph does not matter. These operators only work on indexes built with index_sp (sentence and paragraph indexing feature) enabled, and revert to a mere AND otherwise. Refer to the index_sp directive documentation for the notes on what's considered a sentence and a paragraph.

ZONE limit operator, added in 2.0.1-beta, is quite similar to field limit operator, but restricts matching to a given in-field zone or a list of zones. Note that the subsequent subexpressions are not required to match in a single contiguous span of a given zone, and may match in multiple spans. For instance, (ZONE:th hello world) query will match this example document:

<th>Table 1. Local awareness of Hello Kitty brand.</th>
.. some table data goes here ..
<th>Table 2. World-wide brand awareness.</th>

ZONE operator affects the query until the next field or ZONE limit operator, or the closing parenthesis. It only works on the indexes built with zones support (see Section 12.2.9, “index_zones”) and will be ignored otherwise.

ZONESPAN limit operator, added in 2.1.1-beta, is similar to the ZONE operator, but requires the match to occur in a single contiguous span. In the example above, (ZONESPAN:th hello world)> would not match the document, since "hello" and "world" do not occur within the same span.

MAYBE operator was added in 2.2.3-beta. It works much like | operator but doesn't return documents which match only right subtree expression.

5.4. Search results ranking

5.4.1. Ranking overview

Ranking (aka weighting) of the search results can be defined as a process of computing a so-called relevance (aka weight) for every given matched document with regards to a given query that matched it. So relevance is in the end just a number attached to every document that estimates how relevant the document is to the query. Search results can then be sorted based on this number and/or some additional parameters, so that the most sought after results would come up higher on the results page.

There is no single standard one-size-fits-all way to rank any document in any scenario. Moreover, there can not ever be such a way, because relevance is subjective. As in, what seems relevant to you might not seem relevant to me. Hence, in general case it's not just hard to compute, it's theoretically impossible.

So ranking in Sphinx is configurable. It has a notion of a so-called ranker. A ranker can formally be defined as a function that takes document and query as its input and produces a relevance value as output. In layman's terms, a ranker controls exactly how (using which specific algorithm) will Sphinx assign weights to the document.

Previously, this ranking function was rigidly bound to the matching mode. So in the legacy matching modes (that is, SPH_MATCH_ALL, SPH_MATCH_ANY, SPH_MATCH_PHRASE, and SPH_MATCH_BOOLEAN) you can not choose the ranker. You can only do that in the SPH_MATCH_EXTENDED mode. (Which is the only mode in SphinxQL and the suggested mode in SphinxAPI anyway.) To choose a non-default ranker you can either use SetRankingMode() with SphinxAPI, or OPTION ranker clause in SELECT statement when using SphinxQL.

As a sidenote, legacy matching modes are internally implemented via the unified syntax anyway. When you use one of those modes, Sphinx just internally adjusts the query and sets the associated ranker, then executes the query using the very same unified code path.

5.4.2. Available built-in rankers

Sphinx ships with a number of built-in rankers suited for different purposes. A number of them uses two factors, phrase proximity (aka LCS) and BM25. Phrase proximity works on the keyword positions, while BM25 works on the keyword frequencies. Basically, the better the degree of the phrase match between the document body and the query, the higher is the phrase proximity (it maxes out when the document contains the entire query as a verbatim quote). And BM25 is higher when the document contains more rare words. We'll save the detailed discussion for later.

Currently implemented rankers are:

  • SPH_RANK_PROXIMITY_BM25, the default ranking mode that uses and combines both phrase proximity and BM25 ranking.

  • SPH_RANK_BM25, statistical ranking mode which uses BM25 ranking only (similar to most other full-text engines). This mode is faster but may result in worse quality on queries which contain more than 1 keyword.

  • SPH_RANK_NONE, no ranking mode. This mode is obviously the fastest. A weight of 1 is assigned to all matches. This is sometimes called boolean searching that just matches the documents but does not rank them.

  • SPH_RANK_WORDCOUNT, ranking by the keyword occurrences count. This ranker computes the per-field keyword occurrence counts, then multiplies them by field weights, and sums the resulting values.

  • SPH_RANK_PROXIMITY, added in version 0.9.9-rc1, returns raw phrase proximity value as a result. This mode is internally used to emulate SPH_MATCH_ALL queries.

  • SPH_RANK_MATCHANY, added in version 0.9.9-rc1, returns rank as it was computed in SPH_MATCH_ANY mode earlier, and is internally used to emulate SPH_MATCH_ANY queries.

  • SPH_RANK_FIELDMASK, added in version 0.9.9-rc2, returns a 32-bit mask with N-th bit corresponding to N-th fulltext field, numbering from 0. The bit will only be set when the respective field has any keyword occurrences satisfying the query.

  • SPH_RANK_SPH04, added in version 1.10-beta, is generally based on the default SPH_RANK_PROXIMITY_BM25 ranker, but additionally boosts the matches when they occur in the very beginning or the very end of a text field. Thus, if a field equals the exact query, SPH04 should rank it higher than a field that contains the exact query but is not equal to it. (For instance, when the query is "Hyde Park", a document entitled "Hyde Park" should be ranked higher than a one entitled "Hyde Park, London" or "The Hyde Park Cafe".)

  • SPH_RANK_EXPR, added in version 2.0.2-beta, lets you specify the ranking formula in run time. It exposes a number of internal text factors and lets you define how the final weight should be computed from those factors. You can find more details about its syntax and a reference available factors in a subsection below.

You should specify the SPH_RANK_ prefix and use capital letters only when using the SetRankingMode() call from the SphinxAPI. The API ports expose these as global constants. Using SphinxQL syntax, the prefix should be omitted and the ranker name is case insensitive. Example:

// SphinxAPI
$client->SetRankingMode ( SPH_RANK_SPH04 );

// SphinxQL
mysql_query ( "SELECT ... OPTION ranker=sph04" );

Legacy matching modes rankers

Legacy matching modes automatically select a ranker as follows:





5.4.3. Expression based ranker (SPH_RANK_EXPR)

Expression ranker, added in version 2.0.2-beta, lets you change the ranking formula on the fly, on a per-query basis. For a quick kickoff, this is how you emulate PROXIMITY_BM25 ranker using the expression based one:

SELECT *, WEIGHT() FROM myindex WHERE MATCH('hello world')
OPTION ranker=expr('sum(lcs*user_weight)*1000+bm25')

The output of this query must not change if you omit the OPTION clause, because the default ranker (PROXIMITY_BM25) behaves exactly like specified in the ranker formula above. But the expression ranker is somewhat more flexible than just that and provides access to many more factors.

The ranking formula is an arbitrary arithmetic expression that can use constants, document attributes, built-in functions and operators (described in Section 5.5, “Expressions, functions, and operators”), and also a few ranking-specific things that are only accessible in a ranking formula. Namely, those are field aggregation functions, field-level, and document-level ranking factors.

5.4.4. Quick summary of the ranking factors

Table 5.1. 

max_lcsqueryintmaximum possible LCS value for the current query
bm25documentintquick estimate of BM25(1.2, 0) without syntax support
bm25a(k1, b)documentintprecise BM25() value with configurable K1, B constants and syntax support
bm25f(k1, b, {field=weight, ...})documentintprecise BM25F() value with extra configurable field weights
field_maskdocumentintbit mask of matched fields
query_word_countdocumentintnumber of unique inclusive keywords in a query
doc_word_countdocumentintnumber of unique keywords matched in the document
lcsfieldintLongest Common Subsequence between query and document, in words
user_weightfieldintuser field weight
hit_countfieldinttotal number of keyword occurrences
word_countfieldintnumber of unique matched keywords
tf_idffieldfloatsum(tf*idf) over matched keywords == sum(idf) over occurrences
min_hit_posfieldintfirst matched occurrence position, in words, 1-based
min_best_span_posfieldintfirst maximum LCS span position, in words, 1-based
exact_hitfieldboolwhether query == field
min_idffieldfloatmin(idf) over matched keywords
max_idffieldfloatmax(idf) over matched keywords
sum_idffieldfloatsum(idf) over matched keywords
exact_orderfieldboolwhether all query keywords were a) matched and b) in query order
min_gapsfieldintminimum number of gaps between the matched keywords over the matching spans
lccsfieldintLongest Common Contiguous Subsequence between query and document, in words
wlccsfieldfloatWeighted Longest Common Contiguous Subsequence, sum(idf) over contiguous keyword spans
atcfieldfloatAggregate Term Closeness, log(1+sum(idf1*idf2*pow(distance, -1.75)) over the best pairs of keywords

5.4.5. Document-level ranking factors

A document-level factor is a numeric value computed by the ranking engine for every matched document with regards to the current query. (So it differs from a plain document attribute in that the attribute do not depend on the full text query, while factors might.) Those factors can be used anywhere in the ranking expression. Currently implemented document-level factors are:

  • bm25 (integer), a document-level BM25 estimate (computed without keyword occurrence filtering).

  • max_lcs (integer), a query-level maximum possible value that the sum(lcs*user_weight) expression can ever take. This can be useful for weight boost scaling. For instance, MATCHANY ranker formula uses this to guarantee that a full phrase match in any field ranks higher than any combination of partial matches in all fields.

  • field_mask (integer), a document-level 32-bit mask of matched fields.

  • query_word_count (integer), the number of unique keywords in a query, adjusted for a number of excluded keywords. For instance, both (one one one one) and (one !two) queries should assign a value of 1 to this factor, because there is just one unique non-excluded keyword.

  • doc_word_count (integer), the number of unique keywords matched in the entire document.

5.4.6. Field-level ranking factors

A field-level factor is a numeric value computed by the ranking engine for every matched in-document text field with regards to the current query. As more than one field can be matched by a query, but the final weight needs to be a single integer value, these values need to be folded into a single one. To achieve that, field-level factors can only be used within a field aggregation function, they can not be used anywhere in the expression. For example, you can not use (lcs+bm25) as your ranking expression, as lcs takes multiple values (one in every matched field). You should use (sum(lcs)+bm25) instead, that expression sums lcs over all matching fields, and then adds bm25 to that per-field sum. Currently implemented field-level factors are:

  • lcs (integer), the length of a maximum verbatim match between the document and the query, counted in words. LCS stands for Longest Common Subsequence (or Subset). Takes a minimum value of 1 when only stray keywords were matched in a field, and a maximum value of query keywords count when the entire query was matched in a field verbatim (in the exact query keywords order). For example, if the query is 'hello world' and the field contains these two words quoted from the query (that is, adjacent to each other, and exactly in the query order), lcs will be 2. For example, if the query is 'hello world program' and the field contains 'hello world', lcs will be 2. Note that any subset of the query keyword works, not just a subset of adjacent keywords. For example, if the query is 'hello world program' and the field contains 'hello (test program)', lcs will be 2 just as well, because both 'hello' and 'program' matched in the same respective positions as they were in the query. Finally, if the query is 'hello world program' and the field contains 'hello world program', lcs will be 3. (Hopefully that is unsurprising at this point.)

  • user_weight (integer), the user specified per-field weight (refer to SetFieldWeights() in SphinxAPI and OPTION field_weights in SphinxQL respectively). The weights default to 1 if not specified explicitly.

  • hit_count (integer), the number of keyword occurrences that matched in the field. Note that a single keyword may occur multiple times. For example, if 'hello' occurs 3 times in a field and 'world' occurs 5 times, hit_count will be 8.

  • word_count (integer), the number of unique keywords matched in the field. For example, if 'hello' and 'world' occur anywhere in a field, word_count will be 2, irregardless of how many times do both keywords occur.

  • tf_idf (float), the sum of TF*IDF over all the keywords matched in the field. IDF is the Inverse Document Frequency, a floating point value between 0 and 1 that describes how frequent is the keywords (basically, 0 for a keyword that occurs in every document indexed, and 1 for a unique keyword that occurs in just a single document). TF is the Term Frequency, the number of matched keyword occurrences in the field. As a side note, tf_idf is actually computed by summing IDF over all matched occurrences. That's by construction equivalent to summing TF*IDF over all matched keywords.

  • min_hit_pos (integer), the position of the first matched keyword occurrence, counted in words. Indexing begins from position 1.

  • min_best_span_pos (integer), the position of the first maximum LCS occurrences span. For example, assume that our query was 'hello world program' and 'hello world' subphrase was matched twice in the field, in positions 13 and 21. Assume that 'hello' and 'world' additionally occurred elsewhere in the field, but never next to each other and thus never as a subphrase match. In that case, min_best_span_pos will be 13. Note how for the single keyword queries min_best_span_pos will always equal min_hit_pos.

  • exact_hit (boolean), whether a query was an exact match of the entire current field. Used in the SPH04 ranker.

  • min_idf, max_idf, and sum_idf (float), added in version 2.1.1-beta. These factors respectively represent the min(idf), max(idf) and sum(idf) over all keywords that were matched in the field.

  • exact_order (boolean), added in version 2.2.1-beta. Whether all of the query keywords were matched in the field in the exact query order. For example, (microsoft office) query would yield exact_order=1 in a field with the following contents: (We use Microsoft software in our office.). However, the very same query in a (Our office is Microsoft free.) field would yield exact_order=0.

  • min_gaps (integer), added in version 2.2.1-beta, the minimum number of positional gaps between (just) the keywords matched in field. Always 0 when less than 2 keywords match; always greater or equal than 0 otherwise.

    For example, with a [big wolf] query, [big bad wolf] field would yield min_gaps=1; [big bad hairy wolf] field would yield min_gaps=2; [the wolf was scary and big] field would yield min_gaps=3; etc. However, a field like [i heard a wolf howl] would yield min_gaps=0, because only one keyword would be matching in that field, and, naturally, there would be no gaps between the matchedkeywords.

    Therefore, this is a rather low-level, "raw" factor that you would most likely want to adjust before actually using for ranking. Specific adjustments depend heavily on your data and the resulting formula, but here are a few ideas you can start with: (a) any min_gaps based boosts could be simply ignored when word_count<2; (b) non-trivial min_gaps values (i.e. when word_count>=2) could be clamped with a certain "worst case" constant while trivial values (i.e. when min_gaps=0 and word_count<2) could be replaced by that constant; (c) a transfer function like 1/(1+min_gaps) could be applied (so that better, smaller min_gaps values would maximize it and worse, bigger min_gaps values would fall off slowly); and so on.

  • lccs (integer), added in version 2.2.1-beta. Longest Common Contiguous Subsequence. A length of the longest subphrase that is common between the query and the document, computed in keywords.

    LCCS factor is rather similar to LCS but more restrictive, in a sense. While LCS could be greater than 1 though no two query words are matched next to each other, LCCS would only get greater than 1 if there are exact, contiguous query subphrases in the document. For example, (one two three four five) query vs (one hundred three hundred five hundred) document would yield lcs=3, but lccs=1, because even though mutual dispositions of 3 keywords (one, three, five) match between the query and the document, no 2 matching positions are actually next to each other.

    Note that LCCS still does not differentiate between the frequent and rare keywords; for that, see WLCS and WLLCS.

  • wlccs (float), added in version 2.2.1-beta. Weighted Longest Common Contiguous Subsequence. A sum of IDFs of the keywords of the longest subphrase that is common between the query and the document.

    WLCCS is computed very similarly to LCCS, but every "suitable" keyword occurrence increases it by the keyword IDF rather than just by 1 (which is the case with LCS and LCCS). That lets us rank sequences of more rare and important keywords higher than sequences of frequent keywords, even if the latter are longer. For example, a query (Zanzibar bed and breakfast) would yield lccs=1 for a (hotels of Zanzibar) document, but lccs=3 against (London bed and breakfast), even though "Zanzibar" is actually somewhat more rare than the entire "bed and breakfast" phrase. WLCCS factor alleviates that problem by using the keyword frequencies.

  • atc (float), added in version 2.2.1-beta. Aggregate Term Closeness. A proximity based measure that grows higher when the document contains more groups of more closely located and more important (rare) query keywords. WARNING: you should use ATC with OPTION idf='plain,tfidf_unnormalized'; otherwise you would get unexpected results.

    ATC basically works as follows. For every keyword occurrence in the document, we compute the so called term closeness. For that, we examine all the other closest occurrences of all the query keywords (keyword itself included too) to the left and to the right of the subject occurrence, compute a distance dampening coefficient as k = pow(distance, -1.75) for those occurrences, and sum the dampened IDFs. Thus for every occurrence of every keyword, we get a "closeness" value that describes the "neighbors" of that occurrence. We then multiply those per-occurrence closenesses by their respective subject keyword IDF, sum them all, and finally, compute a logarithm of that sum.

    Or in other words, we process the best (closest) matched keyword pairs in the document, and compute pairwise "closenesses" as the product of their IDFs scaled by the distance coefficient:

    pair_tc = idf(pair_word1) * idf(pair_word2) * pow(pair_distance, -1.75)

    We then sum such closenesses, and compute the final, log-dampened ATC value:

    atc = log(1+sum(pair_tc))

    Note that this final dampening logarithm is exactly the reason you should use OPTION idf=plain, because without it, the expression inside the log() could be negative.

    Having closer keyword occurrences actually contributes much more to ATC than having more frequent keywords. Indeed, when the keywords are right next to each other, distance=1 and k=1; when there just one word in between them, distance=2 and k=0.297, with two words between, distance=3 and k=0.146, and so on. At the same time IDF attenuates somewhat slower. For example, in a 1 million document collection, the IDF values for keywords that match in 10, 100, and 1000 documents would be respectively 0.833, 0.667, and 0.500. So a keyword pair with two rather rare keywords that occur in just 10 documents each but with 2 other words in between would yield pair_tc = 0.101 and thus just barely outweigh a pair with a 100-doc and a 1000-doc keyword with 1 other word between them and pair_tc = 0.099. Moreover, a pair of two unique, 1-doc keywords with 3 words between them would get a pair_tc = 0.088 and lose to a pair of two 1000-doc keywords located right next to each other and yielding a pair_tc = 0.25. So, basically, while ATC does combine both keyword frequency and proximity, it is still somewhat favoring the proximity.

5.4.7. Ranking factor aggregation functions

A field aggregation function is a single argument function that takes an expression with field-level factors, iterates it over all the matched fields, and computes the final results. Currently implemented field aggregation functions are:

  • sum, sums the argument expression over all matched fields. For instance, sum(1) should return a number of matched fields.

  • top, returns the greatest value of the argument over all matched fields.

5.4.8. Formula expressions for all the built-in rankers

Most of the other rankers can actually be emulated with the expression based ranker. You just need to pass a proper expression. Such emulation is, of course, going to be slower than using the built-in, compiled ranker but still might be of interest if you want to fine-tune your ranking formula starting with one of the existing ones. Also, the formulas define the nitty gritty ranker details in a nicely readable fashion.

  • SPH_RANK_PROXIMITY_BM25 = sum(lcs*user_weight)*1000+bm25

  • SPH_RANK_BM25 = bm25


  • SPH_RANK_WORDCOUNT = sum(hit_count*user_weight)

  • SPH_RANK_PROXIMITY = sum(lcs*user_weight)

  • SPH_RANK_MATCHANY = sum((word_count+(lcs-1)*max_lcs)*user_weight)

  • SPH_RANK_FIELDMASK = field_mask

  • SPH_RANK_SPH04 = sum((4*lcs+2*(min_hit_pos==1)+exact_hit)*user_weight)*1000+bm25

5.5. Expressions, functions, and operators

Sphinx lets you use arbitrary arithmetic expressions both via SphinxQL and SphinxAPI, involving attribute values, internal attributes (document ID and relevance weight), arithmetic operations, a number of built-in functions, and user-defined functions. This section documents the supported operators and functions. Here's the complete reference list for quick access.

5.5.1. Operators

Arithmetic operators: +, -, *, /, %, DIV, MOD

The standard arithmetic operators. Arithmetic calculations involving those can be performed in three different modes: (a) using single-precision, 32-bit IEEE 754 floating point values (the default), (b) using signed 32-bit integers, (c) using 64-bit signed integers. The expression parser will automatically switch to integer mode if there are no operations the result in a floating point value. Otherwise, it will use the default floating point mode. For instance, a+b will be computed using 32-bit integers if both arguments are 32-bit integers; or using 64-bit integers if both arguments are integers but one of them is 64-bit; or in floats otherwise. However, a/b or sqrt(a) will always be computed in floats, because these operations return a result of non-integer type. To avoid the first, you can either use IDIV(a,b) or a DIV b form. Also, a*b will not be automatically promoted to 64-bit when the arguments are 32-bit. To enforce 64-bit results, you can use BIGINT(). (But note that if there are non-integer operations, BIGINT() will simply be ignored.)

Comparison operators: <, > <=, >=, =, <>

Comparison operators (eg. = or <=) return 1.0 when the condition is true and 0.0 otherwise. For instance, (a=b)+3 will evaluate to 4 when attribute 'a' is equal to attribute 'b', and to 3 when 'a' is not. Unlike MySQL, the equality comparisons (ie. = and <> operators) introduce a small equality threshold (1e-6 by default). If the difference between compared values is within the threshold, they will be considered equal.

Boolean operators: AND, OR, NOT

Boolean operators (AND, OR, NOT) were introduced in 0.9.9-rc2 and behave as usual. They are left-associative and have the least priority compared to other operators. NOT has more priority than AND and OR but nevertheless less than any other operator. AND and OR have the same priority so brackets use is recommended to avoid confusion in complex expressions.

Bitwise operators: &, |

These operators perform bitwise AND and OR respectively. The operands must be of an integer types. Introduced in version 1.10-beta.

5.5.2. Numeric functions


Returns the absolute value of the argument.


BITDOT(mask, w0, w1, ...) returns the sum of products of an each bit of a mask multiplied with its weight. bit0*w0 + bit1*w1 + ...


Returns the smallest integer value greater or equal to the argument.


CONTAINS(polygon, x, y) checks whether the (x,y) point is within the given polygon, and returns 1 if true, or 0 if false. The polygon has to be specified using either the POLY2D() function or the GEOPOLY2D() function. The former function is intended for "small" polygons, meaning less than 500 km (300 miles) a side, and it doesn't take into account the Earth's curvature for speed. For larger distances, you should use GEOPOLY2D, which tessellates the given polygon in smaller parts, accounting for the Earth's curvature. These functions were added in version 2.1.1-beta.


Returns the cosine of the argument.


Forcibly promotes given argument to floating point type. Intended to help enforce evaluation of numeric JSON fields. Introduced in version 2.2.1-beta.


Returns the exponent of the argument (e=2.718... to the power of the argument).


Returns the N-th Fibonacci number, where N is the integer argument. That is, arguments of 0 and up will generate the values 0, 1, 1, 2, 3, 5, 8, 13 and so on. Note that the computations are done using 32-bit integer math and thus numbers 48th and up will be returned modulo 2^32.


Returns the largest integer value lesser or equal to the argument.


GEOPOLY2D(x1,y1,x2,y2,x3,y3...) produces a polygon to be used with the CONTAINS() function. This function takes into account the Earth's curvature by tessellating the polygon into smaller ones, and should be used for larger areas; see the POLY2D() function.


Returns the result of an integer division of the first argument by the second argument. Both arguments must be of an integer type.


Returns the natural logarithm of the argument (with the base of e=2.718...).


Returns the common logarithm of the argument (with the base of 10).


Returns the binary logarithm of the argument (with the base of 2).


Returns the bigger of two arguments.


Returns the smaller of two arguments.


POLY2D(x1,y1,x2,y2,x3,y3...) produces a polygon to be used with the CONTAINS() function. This polygon assumes a flat Earth, so it should not be too large; see the POLY2D() function.


Returns the first argument raised to the power of the second argument.


Returns the sine of the argument.


Returns the square root of the argument.


Forcibly reinterprets given argument to 64-bit unsigned type. Introduced in version 2.2.1-beta.

5.5.3. Date and time functions


Returns the integer day of month (in 1..31 range) from a timestamp argument, according to the current timezone. Introduced in version 2.0.1-beta.


Returns the integer month (in 1..12 range) from a timestamp argument, according to the current timezone. Introduced in version 2.0.1-beta.


Returns the current timestamp as an INTEGER. Introduced in version 0.9.9-rc1.


Returns the integer year (in 1969..2038 range) from a timestamp argument, according to the current timezone. Introduced in version 2.0.1-beta.


Returns the integer year and month code (in 196912..203801 range) from a timestamp argument, according to the current timezone. Introduced in version 2.0.1-beta.


Returns the integer year, month, and date code (in 19691231..20380119 range) from a timestamp argument, according to the current timezone. Introduced in version 2.0.1-beta.

5.5.4. Type conversion functions


Forcibly promotes the integer argument to 64-bit type, and does nothing on floating point argument. It's intended to help enforce evaluation of certain expressions (such as a*b) in 64-bit mode even though all the arguments are 32-bit. Introduced in version 0.9.9-rc1.


Forcibly promotes given argument to 64-bit signed type. Intended to help enforce evaluation of numeric JSON fields. Introduced in version 2.2.1-beta.


Forcibly reinterprets its 32-bit unsigned integer argument as signed, and also expands it to 64-bit type (because 32-bit type is unsigned). It's easily illustrated by the following example: 1-2 normally evaluates to 4294967295, but SINT(1-2) evaluates to -1. Introduced in version 1.10-beta.

5.5.5. Comparison functions


IF() behavior is slightly different that that of its MySQL counterpart. It takes 3 arguments, check whether the 1st argument is equal to 0.0, returns the 2nd argument if it is not zero, or the 3rd one when it is. Note that unlike comparison operators, IF() does not use a threshold! Therefore, it's safe to use comparison results as its 1st argument, but arithmetic operators might produce unexpected results. For instance, the following two calls will produce different results even though they are logically equivalent:

IF ( sqrt(3)*sqrt(3)-3<>0, a, b )
IF ( sqrt(3)*sqrt(3)-3, a, b )

In the first case, the comparison operator <> will return 0.0 (false) because of a threshold, and IF() will always return 'b' as a result. In the second one, the same sqrt(3)*sqrt(3)-3 expression will be compared with zero without threshold by the IF() function itself. But its value will be slightly different from zero because of limited floating point calculations precision. Because of that, the comparison with 0.0 done by IF() will not pass, and the second variant will return 'a' as a result.


IN(expr,val1,val2,...), introduced in version 0.9.9-rc1, takes 2 or more arguments, and returns 1 if 1st argument (expr) is equal to any of the other arguments (val1..valN), or 0 otherwise. Currently, all the checked values (but not the expression itself!) are required to be constant. (Its technically possible to implement arbitrary expressions too, and that might be implemented in the future.) Constants are pre-sorted and then binary search is used, so IN() even against a big arbitrary list of constants will be very quick. Starting with 0.9.9-rc2, first argument can also be a MVA attribute. In that case, IN() will return 1 if any of the MVA values is equal to any of the other arguments. Starting with 2.0.1-beta, IN() also supports IN(expr,@uservar) syntax to check whether the value belongs to the list in the given global user variable. First argument can be JSON attribute since 2.2.1-beta.


INTERVAL(expr,point1,point2,point3,...), introduced in version 0.9.9-rc1, takes 2 or more arguments, and returns the index of the argument that is less than the first argument: it returns 0 if expr<point1, 1 if point1<=expr<point2, and so on. It is required that point1<point2<...<pointN for this function to work correctly.

5.5.6. Miscellaneous functions


ALL(cond FOR var IN json.array) function was introduced in 2.2.1-beta. It applies to JSON arrays and returns 1 if condition is true for all elements in array and 0 otherwise. 'cond' is a general expression which additionally can use 'var' as current value of an array element within itself.

SELECT ALL(x>3 AND x<7 FOR x IN j.intarray) FROM test;

ANY(cond FOR var IN json.array) function was introduced in 2.2.1-beta. It works similar to ALL() except for it returns 1 if condition is true for any element in array.


Returns the arctangent function of two arguments, expressed in radians.


Returns the CRC32 value of a string argument. Introduced in version 2.0.1-beta.


GEODIST(lat1, lon1, lat2, lon2, [...]) function, introduced in version 0.9.9-rc2, computes geosphere distance between two given points specified by their coordinates. Note that by default both latitudes and longitudes must be in radians and the result will be in meters. You can use arbitrary expression as any of the four coordinates. An optimized path will be selected when one pair of the arguments refers directly to a pair attributes and the other one is constant.

Starting with version 2.2.1-beta, GEODIST() also takes an optional 5th argument that lets you easily convert between input and output units, and pick the specific geodistance formula to use. The complete syntax and a few examples are as follows:

GEODIST(lat1, lon1, lat2, lon2, { option=value, ... })

GEODIST(40.7643929, -73.9997683, 40.7642578, -73.9994565, {in=degrees, out=feet})
GEODIST(51.50, -0.12, 29.98, 31.13, {in=deg, out=mi}}

The known options and their values are:

  • in = {deg | degrees | rad | radians}, specifies the input units;
  • out = {m | meters | km | kilometers | ft | feet | mi | miles}, specifies the output units;
  • method = {haversine | adaptive}, specifies the geodistance calculation method.

Upto version 2.1.x (inclusive), "haversine" method was the default. Starting with 2.2.1-beta, the default method changed to "adaptive", a new, well optimized implementation that is both more precise and much faster at all times.


GREATEST(attr_json.some_array) was introduced in version 2.2.1-beta. First argument is JSON array and return value is the greatest value in that array. Also works for MVA.


INDEXOF(cond FOR var IN json.array) function was introduced in 2.2.1-beta. It iterates through all elements in array and returns index of first element for which 'cond' is true and -1 if 'cond' is false for every element in array.

SELECT INDEXOF(name='John' FOR name IN j.peoples) FROM test;

LEAST(attr_json.some_array) was introduced in version 2.2.1-beta. First argument is JSON array and return value is the least value in that array. Also works for MVA.


LENGTH(attr_mva) function, introduced in version 2.1.2-stable, returns amount of elements in MVA set. It works with both 32-bit and 64-bit MVA attributes. LENGTH(attr_json) was introduced in version 2.2.1-beta. It returns length of a field in JSON. Return value depends on type of a field. For example LENGTH(json_attr.some_int) always returns 1 and LENGTH(json_attr.some_array) returns number of elements in array.


Returns sort key value of the worst found element in the current top-N matches if sort key is float and 0 otherwise.


Returns weight of the worst found element in the current top-N matches.


PACKEDFACTORS(), introduced in version 2.1.1-beta, can be used in queries, either to just see all the weighting factors calculated when doing the matching, or to provide a binary attribute that can be used to write a custom ranking UDF. This function works only if expression ranker is specified and the query is not a full scan, otherwise it will return an error. Starting with 2.2.2-beta PACKEDFACTORS() can take an optional argument that disables ATC ranking factor calculation:


Calculating ATC slows down query processing considerably, so this option can be useful if you need to see the ranking factors, but do not need ATC. Starting with 2.2.3-beta PACKEDFACTORS() can also be told to format its output as JSON:


The respective outputs in either key-value pair or JSON format would look as follows below. (Note that the examples below are wrapped for readability; actual returned values would be single-line.)

    -> WHERE MATCH('test one') OPTION ranker=expr('1') \G
*************************** 1. row ***************************
             id: 1
packedfactors(): bm25=569, bm25a=0.617197, field_mask=2, doc_word_count=2,
    field1=(lcs=1, hit_count=2, word_count=2, tf_idf=0.152356,
        min_idf=-0.062982, max_idf=0.215338, sum_idf=0.152356, min_hit_pos=4,
        min_best_span_pos=4, exact_hit=0, max_window_hits=1, min_gaps=2,
        exact_order=1, lccs=1, wlccs=0.215338, atc=-0.003974),
    word0=(tf=1, idf=-0.062982),
    word1=(tf=1, idf=0.215338)
1 row in set (0.00 sec)

mysql> SELECT id, PACKEDFACTORS({json=1}) FROM test1
    -> WHERE MATCH('test one') OPTION ranker=expr('1') \G
*************************** 1. row ***************************
                     id: 1

    "bm25": 569,
    "bm25a": 0.617197,
    "field_mask": 2,
    "doc_word_count": 2,
    "fields": [
            "lcs": 1,
            "hit_count": 2,
            "word_count": 2,
            "tf_idf": 0.152356,
            "min_idf": -0.062982,
            "max_idf": 0.215338,
            "sum_idf": 0.152356,
            "min_hit_pos": 4,
            "min_best_span_pos": 4,
            "exact_hit": 0,
            "max_window_hits": 1,
            "min_gaps": 2,
            "exact_order": 1,
            "lccs": 1,
            "wlccs": 0.215338,
            "atc": -0.003974
    "words": [
            "tf": 1,
            "idf": -0.062982
            "tf": 1,
            "idf": 0.215338

1 row in set (0.01 sec)

This function can be used to implement custom ranking functions in UDFs, as in

FROM my_index
WHERE match('hello')
OPTION ranker=expr('1');

Where CUSTOM_RANK() is a function implemented in an UDF. It should declare a SPH_UDF_FACTORS structure (defined in sphinxudf.h), initialize this structure, unpack the factors into it before usage, and deinitialize it afterwards, as follows:

sphinx_factors_unpack((DWORD*)args->arg_values[0], &factors);
// ... can use the contents of factors variable here ...

PACKEDFACTORS() data is available at all query stages, not just when doing the initial matching and ranking pass. That enables another particularly interesting application of PACKEDFACTORS(), namely re-ranking.

In the example just above, we used an expression-based ranker with a dummy expression, and sorted the result set by the value computed by our UDF. In other words, we used the UDF to rank all our results. Assume now, for the sake of an example, that our UDF is extremely expensive to compute and has a throughput of just 10,000 calls per second. Assume that our query matches 1,000,000 documents. To maintain reasonable performance, we would then want to use a (much) simpler expression to do most of our ranking, and then apply the expensive UDF to only a few top results, say, top-100 results. Or, in other words, build top-100 results using a simpler ranking function and then re-rank those with a complex one. We can do that just as well with subselects:

    FROM my_index WHERE match('hello')
    OPTION ranker=expr('sum(lcs)*1000+bm25')
    LIMIT 100

In this example, expression-based ranker will be called for every matched document to compute WEIGHT(). So it will get called 1,000,000 times. But the UDF computation can be postponed until the outer sort. And it also will be done for just the top-100 matches by WEIGHT(), according to the inner limit. So the UDF will only get called 100 times. And then the final top-10 matches by UDF value will be selected and returned to the application.

For reference, in the distributed case PACKEDFACTORS() data gets sent from the agents to master in a binary format, too. This makes it technically feasible to implement additional re-ranking pass (or passes) on the master node, if needed.

If used with SphinxQL but not called from any UDFs, the result of PACKEDFACTORS() is simply formatted as plain text, which can be used to manually assess the ranking factors. Note that this feature is not currently supported by the Sphinx API.


REMAP(condition, expression, (cond1, cond2, ...), (expr1, expr2, ...)) function was added in 2.2.2-beta. It allows you to make some exceptions of an expression values depending on condition values. Condition expression should always result integer, expression can result in integer or float.

SELECT REMAP(userid, karmapoints, (1, 67), (999, 0)) FROM users;
SELECT REMAP(id%10, salary, (0), (0.0)) FROM employes;

5.6. Sorting modes

There are the following result sorting modes available:

  • SPH_SORT_RELEVANCE mode, that sorts by relevance in descending order (best matches first);

  • SPH_SORT_ATTR_DESC mode, that sorts by an attribute in descending order (bigger attribute values first);

  • SPH_SORT_ATTR_ASC mode, that sorts by an attribute in ascending order (smaller attribute values first);

  • SPH_SORT_TIME_SEGMENTS mode, that sorts by time segments (last hour/day/week/month) in descending order, and then by relevance in descending order;

  • SPH_SORT_EXTENDED mode, that sorts by SQL-like combination of columns in ASC/DESC order;

  • SPH_SORT_EXPR mode, that sorts by an arithmetic expression.

SPH_SORT_RELEVANCE ignores any additional parameters and always sorts matches by relevance rank. All other modes require an additional sorting clause, with the syntax depending on specific mode. SPH_SORT_ATTR_ASC, SPH_SORT_ATTR_DESC and SPH_SORT_TIME_SEGMENTS modes require simply an attribute name. SPH_SORT_RELEVANCE is equivalent to sorting by "@weight DESC, @id ASC" in extended sorting mode, SPH_SORT_ATTR_ASC is equivalent to "attribute ASC, @weight DESC, @id ASC", and SPH_SORT_ATTR_DESC to "attribute DESC, @weight DESC, @id ASC" respectively.


In SPH_SORT_TIME_SEGMENTS mode, attribute values are split into so-called time segments, and then sorted by time segment first, and by relevance second.

The segments are calculated according to the current timestamp at the time when the search is performed, so the results would change over time. The segments are as follows:

  • last hour,

  • last day,

  • last week,

  • last month,

  • last 3 months,

  • everything else.

These segments are hardcoded, but it is trivial to change them if necessary.

This mode was added to support searching through blogs, news headlines, etc. When using time segments, recent records would be ranked higher because of segment, but within the same segment, more relevant records would be ranked higher - unlike sorting by just the timestamp attribute, which would not take relevance into account at all.


In SPH_SORT_EXTENDED mode, you can specify an SQL-like sort expression with up to 5 attributes (including internal attributes), eg:

@relevance DESC, price ASC, @id DESC

Both internal attributes (that are computed by the engine on the fly) and user attributes that were configured for this index are allowed. Internal attribute names must start with magic @-symbol; user attribute names can be used as is. In the example above, @relevance and @id are internal attributes and price is user-specified.

Known internal attributes are:

  • @id (match ID)

  • @weight (match weight)

  • @rank (match weight)

  • @relevance (match weight)

  • @random (return results in random order)

@rank and @relevance are just additional aliases to @weight.


Expression sorting mode lets you sort the matches by an arbitrary arithmetic expression, involving attribute values, internal attributes (@id and @weight), arithmetic operations, and a number of built-in functions. Here's an example:

$cl->SetSortMode ( SPH_SORT_EXPR,
    "@weight + ( user_karma + ln(pageviews) )*0.1" );

The operators and functions supported in the expressions are discussed in a separate section, Section 5.5, “Expressions, functions, and operators”.

5.7. Grouping (clustering) search results

Sometimes it could be useful to group (or in other terms, cluster) search results and/or count per-group match counts - for instance, to draw a nice graph of how much matching blog posts were there per each month; or to group Web search results by site; or to group matching forum posts by author; etc.

In theory, this could be performed by doing only the full-text search in Sphinx and then using found IDs to group on SQL server side. However, in practice doing this with a big result set (10K-10M matches) would typically kill performance.

To avoid that, Sphinx offers so-called grouping mode. It is enabled with SetGroupBy() API call. When grouping, all matches are assigned to different groups based on group-by value. This value is computed from specified attribute using one of the following built-in functions:

  • SPH_GROUPBY_DAY, extracts year, month and day in YYYYMMDD format from timestamp;

  • SPH_GROUPBY_WEEK, extracts year and first day of the week number (counting from year start) in YYYYNNN format from timestamp;

  • SPH_GROUPBY_MONTH, extracts month in YYYYMM format from timestamp;

  • SPH_GROUPBY_YEAR, extracts year in YYYY format from timestamp;

  • SPH_GROUPBY_ATTR, uses attribute value itself for grouping.

The final search result set then contains one best match per group. Grouping function value and per-group match count are returned along as "virtual" attributes named @group and @count respectively.

The result set is sorted by group-by sorting clause, with the syntax similar to SPH_SORT_EXTENDED sorting clause syntax. In addition to @id and @weight, group-by sorting clause may also include:

  • @group (groupby function value),

  • @count (amount of matches in group).

The default mode is to sort by groupby value in descending order, ie. by "@group desc".

On completion, total_found result parameter would contain total amount of matching groups over he whole index.

WARNING: grouping is done in fixed memory and thus its results are only approximate; so there might be more groups reported in total_found than actually present. @count might also be underestimated. To reduce inaccuracy, one should raise max_matches. If max_matches allows to store all found groups, results will be 100% correct.

For example, if sorting by relevance and grouping by "published" attribute with SPH_GROUPBY_DAY function, then the result set will contain

  • one most relevant match per each day when there were any matches published,

  • with day number and per-day match count attached,

  • sorted by day number in descending order (ie. recent days first).

Starting with version 0.9.9-rc2, aggregate functions (AVG(), MIN(), MAX(), SUM()) are supported through SetSelect() API call when using GROUP BY.

5.8. Distributed searching

To scale well, Sphinx has distributed searching capabilities. Distributed searching is useful to improve query latency (ie. search time) and throughput (ie. max queries/sec) in multi-server, multi-CPU or multi-core environments. This is essential for applications which need to search through huge amounts data (ie. billions of records and terabytes of text).

The key idea is to horizontally partition (HP) searched data across search nodes and then process it in parallel.

Partitioning is done manually. You should

  • setup several instances of Sphinx programs (indexer and searchd) on different servers;

  • make the instances index (and search) different parts of data;

  • configure a special distributed index on some of the searchd instances;

  • and query this index.

This index only contains references to other local and remote indexes - so it could not be directly reindexed, and you should reindex those indexes which it references instead.

When searchd receives a query against distributed index, it does the following:

  1. connects to configured remote agents;

  2. issues the query;

  3. sequentially searches configured local indexes (while the remote agents are searching);

  4. retrieves remote agents' search results;

  5. merges all the results together, removing the duplicates;

  6. sends the merged results to client.

From the application's point of view, there are no differences between searching through a regular index, or a distributed index at all. That is, distributed indexes are fully transparent to the application, and actually there's no way to tell whether the index you queried was distributed or local. (Even though as of 0.9.9 Sphinx does not allow to combine searching through distributed indexes with anything else, this constraint will be lifted in the future.)

Any searchd instance could serve both as a master (which aggregates the results) and a slave (which only does local searching) at the same time. This has a number of uses:

  1. every machine in a cluster could serve as a master which searches the whole cluster, and search requests could be balanced between masters to achieve a kind of HA (high availability) in case any of the nodes fails;

  2. if running within a single multi-CPU or multi-core machine, there would be only 1 searchd instance querying itself as an agent and thus utilizing all CPUs/core.

It is scheduled to implement better HA support which would allow to specify which agents mirror each other, do health checks, keep track of alive agents, load-balance requests, etc.

5.9. searchd query log formats

In version 2.0.1-beta and above two query log formats are supported. Previous versions only supported a custom plain text format. That format is still the default one. However, while it might be more convenient for manual monitoring and review, but hard to replay for benchmarks, it only logs search queries but not the other types of requests, does not always contain the complete search query data, etc. The default text format is also harder (and sometimes impossible) to replay for benchmarking purposes. The new sphinxql format alleviates that. It aims to be complete and automatable, even though at the cost of brevity and readability.

5.9.1. Plain log format

By default, searchd logs all successfully executed search queries into a query log file. Here's an example:

[Fri Jun 29 21:17:58 2007] 0.004 sec 0.004 sec [all/0/rel 35254 (0,20)] [lj] test
[Fri Jun 29 21:20:34 2007] 0.024 sec 0.024 sec [all/0/rel 19886 (0,20) @channel_id] [lj] test

This log format is as follows:

[query-date] real-time wall-time [match-mode/filters-count/sort-mode
    total-matches (offset,limit) @groupby-attr] [index-name] query

  • real-time is a time measured just from start to finish of the query

  • wall-time like real-time but not including waiting for agents and merging result sets time

Match mode can take one of the following values:

  • "all" for SPH_MATCH_ALL mode;

  • "any" for SPH_MATCH_ANY mode;

  • "phr" for SPH_MATCH_PHRASE mode;

  • "bool" for SPH_MATCH_BOOLEAN mode;

  • "ext" for SPH_MATCH_EXTENDED mode;

  • "ext2" for SPH_MATCH_EXTENDED2 mode;

  • "scan" if the full scan mode was used, either by being specified with SPH_MATCH_FULLSCAN, or if the query was empty (as documented under Matching Modes)

Sort mode can take one of the following values:

  • "rel" for SPH_SORT_RELEVANCE mode;

  • "attr-" for SPH_SORT_ATTR_DESC mode;

  • "attr+" for SPH_SORT_ATTR_ASC mode;

  • "tsegs" for SPH_SORT_TIME_SEGMENTS mode;

  • "ext" for SPH_SORT_EXTENDED mode.

Additionally, if searchd was started with --iostats, there will be a block of data after where the index(es) searched are listed.

A query log entry might take the form of:

[Fri Jun 29 21:17:58 2007] 0.004 sec [all/0/rel 35254 (0,20)] [lj]
   [ios=6 kb=111.1 ms=0.5] test

This additional block is information regarding I/O operations in performing the search: the number of file I/O operations carried out, the amount of data in kilobytes read from the index files and time spent on I/O operations (although there is a background processing component, the bulk of this time is the I/O operation time).

5.9.2. SphinxQL log format

This is a new log format introduced in 2.0.1-beta, with the goals begin logging everything and then some, and in a format easy to automate (for instance, automatically replay). New format can either be enabled via the query_log_format directive in the configuration file, or switched back and forth on the fly with the SET GLOBAL query_log_format=... statement via SphinxQL. In the new format, the example from the previous section would look as follows. (Wrapped below for readability, but with just one query per line in the actual log.)

/* Fri Jun 29 21:17:58.609 2007 2011 conn 2 real 0.004 wall 0.004 found 35254 */
SELECT * FROM lj WHERE MATCH('test') OPTION ranker=proximity;

/* Fri Jun 29 21:20:34 2007.555 conn 3 real 0.024 wall 0.024 found 19886 */
SELECT * FROM lj WHERE MATCH('test') GROUP BY channel_id
OPTION ranker=proximity;

Note that all requests would be logged in this format, including those sent via SphinxAPI and SphinxSE, not just those sent via SphinxQL. Also note, that this kind of logging works only with plain log files and will not work if you use 'syslog' for logging.

The features of SphinxQL log format compared to the default text one are as follows.

  • All request types should be logged. (This is still work in progress.)

  • Full statement data will be logged where possible.

  • Errors and warnings are logged.

  • The log should be automatically replayable via SphinxQL.

  • Additional performance counters (currently, per-agent distributed query times) are logged.

Use sphinxql:compact_in to shorten your IN() clauses in log if you have too much values in it.

Every request (including both SphinxAPI and SphinxQL) request must result in exactly one log line. All request types, including INSERT, CALL SNIPPETS, etc will eventually get logged, though as of time of this writing, that is a work in progress). Every log line must be a valid SphinxQL statement that reconstructs the full request, except if the logged request is too big and needs shortening for performance reasons. Additional messages, counters, etc can be logged in the comments section after the request.

5.10. MySQL protocol support and SphinxQL

Starting with version 0.9.9-rc2, Sphinx searchd daemon supports MySQL binary network protocol and can be accessed with regular MySQL API. For instance, 'mysql' CLI client program works well. Here's an example of querying Sphinx using MySQL client:

$ mysql -P 9306
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 1
Server version: 0.9.9-dev (r1734)

Type 'help;' or '\h' for help. Type '\c' to clear the buffer.

mysql> SELECT * FROM test1 WHERE MATCH('test')
    -> ORDER BY group_id ASC OPTION ranker=bm25;
| id   | weight | group_id | date_added |
|    4 |   1442 |        2 | 1231721236 |
|    2 |   2421 |      123 | 1231721236 |
|    1 |   2421 |      456 | 1231721236 |
3 rows in set (0.00 sec)

Note that mysqld was not even running on the test machine. Everything was handled by searchd itself.

The new access method is supported in addition to native APIs which all still work perfectly well. In fact, both access methods can be used at the same time. Also, native API is still the default access method. MySQL protocol support needs to be additionally configured. This is a matter of 1-line config change, adding a new listener with mysql41 specified as a protocol:

listen = localhost:9306:mysql41

Just supporting the protocol and not the SQL syntax would be useless so Sphinx now also supports a subset of SQL that we dubbed SphinxQL. It supports the standard querying all the index types with SELECT, modifying RT indexes with INSERT, REPLACE, and DELETE, and much more. Full SphinxQL reference is available in Chapter 8, SphinxQL reference.

5.11. Multi-queries

Multi-queries, or query batches, let you send multiple queries to Sphinx in one go (more formally, one network request).

Two API methods that implement multi-query mechanism are AddQuery() and RunQueries(). You can also run multiple queries with SphinxQL, see Section 8.40, “Multi-statement queries”. (In fact, regular Query() call is internally implemented as a single AddQuery() call immediately followed by RunQueries() call.) AddQuery() captures the current state of all the query settings set by previous API calls, and memorizes the query. RunQueries() actually sends all the memorized queries, and returns multiple result sets. There are no restrictions on the queries at all, except just a sanity check on a number of queries in a single batch (see Section 12.4.20, “max_batch_queries”).

Why use multi-queries? Generally, it all boils down to performance. First, by sending requests to searchd in a batch instead of one by one, you always save a bit by doing less network roundtrips. Second, and somewhat more important, sending queries in a batch enables searchd to perform certain internal optimizations. As new types of optimizations are being added over time, it generally makes sense to pack all the queries into batches where possible, so that simply upgrading Sphinx to a new version would automatically enable new optimizations. In the case when there aren't any possible batch optimizations to apply, queries will be processed one by one internally.

Why (or rather when) not use multi-queries? Multi-queries requires all the queries in a batch to be independent, and sometimes they aren't. That is, sometimes query B is based on query A results, and so can only be set up after executing query A. For instance, you might want to display results from a secondary index if and only if there were no results found in a primary index. Or maybe just specify offset into 2nd result set based on the amount of matches in the 1st result set. In that case, you will have to use separate queries (or separate batches).

As of 0.9.10, there are two major optimizations to be aware of: common query optimization (available since 0.9.8); and common subtree optimization (available since 0.9.10).

Common query optimization means that searchd will identify all those queries in a batch where only the sorting and group-by settings differ, and only perform searching once. For instance, if a batch consists of 3 queries, all of them are for "ipod nano", but 1st query requests top-10 results sorted by price, 2nd query groups by vendor ID and requests top-5 vendors sorted by rating, and 3rd query requests max price, full-text search for "ipod nano" will only be performed once, and its results will be reused to build 3 different result sets.

So-called faceted searching is a particularly important case that benefits from this optimization. Indeed, faceted searching can be implemented by running a number of queries, one to retrieve search results themselves, and a few other ones with same full-text query but different group-by settings to retrieve all the required groups of results (top-3 authors, top-5 vendors, etc). And as long as full-text query and filtering settings stay the same, common query optimization will trigger, and greatly improve performance.

Common subtree optimization is even more interesting. It lets searchd exploit similarities between batched full-text queries. It identifies common full-text query parts (subtrees) in all queries, and caches them between queries. For instance, look at the following query batch:

barack obama president
barack obama john mccain
barack obama speech

There's a common two-word part ("barack obama") that can be computed only once, then cached and shared across the queries. And common subtree optimization does just that. Per-query cache size is strictly controlled by subtree_docs_cache and subtree_hits_cache directives (so that caching all sixteen gazillions of documents that match "i am" does not exhaust the RAM and instantly kill your server).

Here's a code sample (in PHP) that fire the same query in 3 different sorting modes:

require ( "sphinxapi.php" );
$cl = new SphinxClient ();
$cl->SetMatchMode ( SPH_MATCH_EXTENDED );

$cl->SetSortMode ( SPH_SORT_RELEVANCE );
$cl->AddQuery ( "the", "lj" );
$cl->SetSortMode ( SPH_SORT_EXTENDED, "published desc" );
$cl->AddQuery ( "the", "lj" );
$cl->SetSortMode ( SPH_SORT_EXTENDED, "published asc" );
$cl->AddQuery ( "the", "lj" );
$res = $cl->RunQueries();

How to tell whether the queries in the batch were actually optimized? If they were, respective query log will have a "multiplier" field that specifies how many queries were processed together:

[Sun Jul 12 15:18:17.000 2009] 0.040 sec x3 [ext/0/rel 747541 (0,20)] [lj] the
[Sun Jul 12 15:18:17.000 2009] 0.040 sec x3 [ext/0/ext 747541 (0,20)] [lj] the
[Sun Jul 12 15:18:17.000 2009] 0.040 sec x3 [ext/0/ext 747541 (0,20)] [lj] the

Note the "x3" field. It means that this query was optimized and processed in a sub-batch of 3 queries. For reference, this is how the regular log would look like if the queries were not batched:

[Sun Jul 12 15:18:17.062 2009] 0.059 sec [ext/0/rel 747541 (0,20)] [lj] the
[Sun Jul 12 15:18:17.156 2009] 0.091 sec [ext/0/ext 747541 (0,20)] [lj] the
[Sun Jul 12 15:18:17.250 2009] 0.092 sec [ext/0/ext 747541 (0,20)] [lj] the

Note how per-query time in multi-query case was improved by a factor of 1.5x to 2.3x, depending on a particular sorting mode. In fact, for both common query and common subtree optimizations, there were reports of 3x and even more improvements, and that's from production instances, not just synthetic tests.

5.12. Collations

Introduced to Sphinx in version 2.0.1-beta to supplement string sorting, collations essentially affect the string attribute comparisons. They specify both the character set encoding and the strategy that Sphinx uses to compare strings when doing ORDER BY or GROUP BY with a string attribute involved.

String attributes are stored as is when indexing, and no character set or language information is attached to them. That's okay as long as Sphinx only needs to store and return the strings to the calling application verbatim. But when you ask Sphinx to sort by a string value, that request immediately becomes quite ambiguous.

First, single-byte (ASCII, or ISO-8859-1, or Windows-1251) strings need to be processed differently that the UTF-8 ones that may encode every character with a variable number of bytes. So we need to know what is the character set type to interpret the raw bytes as meaningful characters properly.

Second, we additionally need to know the language-specific string sorting rules. For instance, when sorting according to US rules in en_US locale, the accented character '�' (small letter i with diaeresis) should be placed somewhere after 'z'. However, when sorting with French rules and fr_FR locale in mind, it should be placed between 'i' and 'j'. And some other set of rules might choose to ignore accents at all, allowing '�' and 'i' to be mixed arbitrarily.

Third, but not least, we might need case-sensitive sorting in some scenarios and case-insensitive sorting in some others.

Collations combine all of the above: the character set, the language rules, and the case sensitivity. Sphinx currently provides the following four collations.

  1. libc_ci

  2. libc_cs

  3. utf8_general_ci

  4. binary

The first two collations rely on several standard C library (libc) calls and can thus support any locale that is installed on your system. They provide case-insensitive (_ci) and case-sensitive (_cs) comparisons respectively. By default they will use C locale, effectively resorting to bytewise comparisons. To change that, you need to specify a different available locale using collation_libc_locale directive. The list of locales available on your system can usually be obtained with the locale command:

$ locale -a

The specific list of the system locales may vary. Consult your OS documentation to install additional needed locales.

utf8_general_ci and binary locales are built-in into Sphinx. The first one is a generic collation for UTF-8 data (without any so-called language tailoring); it should behave similar to utf8_general_ci collation in MySQL. The second one is a simple bytewise comparison.

Collation can be overridden via SphinxQL on a per-session basis using SET collation_connection statement. All subsequent SphinxQL queries will use this collation. SphinxAPI and SphinxSE queries will use the server default collation, as specified in collation_server configuration directive. Sphinx currently defaults to libc_ci collation.

Collations should affect all string attribute comparisons, including those within ORDER BY and GROUP BY, so differently ordered or grouped results can be returned depending on the collation chosen. Note that collations don't affect full-text searching, for that use charset_table.

Chapter 6. Extending Sphinx

6.1. Sphinx UDFs (User Defined Functions)

Starting with 2.0.1-beta, our expression engine can be extended with user defined functions, or UDFs for short, like this:

SELECT id, attr1, myudf(attr2, attr3+attr4) ...

You can load and unload UDFs dynamically into searchd without having to restart the daemon, and used them in expressions when searching, ranking, etc. Quick summary of the UDF features is as follows.

  • UDFs can take integer (both 32-bit and 64-bit), float, string, MVA, or PACKEDFACTORS() arguments.

  • UDFs can return integer, float, or string values.

  • UDFs can check the argument number, types, and names during the query setup phase, and raise errors.

  • Aggregation UDFs are not yet supported (but might be in the future).

UDFs have a wide variety of uses, for instance:

  • adding custom mathematical or string functions;

  • accessing the database or files from within Sphinx;

  • implementing complex ranking functions.

UDFs reside in the external dynamic libraries (.so files on UNIX and .dll on Windows systems). Library files need to reside in a trusted folder specified by plugin_dir directive, for obvious security reasons: securing a single folder is easy; letting anyone install arbitrary code into searchd is a risk. You can load and unload them dynamically into searchd with CREATE FUNCTION and DROP FUNCTION SphinxQL statements respectively. Sphinx keeps track of the currently loaded functions, that is, every time you create or drop an UDF, searchd writes its state to the sphinxql_state file as a plain good old SQL script.

Once you successfully load an UDF, you can use it in your SELECT or other statements just as well as any of the builtin functions:

SELECT id, MYCUSTOMFUNC(groupid, authorname), ... FROM myindex

UDFs are completely supported in workers=threads mode only. They are partially supported in workers=prefork mode too: namely, CREATEs from the sphinxql_state startup script will work and those UDFs will be accessible. However, DROPs will not be available. UDFs are not supported in workers=fork mode.

Multiple UDFs (and other plugins) may reside in a single library. That library will only be loaded once. It gets automatically unloaded once all the UDFs and plugins from it are dropped.

In theory you can write an UDF in any language as long as its compiler is able to import standard C header, and emit standard dynamic libraries with properly exported functions. Of course, the path of least resistance is to write in either C++ or plain C. We provide an example UDF library written in plain C and implementing several functions (demonstrating a few different techniques) along with our source code, see src/udfexample.c. That example includes src/sphinxudf.h header file definitions of a few UDF related structures and types. For most UDFs and plugins, a mere #include "sphinxudf.h", like in the example, should be completely sufficient, too. However, if you're writing a ranking function and need to access the ranking signals (factors) data from within the UDF, you will also need to compile and link with src/sphinxudf.c (also available in our source code), because the implementations of the fuctions that let you access the signal data from within the UDF reside in that file.

Both sphinxudf.h header and sphinxudf.c are standalone. So you can copy around those files only; they do not depend on any other bits of Sphinx source code.

Within your UDF, you must implement and export only a couple functions, literally. First, for UDF interface version control, you must define a function int LIBRARYNAME_ver(), where LIBRARYNAME is the name of your library file, and you must return SPH_UDF_VERSION (a value defined in sphinxudf.h) from it. Here's an example.

#include <sphinxudf.h>

// our library will be called udfexample.so, thus, so it must define
// a version function named udfexample_ver()
int udfexample_ver()
    return SPH_UDF_VERSION;

That protects you from accidentally loading a library with a mismatching UDF interface version into a newer or older searchd. Second, yout must implement the actual function, too. sphinx_int64_t testfunc ( SPH_UDF_INIT * init, SPH_UDF_ARGS * args, char * error_flag ) { return 123; }

UDF function names in SphinxQL are case insensitive. However, the respective C function names are not, they need to be all lower-case, or the UDF will not load. More importantly, it is vital that a) the calling convention is C (aka __cdecl), b) arguments list matches the plugin system expectations exactly, and c) the return type matches the one you specify in CREATE FUNCTION. Unfortunately, there is no (easy) way for us to check for those mistakes when loading the function, and they could crash the server and/or result in unexpected results. Last but not least, all the C functions you implement need to be thread-safe.

The first argument, a pointer to SPH_UDF_INIT structure, is essentially a pointer to our function state. It is option. In the example just above the function is stateless, it simply returns 123 every time it gets called. So we do not have to define an initialization function, and we can simply ignore that argument.

The second argument, a pointer to SPH_UDF_ARGS, is the most important one. All the actual call arguments are passed to your UDF via this structure; it contians the call argument count, names, types, etc. So whether your function gets called like SELECT id, testfunc(1) or like SELECT id, testfunc('abc', 1000*id+gid, WEIGHT()) or anyhow else, it will receive the very same SPH_UDF_ARGS structure in all of these cases. However, the data passed in the args structure will be different. In the first example args->arg_count will be set to 1, in the second example it will be set to 3, args->arg_types array will contain different type data, and so on.

Finally, the third argument is an error flag. UDF can raise it to indicate that some kinda of an internal error happened, the UDF can not continue, and the query should terminate early. You should not use this for argument type checks or for any other error reporting that is likely to happen during normal use. This flag is designed to report sudden critical runtime errors, such as running out of memory.

If we wanted to, say, allocate temporary storage for our function to use, or check upfront whether the arguments are of the supported types, then we would need to add two more functions, with UDF initialization and deinitialization, respectively.

int testfunc_init ( SPH_UDF_INIT * init, SPH_UDF_ARGS * args,
    char * error_message )
    // allocate and initialize a little bit of temporary storage
    init->func_data = malloc ( sizeof(int) );
    *(int*)init->func_data = 123;

    // return a success code
    return 0;

void testfunc_deinit ( SPH_UDF_INIT * init )
    // free up our temporary storage
    free ( init->func_data );

Note how testfunc_init() also receives the call arguments structure. By the time it is called it does not receive any actual values, so the args->arg_values will be NULL. But the argument names and types are known and will be passed. You can check them in the initialization function and return an error if they are of an unsupported type.

UDFs can receive arguments of pretty much any valid internal Sphinx type. Refer to sphinx_udf_argtype enumeration in sphinxudf.h for a full list. Most of the types map straightforwardly to the respective C types. The most notable exception is the SPH_UDF_TYPE_FACTORS argument type. You get that type by calling your UDF with a PACKEDFACTOR() argument. It's data is a binary blob in a certain internal format, and to extract individual ranking signals from that blob, you need to use either of the two sphinx_factors_XXX() or sphinx_get_YYY_factor() families of functions. The first family consists of just 3 functions, sphinx_factors_init() that initializes the unpacked SPH_UDF_FACTORS structure, sphinx_factors_unpack() that unpacks a binary blob into it, and sphinx_factors_deinit() that cleans up an deallocates the SPH_UDF_FACTORS. So you need to call init() and unpack(), then you can use the SPH_UDF_FACTORS fields, and then you need to cleanup with deinit(). That is simple, but results in a bunch of memory allocations per each processed document, and might be slow. The other interface, consisting of a bunch of sphinx_get_YYY_factor() functions, is a little more wordy to use, but accesses the blob data directly and guarantees that there will be zero allocations. So for top-notch ranking UDF performance, you want to use that one.

As for the return types, UDFs can currently return a signle INTEGER, BIGINT, FLOAT, or STRING value. The C function return type should be sphinx_int64_t, sphinx_int64_t, double, or char* respectively. In the last case you must use args->fn_malloc function to allocate the returned string values. Internally in your UDF you can use whatever you want, so the testfunc_init() example above is correct code even though it uses malloc() directly: you manage that pointer yourself, it gets freed up using a matching free() call, and all is well. However, the returned strings values are managed by Sphinx and we have our own allocator, so for the return values specifically, you need to use it too.

Depending on how your UDFs are used in the query, the main function call (testfunc() in our example) might be called in a rather different volume and order. Specifically,

  • UDFs referenced in WHERE, ORDER BY, or GROUP BY clauses must and will be evaluated for every matched document. They will be called in the natural matching order.

  • without subselects, UDFs that can be evaluated at the very last stage over the final result set will be evaluated that way, but before applying the LIMIT clause. They will be called in the result set order.

  • with subselects, such UDFs will also be evaluated after applying the inner LIMIT clause.

The calling sequence of the other functions is fixed, though. Namely,

  • testfunc_init() is called once when initializing the query. It can return a non-zero code to indicate a failure; in that case query will be terminated, and the error message from the error_message buffer will be returned.

  • testfunc() is called for every eligible row (see above), whenever Sphinx needs to compute the UDF value. It can also indicate an (internal) failure error by writing a non-zero byte value to error_flag. In that case, it is guaranteed that will no more be called for subsequent rows, and a default return value of 0 will be substituted. Sphinx might or might not choose to terminate such queries early, neither behavior is currently guaranteed.

  • testfunc_deinit() is called once when the query processing (in a given index shard) ends.

As of 2.2.2-beta, we do not yet support aggregation functions. In other words, your UDFs will be called for just a single document at a time and are expected to return some value for that document. Writing a function that can compute an aggregate value like AVG() over the entire group of documents that share the same GROUP BY key is not yet possible. However, you can use UDFs within the builtin aggregate functions: that is, even though MYCUSTOMAVG() is not supported yet, AVG(MYCUSTOMFUNC()) should work alright!

UDFs are local. In order to use them on a cluster, you have to put the same library on all its nodes and run CREATEs on all the nodes too. This might change in the future versions.

6.2. Sphinx plugins

Starting with version 2.2.2-beta, we generalized our dynamic plugin system, and added a few more types of dynamic plugins. Here's the complete plugin type list.

  • UDF plugins;

  • ranker plugins;

  • indexing-time token filter plugins;

  • query-time token filter plugins.

This section discusses writing and managing plugins in general; things specific to writing this or that type of a plugin are then discussed in their respective subsections.

So, how do you write and use a plugin? Three-line crash course goes as follows:

  • create a dynamic library (either .so or.dll), most likely in C or C++;

  • load that plugin into searchd using CREATE PLUGIN;

  • invoke it using the plugin specific calls (typically using this or that OPTION).

Note that while UDFs are first-class plugins they are nevertheless installed using a separate CREATE FUNCTION statement. It lets you specify the return type neatly so there was especially little reason to ruin backwards compatibility and change the syntax.

Dynamic plugins are supported in workers=threads mode only. Multiple plugins (and/or UDFs) may reside in a single library file. So you might choose to either put all your project-specific plugins in a single common uber-library; or you might choose to have a separate library for every UDF and plugin; that is up to you.

Just as with UDFs, you want to include src/sphinxudf.h header file. At the very least, you will need the SPH_UDF_VERSION constant to implement a proper version function. Depending on the specific plugin type, you might or might not need to link your plugin with src/sphinxudf.c. However, as of 2.2.2-beta all the functions implemented in sphinxudf.c are about unpacking the PACKEDFACTORS() blob, and no plugin types are exposed to that kind of data. So currently, you would never need to link with the C-file, just the header would be sufficient. (In fact, if you copy over the UDF version number, then for some of the plugin types you would not even need the header file.)

Formally, plugins are just sets of C functions that follow a certain naming parttern. You are typically required to define just one key function that does the most important work, but you may define a bunch of other functions, too. For example, to implement a ranker called "myrank", you must define myrank_finalize() function that actually returns the rank value, however, you might also define myrank_init(), myrank_update(), and myrank_deinit() functions. Specific sets of well-known suffixes and the call arguments do differ based on the plugin type, but _init() and _deinit() are generic, every plugin has those. Protip: for a quick reference on the known suffixes and their argument types, refer to sphinxplugin.h, we define the call prototoypes in the very beginning of that file.

Despite having the public interface defined in ye good olde good pure C, our plugins essentially follow the object-oriented model. Indeed, every _init() function receives a void ** userdata out-parameter. And the pointer value that you store at (*userdata) location is then be passed as a 1st argument to all the other plugin functions. So you can think of a plugin as class that gets instantiated every time an object of that class is needed to handle a request: the userdata pointer would be its this pointer; the functions would be its methods, and the _init() and _deinit() functions would be the constructor and destructor respectively.

Why this (minor) OOP-in-C complication? Well, plugins run in a multi-threaded environment, and some of them have to be stateful. You can't keep that state in a global variable in your plugin. So we have to pass around a userdata parameter anyway to let you keep that state. And that naturally brings us to the OOP model. And if you've got a simple, stateless plugin, the interface lets you omit the _init() and _deinit() and whatever other functions just as well.

To summarize, here goes the simplest complete ranker plugin, in just 3 lines of C code.

// gcc -fPIC -shared -o myrank.so myrank.c
#include "sphinxudf.h"
int myrank_ver() { return SPH_UDF_VERSION; }
int myrank_finalize(void *u, int w) { return 123; }

And this is how you use it:

mysql> CREATE PLUGIN myrank TYPE 'ranker' SONAME 'myrank.dll';
Query OK, 0 rows affected (0.00 sec)

mysql> SELECT id, weight() FROM test1 WHERE MATCH('test')
    -> OPTION ranker=myrank('');
| id   | weight() |
|    1 |      123 |
|    2 |      123 |
2 rows in set (0.01 sec)

6.3. Ranker plugins

Ranker plugins let you implement a custom ranker that receives all the occurrences of the keywords matched in the document, and computes a WEIGHT() value. They can be called as follows:

SELECT id, attr1 FROM test WHERE match('hello')
OPTION ranker=myranker('option1=1');

The call workflow is as follows:

  1. XXX_init() gets called once per query per index, in the very beginning. A few query-wide options are passed to it through a SPH_RANKER_INIT structure, including the user options strings (in the example just above, "option1=1" is that string).
  2. XXX_update() gets called multiple times per matched document, with every matched keyword occurrence passed as its parameter, a SPH_RANKER_HIT structure. The occurrences within each document are guaranteed to be passed in the order of ascending hit->hit_pos values.
  3. XXX_finalize() gets called once per matched document, once there are no more keyword occurrences. It must return the WEIGHT() value. This is the only mandatory function.
  4. XXX_deinit() gets called once per query, in the very end.

Chapter 7. Command line tools reference

As mentioned elsewhere, Sphinx is not a single program called 'sphinx', but a collection of 4 separate programs which collectively form Sphinx. This section covers these tools and how to use them.

7.1. indexer command reference

indexer is the first of the two principal tools as part of Sphinx. Invoked from either the command line directly, or as part of a larger script, indexer is solely responsible for gathering the data that will be searchable.

The calling syntax for indexer is as follows:

indexer [OPTIONS] [indexname1 [indexname2 [...]]]

Essentially you would list the different possible indexes (that you would later make available to search) in sphinx.conf, so when calling indexer, as a minimum you need to be telling it what index (or indexes) you want to index.

If sphinx.conf contained details on 2 indexes, mybigindex and mysmallindex, you could do the following:

$ indexer mybigindex
$ indexer mysmallindex mybigindex

As part of the configuration file, sphinx.conf, you specify one or more indexes for your data. You might call indexer to reindex one of them, ad-hoc, or you can tell it to process all indexes - you are not limited to calling just one, or all at once, you can always pick some combination of the available indexes.

The exit codes are as follows:

  • 0, everything went ok
  • 1, there was a problem while indexing (and if --rotate was specified, it was skipped)
  • 2, indexing went ok, but --rotate attempt failed

The majority of the options for indexer are given in the configuration file, however there are some options you might need to specify on the command line as well, as they can affect how the indexing operation is performed. These options are:

  • --config <file> (-c <file> for short) tells indexer to use the given file as its configuration. Normally, it will look for sphinx.conf in the installation directory (e.g. /usr/local/sphinx/etc/sphinx.conf if installed into /usr/local/sphinx), followed by the current directory you are in when calling indexer from the shell. This is most of use in shared environments where the binary files are installed somewhere like /usr/local/sphinx/ but you want to provide users with the ability to make their own custom Sphinx set-ups, or if you want to run multiple instances on a single server. In cases like those you could allow them to create their own sphinx.conf files and pass them to indexer with this option. For example:

    $ indexer --config /home/myuser/sphinx.conf myindex

  • --all tells indexer to update every index listed in sphinx.conf, instead of listing individual indexes. This would be useful in small configurations, or cron-type or maintenance jobs where the entire index set will get rebuilt each day, or week, or whatever period is best. Example usage:

    $ indexer --config /home/myuser/sphinx.conf --all

  • --rotate is used for rotating indexes. Unless you have the situation where you can take the search function offline without troubling users, you will almost certainly need to keep search running whilst indexing new documents. --rotate creates a second index, parallel to the first (in the same place, simply including .new in the filenames). Once complete, indexer notifies searchd via sending the SIGHUP signal, and searchd will attempt to rename the indexes (renaming the existing ones to include .old and renaming the .new to replace them), and then start serving from the newer files. Depending on the setting of seamless_rotate, there may be a slight delay in being able to search the newer indexes. Example usage:

    $ indexer --rotate --all

  • --quiet tells indexer not to output anything, unless there is an error. Again, most used for cron-type, or other script jobs where the output is irrelevant or unnecessary, except in the event of some kind of error. Example usage:

    $ indexer --rotate --all --quiet

  • --noprogress does not display progress details as they occur; instead, the final status details (such as documents indexed, speed of indexing and so on are only reported at completion of indexing. In instances where the script is not being run on a console (or 'tty'), this will be on by default. Example usage:

    $ indexer --rotate --all --noprogress

  • --buildstops <outputfile.text> <N> reviews the index source, as if it were indexing the data, and produces a list of the terms that are being indexed. In other words, it produces a list of all the searchable terms that are becoming part of the index. Note; it does not update the index in question, it simply processes the data 'as if' it were indexing, including running queries defined with sql_query_pre or sql_query_post. outputfile.txt will contain the list of words, one per line, sorted by frequency with most frequent first, and N specifies the maximum number of words that will be listed; if sufficiently large to encompass every word in the index, only that many words will be returned. Such a dictionary list could be used for client application features around "Did you mean..." functionality, usually in conjunction with --buildfreqs, below. Example:

    $ indexer myindex --buildstops word_freq.txt 1000

    This would produce a document in the current directory, word_freq.txt with the 1,000 most common words in 'myindex', ordered by most common first. Note that the file will pertain to the last index indexed when specified with multiple indexes or --all (i.e. the last one listed in the configuration file)

  • --buildfreqs works with --buildstops (and is ignored if --buildstops is not specified). As --buildstops provides the list of words used within the index, --buildfreqs adds the quantity present in the index, which would be useful in establishing whether certain words should be considered stopwords if they are too prevalent. It will also help with developing "Did you mean..." features where you can how much more common a given word compared to another, similar one. Example:

    $ indexer myindex --buildstops word_freq.txt 1000 --buildfreqs

    This would produce the word_freq.txt as above, however after each word would be the number of times it occurred in the index in question.

  • --merge <dst-index> <src-index> is used for physically merging indexes together, for example if you have a main+delta scheme, where the main index rarely changes, but the delta index is rebuilt frequently, and --merge would be used to combine the two. The operation moves from right to left - the contents of src-index get examined and physically combined with the contents of dst-index and the result is left in dst-index. In pseudo-code, it might be expressed as: dst-index += src-index An example:

    $ indexer --merge main delta --rotate

    In the above example, where the main is the master, rarely modified index, and delta is the less frequently modified one, you might use the above to call indexer to combine the contents of the delta into the main index and rotate the indexes.

  • --merge-dst-range <attr> <min> <max> runs the filter range given upon merging. Specifically, as the merge is applied to the destination index (as part of --merge, and is ignored if --merge is not specified), indexer will also filter the documents ending up in the destination index, and only documents will pass through the filter given will end up in the final index. This could be used for example, in an index where there is a 'deleted' attribute, where 0 means 'not deleted'. Such an index could be merged with:

    $ indexer --merge main delta --merge-dst-range deleted 0 0

    Any documents marked as deleted (value 1) would be removed from the newly-merged destination index. It can be added several times to the command line, to add successive filters to the merge, all of which must be met in order for a document to become part of the final index.

  • --merge-killlists (and its shorter alias --merge-klists) changes the way kill lists are processed when merging indexes. By default, both kill lists get discarded after a merge. That supports the most typical main+delta merge scenario. With this option enabled, however, kill lists from both indexes get concatenated and stored into the destination index. Note that a source (delta) index kill list will be used to suppress rows from a destination (main) index at all times.

  • --keep-attrs (added in version 2.1.1-beta) allows to reuse existing attributes on reindexing. Whenever the index is rebuilt, each new document id is checked for presence in the "old" index, and if it already exists, its attributes are transferred to the "new" index; if not found, attributes from the new index are used. If the user has updated attributes in the index, but not in the actual source used for the index, all updates will be lost when reindexing; using --keep-attrs enables saving the updated attribute values from the previous index

  • --dump-rows <FILE> dumps rows fetched by SQL source(s) into the specified file, in a MySQL compatible syntax. Resulting dumps are the exact representation of data as received by indexer and help to repeat indexing-time issues.

  • --verbose guarantees that every row that caused problems indexing (duplicate, zero, or missing document ID; or file field IO issues; etc) will be reported. By default, this option is off, and problem summaries may be reported instead.

  • --sighup-each is useful when you are rebuilding many big indexes, and want each one rotated into searchd as soon as possible. With --sighup-each, indexer will send a SIGHUP signal to searchd after successfully completing the work on each index. (The default behavior is to send a single SIGHUP after all the indexes were built.)

  • --nohup is useful when you want to check your index with indextool before actually rotating it. indexer won't send SIGHUP if this option is on.

  • --print-queries prints out SQL queries that indexer sends to the database, along with SQL connection and disconnection events. That is useful to diagnose and fix problems with SQL sources.

7.2. searchd command reference

searchd is the second of the two principle tools as part of Sphinx. searchd is the part of the system which actually handles searches; it functions as a server and is responsible for receiving queries, processing them and returning a dataset back to the different APIs for client applications.

Unlike indexer, searchd is not designed to be run either from a regular script or command-line calling, but instead either as a daemon to be called from init.d (on Unix/Linux type systems) or to be called as a service (on Windows-type systems), so not all of the command line options will always apply, and so will be build-dependent.

Calling searchd is simply a case of:

$ searchd [OPTIONS]

The options available to searchd on all builds are:

  • --help (-h for short) lists all of the parameters that can be called in your particular build of searchd.

  • --config <file> (-c <file> for short) tells searchd to use the given file as its configuration, just as with indexer above.

  • --stop is used to asynchronously stop searchd, using the details of the PID file as specified in the sphinx.conf file, so you may also need to confirm to searchd which configuration file to use with the --config option. NB, calling --stop will also make sure any changes applied to the indexes with UpdateAttributes() will be applied to the index files themselves. Example:

    $ searchd --config /home/myuser/sphinx.conf --stop

  • --stopwait is used to synchronously stop searchd. --stop essentially tells the running instance to exit (by sending it a SIGTERM) and then immediately returns. --stopwait will also attempt to wait until the running searchd instance actually finishes the shutdown (eg. saves all the pending attribute changes) and exits. Example:

    $ searchd --config /home/myuser/sphinx.conf --stopwait

    Possible exit codes are as follows:

    • 0 on success;

    • 1 if connection to running searchd daemon failed;

    • 2 if daemon reported an error during shutdown;

    • 3 if daemon crashed during shutdown.

  • --status command is used to query running searchd instance status, using the connection details from the (optionally) provided configuration file. It will try to connect to the running instance using the first configured UNIX socket or TCP port. On success, it will query for a number of status and performance counter values and print them. You can use Status() API call to access the very same counters from your application. Examples:

    $ searchd --status
    $ searchd --config /home/myuser/sphinx.conf --status

  • --pidfile is used to explicitly force using a PID file (where the searchd process number is stored) despite any other debugging options that say otherwise (for instance, --console). This is a debugging option.

    $ searchd --console --pidfile

  • --console is used to force searchd into console mode; typically it will be running as a conventional server application, and will aim to dump information into the log files (as specified in sphinx.conf). Sometimes though, when debugging issues in the configuration or the daemon itself, or trying to diagnose hard-to-track-down problems, it may be easier to force it to dump information directly to the console/command line from which it is being called. Running in console mode also means that the process will not be forked (so searches are done in sequence) and logs will not be written to. (It should be noted that console mode is not the intended method for running searchd.) You can invoke it as such:

    $ searchd --config /home/myuser/sphinx.conf --console

  • --logdebug, --logdebugv, and --logdebugvv options enable additional debug output in the daemon log. They differ by the logging verboseness level. These are debugging options, they pollute the log a lot, and thus they should not be normally enabled. (The normal use case for these is to enable them temporarily on request, to assist with some particularly complicated debugging session.)

  • --iostats is used in conjunction with the logging options (the query_log will need to have been activated in sphinx.conf) to provide more detailed information on a per-query basis as to the input/output operations carried out in the course of that query, with a slight performance hit and of course bigger logs. Further details are available under the query log format section. You might start searchd thus:

    $ searchd --config /home/myuser/sphinx.conf --iostats

  • --cpustats is used to provide actual CPU time report (in addition to wall time) in both query log file (for every given query) and status report (aggregated). It depends on clock_gettime() system call and might therefore be unavailable on certain systems. You might start searchd thus:

    $ searchd --config /home/myuser/sphinx.conf --cpustats

  • --port portnumber (-p for short) is used to specify the port that searchd should listen on, usually for debugging purposes. This will usually default to 9312, but sometimes you need to run it on a different port. Specifying it on the command line will override anything specified in the configuration file. The valid range is 0 to 65535, but ports numbered 1024 and below usually require a privileged account in order to run. An example of usage:

    $ searchd --port 9313

  • --listen ( address ":" port | port | path ) [ ":" protocol ] (or -l for short) Works as --port, but allow you to specify not only the port, but full path, as IP address and port, or Unix-domain socket path, that searchd will listen on. Otherwords, you can specify either an IP address (or hostname) and port number, or just a port number, or Unix socket path. If you specify port number but not the address, searchd will listen on all network interfaces. Unix path is identified by a leading slash. As the last param you can also specify a protocol handler (listener) to be used for connections on this socket. Supported protocol values are 'sphinx' (Sphinx 0.9.x API protocol) and 'mysql41' (MySQL protocol used since 4.1 upto at least 5.1).

  • --index <index> (or -i <index> for short) forces this instance of searchd only to serve the specified index. Like --port, above, this is usually for debugging purposes; more long-term changes would generally be applied to the configuration file itself. Example usage:

    $ searchd --index myindex

  • --strip-path strips the path names from all the file names referenced from the index (stopwords, wordforms, exceptions, etc). This is useful for picking up indexes built on another machine with possibly different path layouts.

  • --replay-flags=<OPTIONS> switch, added in version 2.0.2-beta, can be used to specify a list of extra binary log replay options. The supported options are:

    • accept-desc-timestamp, ignore descending transaction timestamps and replay such transactions anyway (the default behavior is to exit with an error).


    $ searchd --replay-flags=accept-desc-timestamp

There are some options for searchd that are specific to Windows platforms, concerning handling as a service, are only be available on Windows binaries.

Note that on Windows searchd will default to --console mode, unless you install it as a service.

  • --install installs searchd as a service into the Microsoft Management Console (Control Panel / Administrative Tools / Services). Any other parameters specified on the command line, where --install is specified will also become part of the command line on future starts of the service. For example, as part of calling searchd, you will likely also need to specify the configuration file with --config, and you would do that as well as specifying --install. Once called, the usual start/stop facilities will become available via the management console, so any methods you could use for starting, stopping and restarting services would also apply to searchd. Example:

    C:\WINDOWS\system32> C:\Sphinx\bin\searchd.exe --install
       --config C:\Sphinx\sphinx.conf

    If you wanted to have the I/O stats every time you started searchd, you would specify its option on the same line as the --install command thus:

    C:\WINDOWS\system32> C:\Sphinx\bin\searchd.exe --install
       --config C:\Sphinx\sphinx.conf --iostats

  • --delete removes the service from the Microsoft Management Console and other places where services are registered, after previously installed with --install. Note, this does not uninstall the software or delete the indexes. It means the service will not be called from the services systems, and will not be started on the machine's next start. If currently running as a service, the current instance will not be terminated (until the next reboot, or searchd is called with --stop). If the service was installed with a custom name (with --servicename), the same name will need to be specified with --servicename when calling to uninstall. Example:

    C:\WINDOWS\system32> C:\Sphinx\bin\searchd.exe --delete

  • --servicename <name> applies the given name to searchd when installing or deleting the service, as would appear in the Management Console; this will default to searchd, but if being deployed on servers where multiple administrators may log into the system, or a system with multiple searchd instances, a more descriptive name may be applicable. Note that unless combined with --install or --delete, this option does not do anything. Example:

    C:\WINDOWS\system32> C:\Sphinx\bin\searchd.exe --install
       --config C:\Sphinx\sphinx.conf --servicename SphinxSearch

  • --ntservice is the option that is passed by the Management Console to searchd to invoke it as a service on Windows platforms. It would not normally be necessary to call this directly; this would normally be called by Windows when the service would be started, although if you wanted to call this as a regular service from the command-line (as the complement to --console) you could do so in theory.

  • --safetrace forces searchd to only use system backtrace() call in crash reports. In certain (rare) scenarios, this might be a "safer" way to get that report. This is a debugging option.

  • --nodetach switch (Linux only) tells searchd not to detach into background. This will also cause log entry to be printed out to console. Query processing operates as usual. This is a debugging option.

Last but not least, as every other daemon, searchd supports a number of signals.


Initiates a clean shutdown. New queries will not be handled; but queries that are already started will not be forcibly interrupted.


Initiates index rotation. Depending on the value of seamless_rotate setting, new queries might be shortly stalled; clients will receive temporary errors.


Forces reopen of searchd log and query log files, letting you implement log file rotation.

7.3. spelldump command reference

spelldump is one of the helper tools within the Sphinx package.

It is used to extract the contents of a dictionary file that uses ispell or MySpell format, which can help build word lists for wordforms - all of the possible forms are pre-built for you.

Its general usage is:

spelldump [options] <dictionary> <affix> [result] [locale-name]

The two main parameters are the dictionary's main file and its affix file; usually these are named as [language-prefix].dict and [language-prefix].aff and will be available with most common Linux distributions, as well as various places online.

[result] specifies where the dictionary data should be output to, and [locale-name] additionally specifies the locale details you wish to use.

There is an additional option, -c [file], which specifies a file for case conversion details.

Examples of its usage are:

spelldump en.dict en.aff
spelldump ru.dict ru.aff ru.txt ru_RU.CP1251
spelldump ru.dict ru.aff ru.txt .1251

The results file will contain a list of all the words in the dictionary in alphabetical order, output in the format of a wordforms file, which you can use to customize for your specific circumstances. An example of the result file:

zone > zone
zoned > zoned
zoning > zoning

7.4. indextool command reference

indextool is one of the helper tools within the Sphinx package, introduced in version 0.9.9-rc2. It is used to dump miscellaneous debug information about the physical index. (Additional functionality such as index verification is planned in the future, hence the indextool name rather than just indexdump.) Its general usage is:

indextool <command> [options]

Options apply to all commands:

  • --config <file> (-c <file> for short) overrides the built-in config file names.

  • --quiet (-q for short) keep indextool quiet - it will not output banner, etc.

The commands are as follows:

  • --checkconfig just loads and verifies the config file to check if it's valid, without syntax errors. This option was added in version 2.1.1-beta.

  • --build-infixes INDEXNAME build infixes for an existing dict=keywords index (upgrades .sph, .spi in place). You can use this option for legacy index files that already use dict=keywords, but now need to support infix searching too; updating the index files with indextool may prove easier or faster than regenerating them from scratch with indexer. This option was added in version 2.1.1-beta.

  • --dumpheader FILENAME.sph quickly dumps the provided index header file without touching any other index files or even the configuration file. The report provides a breakdown of all the index settings, in particular the entire attribute and field list. Prior to 0.9.9-rc2, this command was present in now removed CLI search utility.

  • --dumpconfig FILENAME.sph dumps the index definition from the given index header file in (almost) compliant sphinx.conf file format. Added in version 2.0.1-beta.

  • --dumpheader INDEXNAME dumps index header by index name with looking up the header path in the configuration file.

  • --dumpdict INDEXNAME dumps dictionary. This was added in version 2.1.1-beta.

  • --dumpdocids INDEXNAME dumps document IDs by index name. It takes the data from attribute (.spa) file and therefore requires docinfo=extern to work.

  • --dumphitlist INDEXNAME KEYWORD dumps all the hits (occurrences) of a given keyword in a given index, with keyword specified as text.

  • --dumphitlist INDEXNAME --wordid ID dumps all the hits (occurrences) of a given keyword in a given index, with keyword specified as internal numeric ID.

  • --fold INDEXNAME OPTFILE This options is useful too see how actually tokenizer proceeds input. You can feed indextool with text from file if specified or from stdin otherwise. The output will contain spaces instead of separators (accordingly to your charset_table settings) and lowercased letters in words.

  • --htmlstrip INDEXNAME filters stdin using HTML stripper settings for a given index, and prints the filtering results to stdout. Note that the settings will be taken from sphinx.conf, and not the index header.

  • --morph INDEXNAME applies morphology to the given stdin and prints the result to stdout.

  • --check INDEXNAME checks the index data files for consistency errors that might be introduced either by bugs in indexer and/or hardware faults. Starting with version 2.1.1-beta, --check also works on RT indexes, RAM and disk chunks.

  • --strip-path strips the path names from all the file names referenced from the index (stopwords, wordforms, exceptions, etc). This is useful for checking indexes built on another machine with possibly different path layouts.

  • --optimize-rt-klists optimizes the kill list memory use in the disk chunk of a given RT index. That is a one-off optimization intended for rather old RT indexes, created by development versions prior to 1.10-beta release. As of 1.10-beta releases, this kill list optimization (purging) should happen automatically, and there should never be a need to use this option.

  • --rotate works only with --check and defines whether to check index waiting for rotation, i.e. with .new extension. This is useful when you want to check your index before actually using it.

7.5. wordbreaker command reference

wordbreaker is one of the helper tools within the Sphinx package, introduced in version 2.1.1-beta. It is used to split compound words, as usual in URLs, into its component words. For example, this tool can split "lordoftherings" into its four component words, or "http://manofsteel.warnerbros.com" into "man of steel warner bros". This helps searching, without requiring prefixes or infixes: searching for "sphinx" wouldn't match "sphinxsearch" but if you break the compound word and index the separate components, you'll get a match without the costs of prefix and infix larger index files.

Examples of its usage are:

echo manofsteel | bin/wordbreaker -dict dict.txt split

The input stream will be separated in words using the -dict dictionary file. (The dictionary should match the language of the compound word.) The split command breaks words from the standard input, and outputs the result in the standard output. There are also test and bench commands that let you test the splitting quality and benchmark the splitting functionality.

Wordbreaker Wordbreaker needs a dictionary to recognize individual substrings within a string. To differentiate between different guesses, it uses the relative frequency of each word in the dictionary: higher frequency means higher split probability. You can generate such a file using the indexer tool, as in

indexer --buildstops dict.txt 100000 --buildfreqs myindex -c /path/to/sphinx.conf   

which will write the 100,000 most frequent words, along with their counts, from myindex into dict.txt. The output file is a text file, so you can edit it by hand, if need be, to add or remove words.

See http://sphinxsearch.com/blog/2013/01/29/a-new-tool-in-the-trunk-wordbreaker/ for more on this tool.

Chapter 8. SphinxQL reference

SphinxQL is our SQL dialect that exposes all of the search daemon functionality using a standard SQL syntax with a few Sphinx-specific extensions. Everything available via the SphinxAPI is also available via SphinxQL but not vice versa; for instance, writes into RT indexes are only available via SphinxQL. This chapter documents supported SphinxQL statements syntax.

8.1. SELECT syntax

    select_expr [, select_expr ...]
    FROM index [, index2 ...]
    [WHERE where_condition]
    [GROUP [N] BY {col_name | expr_alias} [, {col_name | expr_alias}]]
    [WITHIN GROUP ORDER BY {col_name | expr_alias} {ASC | DESC}]
    [HAVING having_condition]
    [ORDER BY {col_name | expr_alias} {ASC | DESC} [, ...]]
    [LIMIT [offset,] row_count]
    [OPTION opt_name = opt_value [, ...]]
    [FACET facet_options[ FACET facet_options][ ...]]

SELECT statement was introduced in version 0.9.9-rc2. It's syntax is based upon regular SQL but adds several Sphinx-specific extensions and has a few omissions (such as (currently) missing support for JOINs). Specifically,

  • Column list clause. Column names, arbitrary expressions, and star ('*') are all allowed (ie. SELECT id, group_id*123+456 AS expr1 FROM test1 will work). Unlike in regular SQL, all computed expressions must be aliased with a valid identifier. Starting with version 2.0.1-beta, AS is optional.

  • EXIST() function (added in version 2.1.1-beta) is supported. EXIST ( "attr-name", default-value ) replaces non-existent columns with default values. It returns either a value of an attribute specified by 'attr-name', or 'default-value' if that attribute does not exist. As of 2.1.1-beta it does not support STRING or MVA attributes. This function is handy when you are searching through several indexes with different schemas.

    SELECT *, EXIST('gid', 6) as cnd FROM i1, i2 WHERE cnd>5
  • SNIPPET() function (added in version 2.1.1-beta) is supported. This is a wrapper around the snippets functionality, similar to what is available via CALL SNIPPETS. The first two arguments are: the text to highlight, and a query. Starting with 2.2-1-beta it's possible to pass options to function. The intended use is as follows:

    SELECT id, SNIPPET(myUdf(id), 'my.query', 'limit=100')
    FROM myIndex WHERE MATCH('my.query')

    where myUdf() would be a UDF that fetches a document by its ID from some external storage. This enables applications to fetch the entire result set directly from Sphinx in one query, without having to separately fetch the documents in the application and then send them back to Sphinx for highlighting.

    SNIPPET() is a so-called "post limit" function, meaning that computing snippets is postponed not just until the entire final result set is ready, but even after the LIMIT clause is applied. For example, with a LIMIT 20,10 clause, SNIPPET() will be called at most 10 times.

    Table functions is a mechanism of post-query result set processing. It was added in 2.2.1-beta. Table functions take an arbitrary result set as their input, and return a new, processed set as their output. The first argument should be the input result set, but a table function can optionally take and handle more arguments. Table functions can completely change the result set, including the schema. For now, only built in table functions are supported. UDFs are planned when the internal call interface is stabilized. Table functions work for both outer SELECT and nested SELECT.

    • REMOVE_REPEATS ( result_set, column, offset, limit ) - removes repeated adjusted rows with the same 'column' value.

    SELECT REMOVE_REPEATS((SELECT * FROM dist1), gid, 0, 10)

  • FROM clause. FROM clause should contain the list of indexes to search through. Unlike in regular SQL, comma means enumeration of full-text indexes as in Query() API call rather than JOIN. Index name should be according to the rules of a C identifier.

  • WHERE clause. This clause will map both to fulltext query and filters. Comparison operators (=, !=, <, >, <=, >=), IN, AND, NOT, and BETWEEN are all supported and map directly to filters. OR is not supported yet but will be in the future. MATCH('query') is supported and maps to fulltext query. Query will be interpreted according to full-text query language rules. There must be at most one MATCH() in the clause. Starting with version 2.0.1-beta, {col_name | expr_alias} [NOT] IN @uservar condition syntax is supported. (Refer to Section 8.9, “SET syntax” for a discussion of global user variables.)

  • GROUP BY clause. Supports grouping by multiple columns or computed expressions:

    SELECT *, group_id*1000+article_type AS gkey FROM example GROUP BY gkey
    SELECT id FROM products GROUP BY region, price

    Implicit grouping supported when using aggregate functions without specifiying a GROUP BY clause. Consider these two queries:

    SELECT MAX(id), MIN(id), COUNT(*) FROM books
    SELECT MAX(id), MIN(id), COUNT(*), 1 AS grp FROM books GROUP BY grp

    Aggregate functions (AVG(), MIN(), MAX(), SUM()) in column list clause are supported. Arguments to aggregate functions can be either plain attributes or arbitrary expressions. COUNT(*), COUNT(DISTINCT attr) are supported. Currently there can be at most one COUNT(DISTINCT) per query and an argument needs to be an attribute. Both current restrictions on COUNT(DISTINCT) might be lifted in the future. A special GROUPBY() function is also supported. It returns the GROUP BY key. That is particularly useful when grouping by an MVA value, in order to pick the specific value that was used to create the current group.

    SELECT *, AVG(price) AS avgprice, COUNT(DISTINCT storeid), GROUPBY()
    FROM products
    WHERE MATCH('ipod')
    GROUP BY vendorid

    Starting with 2.0.1-beta, GROUP BY on a string attribute is supported, with respect for current collation (see Section 5.12, “Collations”).

    Starting with 2.2.1-beta, you can query Sphinx to return (no more than) N top matches for each group accordingly to WITHIN GROUP ORDER BY.

    SELECT id FROM products GROUP 3 BY category

    You can sort the result set by (an alias of) the aggregate value.

    SELECT group_id, MAX(id) AS max_id
    FROM my_index WHERE MATCH('the')
    GROUP BY group_id ORDER BY max_id DESC

  • GROUP_CONCAT() function is supported, starting with version 2.1.1-beta. When you group by an attribute, the result set only shows attributes from a single document representing the whole group. GROUP_CONCAT() produces a comma-separated list of the attribute values of all documents in the group.

    SELECT id, GROUP_CONCAT(price) as pricesList, GROUPBY() AS name FROM shops GROUP BY shopName;
  • ZONESPANLIST() function returns pairs of matched zone spans. Each pair contains the matched zone span identifier, a colon, and the order number of the matched zone span. For example, if a document reads <emphasis role="bold"><i>text</i> the <i>text</i></emphasis>, and you query for 'ZONESPAN:(i,b) text', then ZONESPANLIST() will return the string "1:1 1:2 2:1" meaning that the first zone span matched "text" in spans 1 and 2, and the second zone span in span 1 only. This was added in version 2.1.1-beta.

  • WITHIN GROUP ORDER BY clause. This is a Sphinx specific extension that lets you control how the best row within a group will to be selected. The syntax matches that of regular ORDER BY clause:

    SELECT *, INTERVAL(posted,NOW()-7*86400,NOW()-86400) AS timeseg, WEIGHT() AS w
    FROM example WHERE MATCH('my search query')
    GROUP BY siteid
    ORDER BY timeseg DESC, w DESC

    Starting with 2.0.1-beta, WITHIN GROUP ORDER BY on a string attribute is supported, with respect for current collation (see Section 5.12, “Collations”).

  • HAVING clause. This is used to filter on GROUP BY values. It was added in 2.2.1-beta. Currently supports only one filtering condition.

    SELECT id FROM plain GROUP BY title HAVING group_id=16;
    SELECT id FROM plain GROUP BY attribute HAVING COUNT(*)>1;

    Because of HAVING is implemented as a whole result set post-processing, result set for query with HAVING could be less than max_matches allows.

  • ORDER BY clause. Unlike in regular SQL, only column names (not expressions) are allowed and explicit ASC and DESC are required. The columns however can be computed expressions:

    SELECT *, WEIGHT()*10+docboost AS skey FROM example ORDER BY skey

    Starting with 2.1.1-beta, you can use subqueries to speed up specific searches, which involve reranking, by postponing hard (slow) calculations as late as possible. For example, SELECT id,a_slow_expression() AS cond FROM an_index ORDER BY id ASC, cond DESC LIMIT 100; could be better written as SELECT * FROM (SELECT id,a_slow_expression() AS cond FROM an_index ORDER BY id ASC LIMIT 100) ORDER BY cond DESC; because in the first case the slow expression would be evaluated for the whole set, while in the second one it would be evaluated just for a subset of values.

    Starting with 2.0.1-beta, ORDER BY on a string attribute is supported, with respect for current collation (see Section 5.12, “Collations”).

    Starting with 2.0.2-beta, ORDER BY RAND() syntax is supported. Note that this syntax is actually going to randomize the weight values and then order matches by those randomized weights.

  • LIMIT clause. Both LIMIT N and LIMIT M,N forms are supported. Unlike in regular SQL (but like in Sphinx API), an implicit LIMIT 0,20 is present by default.

  • OPTION clause. This is a Sphinx specific extension that lets you control a number of per-query options. The syntax is:

    OPTION <optionname>=<value> [ , ... ]

    Supported options and respectively allowed values are:

    • 'agent_query_timeout' - integer (max time in milliseconds to wait for remote queries to complete, see agent_query_timeout under Index configuration options for details)

    • 'boolean_simplify' - 0 or 1, enables simplifying the query to speed it up

    • 'comment' - string, user comment that gets copied to a query log file

    • 'cutoff' - integer (max found matches threshold)

    • 'field_weights' - a named integer list (per-field user weights for ranking)

    • 'global_idf' - use global statistics (frequencies) from the global_idf file for IDF computations, rather than the local index statistics. Added in version 2.1.1-beta.

    • 'idf' - a quoted, comma-separated list of IDF computation flags. Added in version 2.1.1-beta. Known flags are:

      • normalized: BM25 variant, idf = log((N-n+1)/n), as per Robertson et al

      • plain: plain variant, idf = log(N/n), as per Sparck-Jones

      • tfidf_normalized (added in 2.2.1-beta): additionally divide IDF by query word count, so that TF*IDF fits into [0, 1] range

      • tfidf_unnormalized (added in 2.2.1-beta): do not additionally divide IDF by query word count

      where N is the collection size and n is the number of matched documents.

      The historically default IDF (Inverse Document Frequency) in Sphinx is equivalent to OPTION idf='normalized,tfidf_normalized', and those normalizations may cause several undesired effects.

      First, idf=normalized causes keyword penalization. For instance, if you search for [the | something] and [the] occurs in more than 50% of the documents, then documents with both keywords [the] and [something] will get less weight than documents with just one keyword [something]. Using OPTION idf=plain avoids this. Plain IDF varies in [0, log(N)] range, and keywords are never penalized; while the normalized IDF varies in [-log(N), log(N)] range, and too frequent keywords are penalized.

      Second, idf=tfidf_normalized causes IDF drift over queries. Historically, we additionally divided IDF by query keyword count, so that the entire sum(tf*idf) over all keywords would still fit into [0,1] range. However, that means that queries [word1] and [word1 | nonmatchingword2] would assign different weights to the exactly same result set, because the IDFs for both "word1" and "nonmatchingword2" would be divided by 2. OPTION idf=tfidf_unnormalized fixes that. Note that BM25, BM25A, BM25F() ranking factors will be scale accordingly once you disable this normalization.

      IDF flags can be mixed; 'plain' and 'normalized' are mutually exclusive; 'tfidf_unnormalized' and 'tfidf_normalized' are mutually exclusive; and unspecified flags in such a mutually exclusive group take their defaults. That means that OPTION idf=plain is equivalent to a complete OPTION idf='plain,tfidf_normalized' specification.

    • local_df (added in 2.2.1-beta): 0 or 1,automatically sum DFs over all the local parts of a distributed index, so that the IDF is consistent (and precise) over a locally sharded index.

    • 'index_weights' - a named integer list (per-index user weights for ranking)

    • 'max_matches' - integer (per-query max matches value)

      Maximum amount of matches that the daemon keeps in RAM for each index and can return to the client. Default is 1000.

      Introduced in order to control and limit RAM usage, max_matches setting defines how much matches will be kept in RAM while searching each index. Every match found will still be processed; but only best N of them will be kept in memory and return to the client in the end. Assume that the index contains 2,000,000 matches for the query. You rarely (if ever) need to retrieve all of them. Rather, you need to scan all of them, but only choose "best" at most, say, 500 by some criteria (ie. sorted by relevance, or price, or anything else), and display those 500 matches to the end user in pages of 20 to 100 matches. And tracking only the best 500 matches is much more RAM and CPU efficient than keeping all 2,000,000 matches, sorting them, and then discarding everything but the first 20 needed to display the search results page. max_matches controls N in that "best N" amount.

      This parameter noticeably affects per-query RAM and CPU usage. Values of 1,000 to 10,000 are generally fine, but higher limits must be used with care. Recklessly raising max_matches to 1,000,000 means that searchd will have to allocate and initialize 1-million-entry matches buffer for every query. That will obviously increase per-query RAM usage, and in some cases can also noticeably impact performance.

    • 'max_query_time' - integer (max search time threshold, msec)

    • 'max_predicted_time' - integer (max predicted search time, see Section 12.4.43, “predicted_time_costs”)

    • 'ranker' - any of 'proximity_bm25', 'bm25', 'none', 'wordcount', 'proximity', 'matchany', 'fieldmask', 'sph04', 'expr', or 'export' (refer to Section 5.4, “Search results ranking” for more details on each ranker)

    • 'retry_count' - integer (distributed retries count)

    • 'retry_delay' - integer (distributed retry delay, msec)

    • 'reverse_scan' - 0 or 1, lets you control the order in which full-scan query processes the rows

    • 'sort_method' - 'pq' (priority queue, set by default) or 'kbuffer' (gives faster sorting for already pre-sorted data, e.g. index data sorted by id). The result set is in both cases the same; picking one option or the other may just improve (or worsen!) performance. This option was added in version 2.1.1-beta.

    • 'rand_seed' - lets you specify a specific integer seed value for an ORDER BY RAND() query, for example: ... OPTION rand_seed=1234. By default, a new and different seed value is autogenerated for every query.


    SELECT * FROM test WHERE MATCH('@title hello @body world')
    OPTION ranker=bm25, max_matches=3000,
        field_weights=(title=10, body=3), agent_query_timeout=10000

  • FACET clause. This Sphinx specific extension enables faceted search with subtree optimization. It is capable of returning multiple result sets with a single SQL statement, without the need for complicated multi-queries. FACET clauses should be written at the very end of SELECT statements with spaces between them.

    FACET {expr_list} [BY {expr_list}] [ORDER BY {expr | FACET()} {ASC | DESC}] [LIMIT [offset,] count]
    SELECT * FROM test FACET brand_id FACET categories;
    SELECT * FROM test FACET brand_name BY brand_id ORDER BY brand_name ASC FACET property;

    Working example:

    mysql> SELECT *, IN(brand_id,1,2,3,4) AS b FROM facetdemo WHERE MATCH('Product') AND b=1 LIMIT 0,10
    FACET brand_name, brand_id BY brand_id ORDER BY brand_id ASC
    FACET INTERVAL(price,200,400,600,800) ORDER BY FACET() ASC
    FACET categories ORDER BY FACET() ASC;
    | id   | price | brand_id | title             | brand_name  | property | categories | b    |
    |    1 |   668 |        3 | Product Four Six  | Brand Three | Three    | 11,12,13   |    1 |
    |    2 |   101 |        4 | Product Two Eight | Brand Four  | One      | 12,13,14   |    1 |
    |    8 |   750 |        3 | Product Ten Eight | Brand Three | Five     | 13         |    1 |
    |    9 |    49 |        1 | Product Ten Two   | Brand One   | Three    | 13,14,15   |    1 |
    |   13 |   613 |        1 | Product Six Two   | Brand One   | Eight    | 13         |    1 |
    |   20 |   985 |        2 | Product Two Six   | Brand Two   | Nine     | 10         |    1 |
    |   22 |   501 |        3 | Product Five Two  | Brand Three | Four     | 12,13,14   |    1 |
    |   23 |   765 |        1 | Product Six Seven | Brand One   | Nine     | 11,12      |    1 |
    |   28 |   992 |        1 | Product Six Eight | Brand One   | Two      | 12,13      |    1 |
    |   29 |   259 |        1 | Product Nine Ten  | Brand One   | Five     | 12,13,14   |    1 |
    | brand_name  | brand_id | count(*) |
    | Brand One   |        1 |     1012 |
    | Brand Two   |        2 |     1025 |
    | Brand Three |        3 |      994 |
    | Brand Four  |        4 |      973 |
    | property | count(*) |
    | One      |      427 |
    | Five     |      420 |
    | Seven    |      420 |
    | Two      |      418 |
    | Three    |      407 |
    | Six      |      401 |
    | Nine     |      396 |
    | Eight    |      387 |
    | Four     |      371 |
    | Ten      |      357 |
    | interval(price,200,400,600,800) | count(*) |
    |                               0 |      799 |
    |                               1 |      795 |
    |                               2 |      757 |
    |                               3 |      833 |
    |                               4 |      820 |
    | categories | count(*) |
    |         10 |      961 |
    |         11 |     1653 |
    |         12 |     1998 |
    |         13 |     2090 |
    |         14 |     1058 |
    |         15 |      347 |

8.2. SELECT @@system_variable syntax

SELECT @@system_variable [LIMIT [offset,] row_count]

Added in version 2.0.2-beta, this is currently a placeholder query that does nothing and reports success. That is in order to keep compatibility with frameworks and connectors that automatically execute this statement.

8.3. SHOW META syntax

SHOW META [ LIKE pattern ]

SHOW META shows additional meta-information about the latest query such as query time and keyword statistics. IO and CPU counters will only be available if searchd was started with --iostats and --cpustats switches respectively. Additional predicted_time, dist_predicted_time, [{local|dist}]_fetched_[{docs|hits|skips}] counters will only be available if searchd was configured with predicted time costs and query had predicted_time in OPTION clause.

mysql> SELECT * FROM test1 WHERE MATCH('test|one|two');
| id   | weight | group_id | date_added |
|    1 |   3563 |      456 | 1231721236 |
|    2 |   2563 |      123 | 1231721236 |
|    4 |   1480 |        2 | 1231721236 |
3 rows in set (0.01 sec)

mysql> SHOW META;
| Variable_name         | Value |
| total                 | 3     |
| total_found           | 3     |
| time                  | 0.005 |
| keyword[0]            | test  |
| docs[0]               | 3     |
| hits[0]               | 5     |
| keyword[1]            | one   |
| docs[1]               | 1     |
| hits[1]               | 2     |
| keyword[2]            | two   |
| docs[2]               | 1     |
| hits[2]               | 2     |
| cpu_time              | 0.350 |
| io_read_time          | 0.004 |
| io_read_ops           | 2     |
| io_read_kbytes        | 0.4   |
| io_write_time         | 0.000 |
| io_write_ops          | 0     |
| io_write_kbytes       | 0.0   |
| agents_cpu_time       | 0.000 |
| agent_io_read_time    | 0.000 |
| agent_io_read_ops     | 0     |
| agent_io_read_kbytes  | 0.0   |
| agent_io_write_time   | 0.000 |
| agent_io_write_ops    | 0     |
| agent_io_write_kbytes | 0.0   |
12 rows in set (0.00 sec)

Starting version 2.1.1-beta, you can also use the optional LIKE clause. It lets you pick just the variables that match a pattern. The pattern syntax is that of regular SQL wildcards, that is, '%' means any number of any characters, and '_' means a single character:

mysql> SHOW META LIKE 'total%';
| Variable_name         | Value |
| total                 | 3     |
| total_found           | 3     |
2 rows in set (0.00 sec)

8.4. SHOW WARNINGS syntax


SHOW WARNINGS statement, introduced in version 0.9.9-rc2, can be used to retrieve the warning produced by the latest query. The error message will be returned along with the query itself:

mysql> SELECT * FROM test1 WHERE MATCH('@@title hello') \G
ERROR 1064 (42000): index test1: syntax error, unexpected TOK_FIELDLIMIT
near '@title hello'

mysql> SELECT * FROM test1 WHERE MATCH('@title -hello') \G
ERROR 1064 (42000): index test1: query is non-computable (single NOT operator)

mysql> SELECT * FROM test1 WHERE MATCH('"test doc"/3') \G
*************************** 1. row ***************************
        id: 4
    weight: 2500
  group_id: 2
date_added: 1231721236
1 row in set, 1 warning (0.00 sec)

*************************** 1. row ***************************
  Level: warning
   Code: 1000
Message: quorum threshold too high (words=2, thresh=3); replacing quorum operator
         with AND operator
1 row in set (0.00 sec)

8.5. SHOW STATUS syntax

SHOW STATUS [ LIKE pattern ]

SHOW STATUS, introduced in version 0.9.9-rc2, displays a number of useful performance counters. IO and CPU counters will only be available if searchd was started with --iostats and --cpustats switches respectively.

| Counter            | Value |
| uptime             | 216   |
| connections        | 3     |
| maxed_out          | 0     |
| command_search     | 0     |
| command_excerpt    | 0     |
| command_update     | 0     |
| command_keywords   | 0     |
| command_persist    | 0     |
| command_status     | 0     |
| agent_connect      | 0     |
| agent_retry        | 0     |
| queries            | 10    |
| dist_queries       | 0     |
| query_wall         | 0.075 |
| query_cpu          | OFF   |
| dist_wall          | 0.000 |
| dist_local         | 0.000 |
| dist_wait          | 0.000 |
| query_reads        | OFF   |
| query_readkb       | OFF   |
| query_readtime     | OFF   |
| avg_query_wall     | 0.007 |
| avg_query_cpu      | OFF   |
| avg_dist_wall      | 0.000 |
| avg_dist_local     | 0.000 |
| avg_dist_wait      | 0.000 |
| avg_query_reads    | OFF   |
| avg_query_readkb   | OFF   |
| avg_query_readtime | OFF   |
29 rows in set (0.00 sec)

Starting from version 2.1.1-beta, an optional LIKE clause is supported. Refer to Section 8.3, “SHOW META syntax” for its syntax details.

8.6. INSERT and REPLACE syntax

{INSERT | REPLACE} INTO index [(column, ...)]
    VALUES (value, ...)
    [, (...)]

INSERT statement, introduced in version 1.10-beta, is only supported for RT indexes. It inserts new rows (documents) into an existing index, with the provided column values.

ID column must be present in all cases. Rows with duplicate IDs will not be overwritten by INSERT; use REPLACE to do that. REPLACE works exactly like INSERT, except that if an old row has the same ID as a new row, the old row is deleted before the new row is inserted.

index is the name of RT index into which the new row(s) should be inserted. The optional column names list lets you only explicitly specify values for some of the columns present in the index. All the other columns will be filled with their default values (0 for scalar types, empty string for text types).

Expressions are not currently supported in INSERT and values should be explicitly specified.

Multiple rows can be inserted using a single INSERT statement by providing several comma-separated, parentheses-enclosed lists of rows values.

8.7. REPLACE syntax

{INSERT | REPLACE} INTO index [(column, ...)]
	VALUES (value, ...)
	[, (...)]

REPLACE syntax is identical to INSERT syntax and is discussed in Section 8.6, “INSERT and REPLACE syntax”.

8.8. DELETE syntax

DELETE FROM index WHERE where_condition

DELETE statement, introduced in version 1.10-beta, is only supported for RT indexes and for distributed which contains only RT indexes as agents It deletes existing rows (documents) from an existing index based on ID.

index is the name of RT index from which the row should be deleted.

where_condition has the same syntax as in the SELECT statement (see Section 8.1, “SELECT syntax” for details).

mysql> select * from rt;
| id   | gid  | mva1        | mva2 |
|  100 | 1000 | 100,201     | 100  |
|  101 | 1001 | 101,202     | 101  |
|  102 | 1002 | 102,203     | 102  |
|  103 | 1003 | 103,204     | 103  |
|  104 | 1004 | 104,204,205 | 104  |
|  105 | 1005 | 105,206     | 105  |
|  106 | 1006 | 106,207     | 106  |
|  107 | 1007 | 107,208     | 107  |
8 rows in set (0.00 sec)

mysql> delete from rt where match ('dumy') and mva1>206;
Query OK, 2 rows affected (0.00 sec)

mysql> select * from rt;
| id   | gid  | mva1        | mva2 |
|  100 | 1000 | 100,201     | 100  |
|  101 | 1001 | 101,202     | 101  |
|  102 | 1002 | 102,203     | 102  |
|  103 | 1003 | 103,204     | 103  |
|  104 | 1004 | 104,204,205 | 104  |
|  105 | 1005 | 105,206     | 105  |
6 rows in set (0.00 sec)

mysql> delete from rt where id in (100,104,105);
Query OK, 3 rows affected (0.01 sec)

mysql> select * from rt;
| id   | gid  | mva1    | mva2 |
|  101 | 1001 | 101,202 | 101  |
|  102 | 1002 | 102,203 | 102  |
|  103 | 1003 | 103,204 | 103  |
3 rows in set (0.00 sec)

mysql> delete from rt where mva1 in (102,204);
Query OK, 2 rows affected (0.01 sec)

mysql> select * from rt;
| id   | gid  | mva1    | mva2 |
|  101 | 1001 | 101,202 | 101  |
1 row in set (0.00 sec)

8.9. SET syntax

SET [GLOBAL] server_variable_name = value
SET [INDEX index_name] GLOBAL @user_variable_name = (int_val1 [, int_val2, ...])
SET @@dummy_variable = ignored_value

SET statement, introduced in version 1.10-beta, modifies a variable value. The variable names are case-insensitive. No variable value changes survive server restart.

SET NAMES statement and SET @@variable_name syntax, both introduced in version 2.0.2-beta, do nothing. They were implemented to maintain compatibility with 3rd party MySQL client libraries, connectors, and frameworks that may need to run this statement when connecting.

There are the following classes of the variables:

  1. per-session server variable (1.10-beta and above)

  2. global server variable (2.0.1-beta and above)

  3. global user variable (2.0.1-beta and above)

  4. global distributed variable (2.2.3-beta and above)

Global user variables are shared between concurrent sessions. Currently, the only supported value type is the list of BIGINTs, and these variables can only be used along with IN() for filtering purpose. The intended usage scenario is uploading huge lists of values to searchd (once) and reusing them (many times) later, saving on network overheads. Starting with 2.2.3-beta, global user variables might be either transferred to all agents of distributed index or set locally in case of local index defined at distibuted index. Example:

// in session 1
mysql> SET GLOBAL @myfilter=(2,3,5,7,11,13);
Query OK, 0 rows affected (0.00 sec)

// later in session 2
mysql> SELECT * FROM test1 WHERE group_id IN @myfilter;
| id   | weight | group_id | date_added | title           | tag  |
|    3 |      1 |        2 | 1299338153 | another doc     | 15   |
|    4 |      1 |        2 | 1299338153 | doc number four | 7,40 |
2 rows in set (0.02 sec)

Per-session and global server variables affect certain server settings in the respective scope. Known per-session server variables are:

AUTOCOMMIT = {0 | 1}

Whether any data modification statement should be implicitly wrapped by BEGIN and COMMIT. Introduced in version 1.10-beta.


Selects the collation to be used for ORDER BY or GROUP BY on string values in the subsequent queries. Refer to Section 5.12, “Collations” for a list of known collation names. Introduced in version 2.0.1-beta.


Does nothing; a placeholder to support frameworks, clients, and connectors that attempt to automatically enforce a charset when connecting to a Sphinx server. Introduced in version 2.0.1-beta.


Does nothing; a placeholder to support frameworks, clients, and connectors that attempt to automatically enforce a charset when connecting to a Sphinx server. Introduced in version 2.0.2-beta.

SQL_MODE = value

Does nothing; a placeholder to support frameworks, clients, and connectors that attempt to automatically enforce a charset when connecting to a Sphinx server. Introduced in version 2.0.2-beta.

PROFILING = {0 | 1}

Enables query profiling in the current session. Defaults to 0. See also Section 8.30, “SHOW PROFILE syntax”. Introduced in version 2.1.1-beta.

Known global server variables are:

QUERY_LOG_FORMAT = {plain | sphinxql}

Changes the current log format. Introduced in version 2.0.1-beta.

LOG_LEVEL = {info | debug | debugv | debugvv}

Changes the current log verboseness level. Introduced in version 2.0.1-beta.


mysql> SET autocommit=0;
Query OK, 0 rows affected (0.00 sec)

mysql> SET GLOBAL query_log_format=sphinxql;
Query OK, 0 rows affected (0.00 sec)

8.10. SET TRANSACTION syntax


SET TRANSACTION statement, introduced in version 2.0.2-beta, does nothing. It was implemented to maintain compatibility with 3rd party MySQL client libraries, connectors, and frameworks that may need to run this statement when connecting.


Query OK, 0 rows affected (0.00 sec)

8.11. BEGIN, COMMIT, and ROLLBACK syntax


BEGIN, COMMIT, and ROLLBACK statements were introduced in version 1.10-beta. BEGIN statement (or its START TRANSACTION alias) forcibly commits pending transaction, if any, and begins a new one. COMMIT statement commits the current transaction, making all its changes permanent. ROLLBACK statement rolls back the current transaction, canceling all its changes. SET AUTOCOMMIT controls the autocommit mode in the active session.

AUTOCOMMIT is set to 1 by default, meaning that every statement that performs any changes on any index is implicitly wrapped in BEGIN and COMMIT.

Transactions are limited to a single RT index, and also limited in size. They are atomic, consistent, overly isolated, and durable. Overly isolated means that the changes are not only invisible to the concurrent transactions but even to the current session itself.

8.12. BEGIN syntax


BEGIN syntax is discussed in detail in Section 8.11, “BEGIN, COMMIT, and ROLLBACK syntax”.

8.13. ROLLBACK syntax


ROLLBACK syntax is discussed in detail in Section 8.11, “BEGIN, COMMIT, and ROLLBACK syntax”.

8.14. CALL SNIPPETS syntax

CALL SNIPPETS(data, index, query[, opt_value AS opt_name[, ...]])

CALL SNIPPETS statement, introduced in version 1.10-beta, builds a snippet from provided data and query, using specified index settings.

data is the source data to extract a snippet from. It could be a single string, or the list of the strings enclosed in curly brackets. index is the name of the index from which to take the text processing settings. query is the full-text query to build snippets for. Additional options are documented in Section 9.7.1, “BuildExcerpts”. Usage example:

CALL SNIPPETS('this is my document text', 'test1', 'hello world',
    5 AS around, 200 AS limit);
CALL SNIPPETS(('this is my document text','this is my another text'), 'test1', 'hello world',
    5 AS around, 200 AS limit);
CALL SNIPPETS(('data/doc1.txt','data/doc2.txt','/home/sphinx/doc3.txt'), 'test1', 'hello world',
    5 AS around, 200 AS limit, 1 AS load_files);

8.15. CALL KEYWORDS syntax

CALL KEYWORDS(text, index [, 1])

CALL KEYWORDS statement, introduced in version 1.10-beta, splits text into particular keywords. It returns tokenized and normalized forms of the keywords, and, optionally, keyword statistics. Since version 2.2.2-beta it also returns the position of each keyword in the query and all forms of tokenized keywords in the case that lemmatizers were used.

text is the text to break down to keywords. index is the name of the index from which to take the text processing settings. hits is an optional boolean parameter that specifies whether to return document and hit occurrence statistics.

8.16. SHOW TABLES syntax

SHOW TABLES [ LIKE pattern ]

SHOW TABLES statement, introduced in version 2.0.1-beta, enumerates all currently active indexes along with their types. As of 2.0.1-beta, existing index types are local, distributed, and rt respectively. Example:

| Index | Type        |
| dist1 | distributed |
| rt    | rt          |
| test1 | local       |
| test2 | local       |
4 rows in set (0.00 sec)

Starting from version 2.1.1-beta, an optional LIKE clause is supported. Refer to Section 8.3, “SHOW META syntax” for its syntax details.

mysql> SHOW TABLES LIKE '%4';
| Index | Type        |
| dist4 | distributed |
1 row in set (0.00 sec)

8.17. DESCRIBE syntax

{DESC | DESCRIBE} index [ LIKE pattern ]

DESCRIBE statement, introduced in version 2.0.1-beta, lists index columns and their associated types. Columns are document ID, full-text fields, and attributes. The order matches that in which fields and attributes are expected by INSERT and REPLACE statements. As of 2.0.1-beta, column types are field, integer, timestamp, ordinal, bool, float, bigint, string, and mva. ID column will be typed either integer or bigint based on whether the binaries were built with 32-bit or 64-bit document ID support. Example:

mysql> DESC rt;
| Field   | Type    |
| id      | integer |
| title   | field   |
| content | field   |
| gid     | integer |
4 rows in set (0.00 sec)

Starting from version 2.1.1-beta, an optional LIKE clause is supported. Refer to Section 8.3, “SHOW META syntax” for its syntax details.

8.18. CREATE FUNCTION syntax

    SONAME 'udf_lib_file'

CREATE FUNCTION statement, introduced in version 2.0.1-beta, installs a user-defined function (UDF) with the given name and type from the given library file. The library file must reside in a trusted plugin_dir directory. On success, the function is available for use in all subsequent queries that the server receives. Example:

mysql> CREATE FUNCTION avgmva RETURNS INTEGER SONAME 'udfexample.dll';
Query OK, 0 rows affected (0.03 sec)

mysql> SELECT *, AVGMVA(tag) AS q from test1;
| id   | weight | tag     | q         |
|    1 |      1 | 1,3,5,7 | 4.000000  |
|    2 |      1 | 2,4,6   | 4.000000  |
|    3 |      1 | 15      | 15.000000 |
|    4 |      1 | 7,40    | 23.500000 |

8.19. DROP FUNCTION syntax


DROP FUNCTION statement, introduced in version 2.0.1-beta, deinstalls a user-defined function (UDF) with the given name. On success, the function is no longer available for use in subsequent queries. Pending concurrent queries will not be affected and the library unload, if necessary, will be postponed until those queries complete. Example:

mysql> DROP FUNCTION avgmva;
Query OK, 0 rows affected (0.00 sec)

8.20. SHOW VARIABLES syntax

SHOW [{GLOBAL | SESSION}] VARIABLES [WHERE variable_name='xxx']

SHOW VARIABLES statement was added in version 2.0.1-beta to improve compatibility with 3rd party MySQL connectors and frameworks that automatically execute this statement. The WHERE option was added in version 2.1.1-beta.

In version 2.0.1-beta, it did nothing.

Starting from version 2.0.2-beta, it returns the current values of a few server-wide variables. Also, support for GLOBAL and SESSION clauses was added.

| Variable_name        | Value    |
| autocommit           | 1        |
| collation_connection | libc_ci  |
| query_log_format     | sphinxql |
| log_level            | info     |
4 rows in set (0.00 sec)

Starting from 2.1.1-beta, support for WHERE variable_name clause was added, to help certain connectors.

8.21. SHOW COLLATION syntax


Added in version 2.0.1-beta, this is currently a placeholder query that does nothing and reports success. That is in order to keep compatibility with frameworks and connectors that automatically execute this statement.

Query OK, 0 rows affected (0.00 sec)



Added in version 2.1.1-beta, this is currently a placeholder query that does nothing and reports that a UTF-8 character set is available. It was added in order to keep compatibility with frameworks and connectors that automatically execute this statement.

| Charset | Description   | Default collation | Maxlen |
| utf8    | UTF-8 Unicode | utf8_general_ci   | 3      |
1 row in set (0.00 sec)

8.23. UPDATE syntax

UPDATE index SET col1 = newval1 [, ...] WHERE where_condition [OPTION opt_name = opt_value [, ...]]

UPDATE statement was added in version 2.0.1-beta. Multiple attributes and values can be specified in a single statement. Both RT and disk indexes are supported.

As of version 2.0.2-beta, all attributes types (int, bigint, float, MVA), except for strings and JSON attributes, can be dynamically updated. Previously, some of these types were not supported.

where_condition (also added in 2.0.2-beta) has the same syntax as in the SELECT statement (see Section 8.1, “SELECT syntax” for details).

When assigning the out-of-range values to 32-bit attributes, they will be trimmed to their lower 32 bits without a prompt. For example, if you try to update the 32-bit unsigned int with a value of 4294967297, the value of 1 will actually be stored, because the lower 32 bits of 4294967297 (0x100000001 in hex) amount to 1 (0x00000001 in hex).

MVA values sets for updating (and also for INSERT or REPLACE, refer to Section 8.6, “INSERT and REPLACE syntax”) must be specified as comma-separated lists in parentheses. To erase the MVA value, just assign () to it.

Starting from 2.2.1-beta version UPDATE can be used to update integer and float values in JSON array. No strings, arrays and other types yet.

mysql> UPDATE myindex SET enabled=0 WHERE id=123;
Query OK, 1 rows affected (0.00 sec)

mysql> UPDATE myindex
  SET bigattr=-100000000000,
  WHERE MATCH('hehe') AND enabled=1;
Query OK, 148 rows affected (0.01 sec)

OPTION clause. This is a Sphinx specific extension that lets you control a number of per-update options. The syntax is:

OPTION <optionname>=<value> [ , ... ]

The list of allowed options are the same as for SELECT statement. Specifically for UPDATE statement you can use these options:

  • 'ignore_nonexistent_columns' - this option, added in version 2.1.1-beta, points that the update will silently ignore any warnings about trying to update a column which is not exists in current index schema.

    'strict' - this option is used while updating JSON attributes. As of 2.2.1-beta, it's possible to update just some types in JSON. And if you try to update, for example, array type you'll get error with 'strict' option on and warning otherwise.

8.24. ALTER syntax


The ALTER statement was added in version 2.2.1-beta. As of 2.2.1-beta, it supports adding one attribute at a time for both plain and RT indexes. The int, bigint, float, bool, multi-valued, multi-valued 64bit, json and string attribute types are supported. Support for multi, multi64, json and string attributes was added in 2.2.2-beta. As of 2.2.2-beta, you can add json and string attributes, but you cannot modify their values. The ability to remove attributes was added in 2.2.2-beta.

Implementation details. As of 2.2.1-beta, the querying of an index is impossible (because of a write lock) while adding a column. This may change in the future. The newly created attribute values are set to 0. ALTER will not work for distributed indexes and indexes without any attributes. DROP COLUMN will fail if an index has only one attribute.


As of 2.2.3-beta, ALTER can also reconfigure an existing RT index, so that new tokenization, morphology, and other text processing settings from sphinx.conf take effect on the newly INSERT-ed rows, while retaining the existing rows as they were. Internally, it forcibly saves the current RAM chunk as a new disk chunk, and adjusts the index header, so that the new rows are tokenized using the new rules. Note that as the queries are currently parsed separately for every disk chunk, this might result in warnings regarding the keyword sets mismatch.

mysql> desc plain;
| Field      | Type      |
| id         | bigint    |
| text       | field     |
| group_id   | uint      |
| date_added | timestamp |
4 rows in set (0.01 sec)

mysql> alter table plain add column test integer;
Query OK, 0 rows affected (0.04 sec)

mysql> desc plain;
| Field      | Type      |
| id         | bigint    |
| text       | field     |
| group_id   | uint      |
| date_added | timestamp |
| test       | uint      |
5 rows in set (0.00 sec)

mysql> alter table plain drop column group_id;
Query OK, 0 rows affected (0.01 sec)

mysql> desc plain;
| Field      | Type      |
| id         | bigint    |
| text       | field     |
| date_added | timestamp |
| test       | uint      |
4 rows in set (0.00 sec)

8.25. ATTACH INDEX syntax

ATTACH INDEX diskindex TO RTINDEX rtindex

ATTACH INDEX statement, added in version 2.0.2-beta, lets you move data from a regular disk index to a RT index.

After a successful ATTACH, the data originally stored in the source disk index becomes a part of the target RT index, and the source disk index becomes unavailable (until the next rebuild). ATTACH does not result in any index data changes. Basically, it just renames the files (making the source index a new disk chunk of the target RT index), and updates the metadata. So it is a generally quick operation which might (frequently) complete as fast as under a second.

Note that when an index is attached to an empty RT index, the fields, attributes, and text processing settings (tokenizer, wordforms, etc) from the source index are copied over and take effect. The respective parts of the RT index definition from the configuration file will be ignored.

As of 2.0.2-beta, ATTACH INDEX comes with a number of restrictions. Most notably, the target RT index is currently required to be empty, making ATTACH INDEX a one-time conversion operation only. Those restrictions may be lifted in future releases, as we add the needed functionality to the RT indexes. The complete list is as follows.

  • Target RT index needs to be empty. (See Section 8.28, “TRUNCATE RTINDEX syntax”)

  • Source disk index needs to have index_sp=0, boundary_step=0, stopword_step=1.

  • Source disk index needs to have an empty index_zones setting.

mysql> DESC rt;
| Field     | Type    |
| id        | integer |
| testfield | field   |
| testattr  | uint    |
3 rows in set (0.00 sec)

mysql> SELECT * FROM rt;
Empty set (0.00 sec)

mysql> SELECT * FROM disk WHERE MATCH('test');
| id   | weight | group_id | date_added |
|    1 |   1304 |        1 | 1313643256 |
|    2 |   1304 |        1 | 1313643256 |
|    3 |   1304 |        1 | 1313643256 |
|    4 |   1304 |        1 | 1313643256 |
4 rows in set (0.00 sec)

Query OK, 0 rows affected (0.00 sec)

mysql> DESC rt;
| Field      | Type      |
| id         | integer   |
| title      | field     |
| content    | field     |
| group_id   | uint      |
| date_added | timestamp |
5 rows in set (0.00 sec)

mysql> SELECT * FROM rt WHERE MATCH('test');
| id   | weight | group_id | date_added |
|    1 |   1304 |        1 | 1313643256 |
|    2 |   1304 |        1 | 1313643256 |
|    3 |   1304 |        1 | 1313643256 |
|    4 |   1304 |        1 | 1313643256 |
4 rows in set (0.00 sec)

mysql> SELECT * FROM disk WHERE MATCH('test');
ERROR 1064 (42000): no enabled local indexes to search

8.26. FLUSH RTINDEX syntax


FLUSH RTINDEX statement, added in version 2.0.2-beta, forcibly flushes RT index RAM chunk contents to disk.

Backing up a RT index is as simple as copying over its data files, followed by the binary log. However, recovering from that backup means that all the transactions in the log since the last successful RAM chunk write would need to be replayed. Those writes normally happen either on a clean shutdown, or periodically with a (big enough!) interval between writes specified in rt_flush_period directive. So such a backup made at an arbitrary point in time just might end up with way too much binary log data to replay.

FLUSH RTINDEX forcibly writes the RAM chunk contents to disk, and also causes the subsequent cleanup of (now-redundant) binary log files. Thus, recovering from a backup made just after FLUSH RTINDEX should be almost instant.

mysql> FLUSH RTINDEX rt;
Query OK, 0 rows affected (0.05 sec)

8.27. FLUSH RAMCHUNK syntax


FLUSH RAMCHUNK statement, added in version 2.1.2-release, forcibly creates a new disk chunk in a RT index.

Normally, RT index would flush and convert the contents of the RAM chunk into a new disk chunk automatically, once the RAM chunk reaches the maximum allowed rt_mem_limit size. However, for debugging and testing it might be useful to forcibly create a new disk chunk, and FLUSH RAMCHUNK statement does exactly that.

Note that using FLUSH RAMCHUNK increases RT index fragmentation. Most likely, you want to use FLUSH RTINDEX instead. We suggest that you abstain from using just this statement unless you're absolutely sure what you're doing. As the right way is to issue FLUSH RAMCHUNK with following OPTIMIZE command. Such combo allows to keep RT index fragmentation on minimum.

Query OK, 0 rows affected (0.05 sec)



TRUNCATE RTINDEX statement, added in version 2.1.1-beta, clears the RT index completely. It disposes the in-memory data, unlinks all the index data files, and releases the associated binary logs.

Query OK, 0 rows affected (0.05 sec)

You may want to use this if you are using RT indices as "delta index" files; when you build the main index, you need to wipe the delta index, and thus TRUNCATE RTINDEX. You also need to use this command before attaching an index; see Section 8.25, “ATTACH INDEX syntax”.


SHOW AGENT ['agent'|'index'|index] STATUS [ LIKE pattern ]

Displays the statistic of remote agents or distributed index. It includes the values like the age of the last request, last answer, the number of different kind of errors and successes, etc. The statistic is shown for every agent for last 1, 5 and 15 intervals, each of them of ha_period_karma seconds. The command exists only in sphinxql.

| Variable_name                      | Value                      |
| status_period_seconds              | 60                         |
| status_stored_periods              | 15                         |
| ag_0_hostname                      |         |
| ag_0_references                    | 2                          |
| ag_0_lastquery                     | 0.41                       |
| ag_0_lastanswer                    | 0.19                       |
| ag_0_lastperiodmsec                | 222                        |
| ag_0_errorsarow                    | 0                          |
| ag_0_1periods_query_timeouts       | 0                          |
| ag_0_1periods_connect_timeouts     | 0                          |
| ag_0_1periods_connect_failures     | 0                          |
| ag_0_1periods_network_errors       | 0                          |
| ag_0_1periods_wrong_replies        | 0                          |
| ag_0_1periods_unexpected_closings  | 0                          |
| ag_0_1periods_warnings             | 0                          |
| ag_0_1periods_succeeded_queries    | 27                         |
| ag_0_1periods_msecsperquery        | 232.31                     |
| ag_0_5periods_query_timeouts       | 0                          |
| ag_0_5periods_connect_timeouts     | 0                          |
| ag_0_5periods_connect_failures     | 0                          |
| ag_0_5periods_network_errors       | 0                          |
| ag_0_5periods_wrong_replies        | 0                          |
| ag_0_5periods_unexpected_closings  | 0                          |
| ag_0_5periods_warnings             | 0                          |
| ag_0_5periods_succeeded_queries    | 146                        |
| ag_0_5periods_msecsperquery        | 231.83                     |
| ag_1_hostname                      |         |
| ag_1_references                    | 2                          |
| ag_1_lastquery                     | 0.41                       |
| ag_1_lastanswer                    | 0.19                       |
| ag_1_lastperiodmsec                | 220                        |
| ag_1_errorsarow                    | 0                          |
| ag_1_1periods_query_timeouts       | 0                          |
| ag_1_1periods_connect_timeouts     | 0                          |
| ag_1_1periods_connect_failures     | 0                          |
| ag_1_1periods_network_errors       | 0                          |
| ag_1_1periods_wrong_replies        | 0                          |
| ag_1_1periods_unexpected_closings  | 0                          |
| ag_1_1periods_warnings             | 0                          |
| ag_1_1periods_succeeded_queries    | 27                         |
| ag_1_1periods_msecsperquery        | 231.24                     |
| ag_1_5periods_query_timeouts       | 0                          |
| ag_1_5periods_connect_timeouts     | 0                          |
| ag_1_5periods_connect_failures     | 0                          |
| ag_1_5periods_network_errors       | 0                          |
| ag_1_5periods_wrong_replies        | 0                          |
| ag_1_5periods_unexpected_closings  | 0                          |
| ag_1_5periods_warnings             | 0                          |
| ag_1_5periods_succeeded_queries    | 146                        |
| ag_1_5periods_msecsperquery        | 230.85                     |
50 rows in set (0.01 sec)

Starting from version 2.1.1-beta, an optional LIKE clause is supported. Refer to Section 8.3, “SHOW META syntax” for its syntax details.

mysql> SHOW AGENT STATUS LIKE '%5period%msec%';
| Key                         | Value  |
| ag_0_5periods_msecsperquery | 234.72 |
| ag_1_5periods_msecsperquery | 233.73 |
| ag_2_5periods_msecsperquery | 343.81 |
3 rows in set (0.00 sec)

You can specify a particular agent by its address. In this case only that agent's data will be displayed. Also, 'agent_' prefix will be used instead of 'ag_N_':

mysql> SHOW AGENT '' STATUS LIKE '%15periods%';
| Variable_name                       | Value  |
| agent_15periods_query_timeouts      | 0      |
| agent_15periods_connect_timeouts    | 0      |
| agent_15periods_connect_failures    | 0      |
| agent_15periods_network_errors      | 0      |
| agent_15periods_wrong_replies       | 0      |
| agent_15periods_unexpected_closings | 0      |
| agent_15periods_warnings            | 0      |
| agent_15periods_succeeded_queries   | 439    |
| agent_15periods_msecsperquery       | 231.73 |
9 rows in set (0.00 sec)

Finally, you can check the status of the agents in a specific distributed index. It can be done with a SHOW AGENT index STATUS statement. That statement shows the index HA status (ie. whether or not it uses agent mirrors at all), and then the mirror information (specifically: address, blackhole and persistent flags, and the mirror selection probability used when one of the weighted-probability strategies is in effect).

mysql> SHOW AGENT dist_index STATUS;
| Variable_name                        | Value                          |
| dstindex_1_is_ha                     | 1                              |
| dstindex_1mirror1_id                 |         |
| dstindex_1mirror1_probability_weight | 0.372864                       |
| dstindex_1mirror1_is_blackhole       | 0                              |
| dstindex_1mirror1_is_persistent      | 0                              |
| dstindex_1mirror2_id                 |         |
| dstindex_1mirror2_probability_weight | 0.374635                       |
| dstindex_1mirror2_is_blackhole       | 0                              |
| dstindex_1mirror2_is_persistent      | 0                              |
| dstindex_1mirror3_id                 | dev1.sphinxsearch.com:6714:loc |
| dstindex_1mirror3_probability_weight | 0.252501                       |
| dstindex_1mirror3_is_blackhole       | 0                              |
| dstindex_1mirror3_is_persistent      | 0                              |
13 rows in set (0.00 sec)

8.30. SHOW PROFILE syntax


SHOW PROFILE statement, added in version 2.1.1-beta, shows a detailed execution profile of the previous SQL statement executed in the current SphinxQL session. Also, profiling must be enabled in the current session before running the statement to be instrumented. That can be done with a SET profiling=1 statement. By default, profiling is disabled to avoid potential performance implications, and therefore the profile will be empty.

Here's a complete instrumentation example:

mysql> SET profiling=1;
Query OK, 0 rows affected (0.00 sec)

mysql> SELECT id FROM lj WHERE MATCH('the test') LIMIT 1;
| id     |
| 946418 |
1 row in set (0.05 sec)

| Status       | Duration | Switches |
| unknown      | 0.000610 | 6        |
| net_read     | 0.000007 | 1        |
| dist_connect | 0.000036 | 1        |
| sql_parse    | 0.000048 | 1        |
| dict_setup   | 0.000001 | 1        |
| parse        | 0.000023 | 1        |
| transforms   | 0.000002 | 1        |
| init         | 0.000401 | 3        |
| open         | 0.000104 | 1        |
| read_docs    | 0.001570 | 71       |
| read_hits    | 0.003936 | 222      |
| get_docs     | 0.029837 | 1347     |
| get_hits     | 0.000548 | 1433     |
| filter       | 0.000619 | 1274     |
| rank         | 0.009892 | 2909     |
| sort         | 0.001562 | 52       |
| finalize     | 0.000250 | 1        |
| dist_wait    | 0.000000 | 1        |
| aggregate    | 0.000145 | 1        |
| net_write    | 0.000031 | 1        |
20 rows in set (0.00 sec)

Status column briefly describes where exactly (in which state) was the time spent. Duration column shows the wall clock time, in seconds. Switches column displays the number of times query engine changed to the given state. Those are just logical engine state switches and not any OS level context switches nor function calls (even though some of the sections can actually map to function calls) and they do not have any direct effect on the performance. In a sense, number of switches is just a number of times when the respective instrumentation point was hit.

States in the profile are returned in a prerecorded order that roughly maps (but is not identical) to the actual query order.

A list of states may (and will) vary over time, as we refine the states. Here's a brief description of the currently profiled states.

  • unknown, generic catch-all state. Accounts for both not-yet-instrumented code, or just small miscellaneous tasks that do not really belong in any other state, but are too small to deserve their own state.
  • net_read, reading the query from the network (that is, the application).
  • io, generic file IO time.
  • dist_connect, connecting to remote agents in the distributed index case.
  • sql_parse, parsing the SphinxQL syntax.
  • dict_setup, dictionary and tokenizer setup.
  • parse, parsing the full-text query syntax.
  • transforms, full-text query transformations (wildcard and other expansions, simplification, etc).
  • init, initializing the query evaluation.
  • open, opening the index files.
  • read_docs, IO time spent reading document lists.
  • read_hits, IO time spent reading keyword positions.
  • get_docs, computing the matching documents.
  • get_hits, computing the matching positions.
  • filter, filtering the full-text matches.
  • rank, computing the relevance rank.
  • sort, sorting the matches.
  • finalize, finalizing the per-index search result set (last stage expressions, etc).
  • dist_wait, waiting for the remote results from the agents in the distributed index case.
  • aggregate, aggregating multiple result sets.
  • net_write, writing the result set to the network.

8.31. SHOW INDEX STATUS syntax


Added in version 2.1.1-beta. Displays various per-index statistics. Currently, those include:

  • indexed_documents and indexed_bytes, number of the documents indexed and their text size in bytes, respectively.
  • field_tokens_XXX, sums of per-field lengths (in tokens) over the entire index (that is used internally in BM25A and BM25F functions for ranking purposes). Only available for indexes built with index_field_lengths=1.
  • ram_bytes, total size (in bytes) of the RAM-resident index portion.

| Variable_name      | Value       |
| index_type         | disk        |
| indexed_documents  | 2495219     |
| indexed_bytes      | 10380483879 |
| field_tokens_title | 6999145     |
| field_tokens_body  | 1501825050  |
| total_tokens       | 1508824195  |
| ram_bytes          | 305963599   |
| disk_bytes         | 5455804365  |
| mem_limit          | 536870912   |
8 rows in set (0.00 sec)



Displays per-index settings in a sphinx.conf compliant file format, similar to the --dumpconfig option of the indextool. The report provides a breakdown of all the index settings, including tokenizer and dictionary options. You may also specify a particular chunk number for the RT indexes.

8.33. OPTIMIZE INDEX syntax


Available since version 2.1.1-beta, OPTIMIZE statement enqueues a RT index for optimization in a background thread.

Over time, RT indexes can grow fragmented into many disk chunks and/or tainted with deleted, but unpurged data, impacting search performance. When that happens, they can be optimized. Basically, the optimization pass merges together disk chunks pairs, purging off documents suppressed by K-list as it goes.

That is a lengthy and IO intensive process, so to limit the impact, all the actual merge work is executed serially in a special background thread, and the OPTIMIZE statement simply adds a job to its queue. Currently, there is no way to check the index or queue status (that might be added in the future to the SHOW INDEX STATUS and SHOW STATUS statements respectively). The optimization thread can be IO-throttled, you can control the maximum number of IOs per second and the maximum IO size with rt_merge_iops and rt_merge_maxiosize directives respectively. The optimization jobs queue is lost on daemon crash.

The RT index being optimized stays online and available for both searching and updates at (almost) all times during the optimization. It gets locked (very) briefly every time that a pair of disk chunks is merged successfully, to rename the old and the new files, and update the index header.

At the moment, OPTIMIZE needs to be issued manually, the indexes will not be optimized automatically. That might change in the future releases.

Query OK, 0 rows affected (0.00 sec)

8.34. SHOW PLAN syntax


SHOW PLAN statement, added in 2.1.2-release, displays the execution plan of the previous SELECT statement. The plan gets generated and stored during the actual execution, so profiling must be enabled in the current session before running that statement. That can be done with a SET profiling=1 statement.

Here's a complete instrumentation example:

mysql> SET profiling=1 \G
Query OK, 0 rows affected (0.00 sec)

mysql> SELECT id FROM lj WHERE MATCH('the i') LIMIT 1 \G
*************************** 1. row ***************************
id: 39815
1 row in set (1.53 sec)

mysql> SHOW PLAN \G
*************************** 1. row ***************************
Variable: transformed_tree
   Value: AND(
  AND(KEYWORD(the, querypos=1)),
  AND(KEYWORD(i, querypos=2)))
1 row in set (0.00 sec)

And here's a less trivial example that shows how the actually evaluated query tree can be rather different from the original one because of expansions and other transformations:

mysql> SELECT * FROM test WHERE MATCH('@title abc* @body hey') \G SHOW PLAN \G
*************************** 1. row ***************************
Variable: transformed_tree
   Value: AND(
  OR(fields=(title), KEYWORD(abcx, querypos=1, expanded), KEYWORD(abcm, querypos=1, expanded)),
  AND(fields=(body), KEYWORD(hey, querypos=2)))
1 row in set (0.00 sec)

8.35. SHOW DATABASES syntax


Added in 2.2.1-beta. This is a dummy statement to support MySQL Workbench and other clients that require it. Currently, it does absolutely nothing.

8.36. CREATE PLUGIN syntax

CREATE PLUGIN plugin_name TYPE 'plugin_type' SONAME 'plugin_library'

Added in 2.2.2-beta. Loads the given library (if it is not loaded yet) and loads the specified plugin from it. As of 2.2.2-beta, the known plugin types are:

  • ranker

  • index_token_filter

  • query_token_filter

Refer to Section 6.2, “Sphinx plugins” for more information regarding writing the plugins.

mysql> CREATE PLUGIN myranker TYPE 'ranker' SONAME 'myplugins.so';
Query OK, 0 rows affected (0.00 sec)

8.37. DROP PLUGIN syntax

DROP PLUGIN plugin_name TYPE 'plugin_type'

Added in 2.2.2-beta. Markes the specified plugin for unloading. The unloading is not immediate, because the concurrent queries might be using it. However, after a DROP new queries will not be able to use it. Then, once all the currently executing queries using it are completed, the plugin will be unloaded. Once all the plugins from the given library are unloaded, the library is also automatically unloaded.

mysql> DROP PLUGIN myranker TYPE 'ranker';
Query OK, 0 rows affected (0.00 sec)

8.38. SHOW PLUGINS syntax


Added in 2.2.2-beta. Displays all the loaded plugins and UDFs. "Type" column should be one of the udf, ranker, index_token_filter, or query_token_filter. "Users" column is the number of thread that are currently using that plugin in a query. "Extra" column is intended for various additional plugin-type specific information; currently, it shows the return type for the UDFs and is empty for all the other plugin types.

| Type | Name     | Library        | Users | Extra |
| udf  | sequence | udfexample.dll | 0     | INT   |
1 row in set (0.00 sec)

8.39. SHOW THREADS syntax

SHOW THREADS [ OPTION columns=width ]

SHOW THREADS statement, introduced in version 2.2.2-beta, lists all currently active client threads, not counting system threads. It returns a table with columns that describe:

  • thread id
  • connection protocol, possible values are sphinxapi and sphinxql
  • thread state, possible values are handshake, net_read, net_write, query, net_idle
  • time since the current state was changed (in seconds, with microsecond precision)
  • information about queries

The 'Info' column will be cut at the width you've specified in the 'columns=width' option (notice the third row in the example table below). This column will contain raw SphinxQL queries and, if there are API queries, full text syntax and comments will be displayed. With an API-snippet, the data size will be displayed along with the query.

mysql> SHOW THREADS OPTION columns=50;
| Tid  | Proto    | State | Time     | Info                                               |
| 5168 | sphinxql | query | 0.000002 | show threads option columns=50                     |
| 5175 | sphinxql | query | 0.000002 | select * from rt where match ( 'the box' )         |
| 1168 | sphinxql | query | 0.000002 | select * from rt where match ( 'the box and faximi |
3 row in set (0.00 sec)

8.40. Multi-statement queries

Starting version 2.0.1-beta, SphinxQL supports multi-statement queries, or batches. Possible inter-statement optimizations described in Section 5.11, “Multi-queries” do apply to SphinxQL just as well. The batched queries should be separated by a semicolon. Your MySQL client library needs to support MySQL multi-query mechanism and multiple result set. For instance, mysqli interface in PHP and DBI/DBD libraries in Perl are known to work.

Here's a PHP sample showing how to utilize mysqli interface with Sphinx.


$link = mysqli_connect ( "", "root", "", "", 9306 );
if ( mysqli_connect_errno() )
    die ( "connect failed: " . mysqli_connect_error() );

$batch = "SELECT * FROM test1 ORDER BY group_id ASC;";
$batch .= "SELECT * FROM test1 ORDER BY group_id DESC";

if ( !mysqli_multi_query ( $link, $batch ) )
    die ( "query failed" );

    // fetch and print result set
    if ( $result = mysqli_store_result($link) )
        while ( $row = mysqli_fetch_row($result) )
            printf ( "id=%s\n", $row[0] );

    // print divider
    if ( mysqli_more_results($link) )
        printf ( "------\n" );

} while ( mysqli_next_result($link) );

Its output with the sample test1 index included with Sphinx is as follows.

$ php test_multi.php

The following statements can currently be used in a batch: SELECT, SHOW WARNINGS, SHOW STATUS, and SHOW META. Arbitrary sequence of these statements are allowed. The results sets returned should match those that would be returned if the batched queries were sent one by one.

8.41. Comment syntax

Since version 2.0.1-beta, SphinxQL supports C-style comment syntax. Everything from an opening /* sequence to a closing */ sequence is ignored. Comments can span multiple lines, can not nest, and should not get logged. MySQL specific /*! ... */ comments are also currently ignored. (As the comments support was rather added for better compatibility with mysqldump produced dumps, rather than improving general query interoperability between Sphinx and MySQL.)


8.42. List of SphinxQL reserved keywords

A complete alphabetical list of keywords that are currently reserved in SphinxQL syntax (and therefore can not be used as identifiers).


8.43. SphinxQL upgrade notes, version 2.0.1-beta

This section only applies to existing applications that use SphinxQL versions prior to 2.0.1-beta.

In previous versions, SphinxQL just wrapped around SphinxAPI and inherited its magic columns and column set quirks. Essentially, SphinxQL queries could return (slightly) different columns and in a (slightly) different order than it was explicitly requested in the query. Namely, weight magic column (which is not a real column in any index) was added at all times, and GROUP BY related @count, @group, and @distinct magic columns were conditionally added when grouping. Also, the order of columns (attributes) in the result set was actually taken from the index rather than the query. (So if you asked for columns C, B, A in your query but they were in the A, B, C order in the index, they would have been returned in the A, B, C order.)

In version 2.0.1-beta, we fixed that. SphinxQL is now more SQL compliant (and will be further brought in as much compliance with standard SQL syntax as possible).

The important changes are as follows:

  • @ID magic name is deprecated in favor of ID. Document ID is considered an attribute.

  • WEIGHT is no longer implicitly returned, because it is not actually a column (an index attribute), but rather an internal function computed per each row (a match). You have to explicitly ask for it, using the WEIGHT() function. (The requirement to alias the result will be lifted in the next release.)

    SELECT id, WEIGHT() w FROM myindex WHERE MATCH('test')

  • You can now use quoted reserved keywords as aliases. The quote character is backtick ("`", ASCII code 96 decimal, 60 hex). One particularly useful example would be returning weight column like the old mode:

    SELECT id, WEIGHT() `weight` FROM myindex WHERE MATCH('test')

  • The column order is now different and should now match the one explicitly defined in the query. So if you are accessing columns based on their position in the result set rather than the name (for instance, by using mysql_fetch_row() rather than mysql_fetch_assoc() in PHP), check and fix the order of columns in your queries.

  • SELECT * return the columns in index order, as it used to, including the ID column. However, SELECT * does not automatically return WEIGHT(). To update such queries in case you access columns by names, simply add it to the query:

    SELECT *, WEIGHT() `weight` FROM myindex WHERE MATCH('test')

    Otherwise, i.e., in case you rely on column order, select ID, weight, and then other columns:

    SELECT id, *, WEIGHT() `weight` FROM myindex WHERE MATCH('test')

  • Magic @count and @distinct attributes are no longer implicitly returned. You now have to explicitly ask for them when using GROUP BY. (Also note that you currently have to alias them; that requirement will be lifted in the future.)

    SELECT gid, COUNT(*) q FROM myindex WHERE MATCH('test')

Chapter 9. API reference

There is a number of native searchd client API implementations for Sphinx. As of time of this writing, we officially support our own PHP, Python, and Java implementations. There also are third party free, open-source API implementations for Perl, Ruby, and C++.

The reference API implementation is in PHP, because (we believe) Sphinx is most widely used with PHP than any other language. This reference documentation is in turn based on reference PHP API, and all code samples in this section will be given in PHP.

However, all other APIs provide the same methods and implement the very same network protocol. Therefore the documentation does apply to them as well. There might be minor differences as to the method naming conventions or specific data structures used. But the provided functionality must not differ across languages.

9.1. General API functions

9.1.1. GetLastError

Prototype: function GetLastError()

Returns last error message, as a string, in human readable format. If there were no errors during the previous API call, empty string is returned.

You should call it when any other function (such as Query()) fails (typically, the failing function returns false). The returned string will contain the error description.

The error message is not reset by this call; so you can safely call it several times if needed.

9.1.2. GetLastWarning

Prototype: function GetLastWarning ()

Returns last warning message, as a string, in human readable format. If there were no warnings during the previous API call, empty string is returned.

You should call it to verify whether your request (such as Query()) was completed but with warnings. For instance, search query against a distributed index might complete successfully even if several remote agents timed out. In that case, a warning message would be produced.

The warning message is not reset by this call; so you can safely call it several times if needed.

9.1.3. SetServer

Prototype: function SetServer ( $host, $port )

Sets searchd host name and TCP port. All subsequent requests will use the new host and port settings. Default host and port are 'localhost' and 9312, respectively.

9.1.4. SetRetries

Prototype: function SetRetries ( $count, $delay=0 )

Sets distributed retry count and delay.

On temporary failures searchd will attempt up to $count retries per agent. $delay is the delay between the retries, in milliseconds. Retries are disabled by default. Note that this call will not make the API itself retry on temporary failure; it only tells searchd to do so. Currently, the list of temporary failures includes all kinds of connect() failures and maxed out (too busy) remote agents.

9.1.5. SetConnectTimeout

Prototype: function SetConnectTimeout ( $timeout )

Sets the time allowed to spend connecting to the server before giving up.

Under some circumstances, the server can be delayed in responding, either due to network delays, or a query backlog. In either instance, this allows the client application programmer some degree of control over how their program interacts with searchd when not available, and can ensure that the client application does not fail due to exceeding the script execution limits (especially in PHP).

In the event of a failure to connect, an appropriate error code should be returned back to the application in order for application-level error handling to advise the user.

9.1.6. SetArrayResult

Prototype: function SetArrayResult ( $arrayresult )

PHP specific. Controls matches format in the search results set (whether matches should be returned as an array or a hash).

$arrayresult argument must be boolean. If $arrayresult is false (the default mode), matches will returned in PHP hash format with document IDs as keys, and other information (weight, attributes) as values. If $arrayresult is true, matches will be returned as a plain array with complete per-match information including document ID.

Introduced along with GROUP BY support on MVA attributes. Group-by-MVA result sets may contain duplicate document IDs. Thus they need to be returned as plain arrays, because hashes will only keep one entry per document ID.

9.1.7. IsConnectError

Prototype: function IsConnectError ()

Checks whether the last error was a network error on API side, or a remote error reported by searchd. Returns true if the last connection attempt to searchd failed on API side, false otherwise (if the error was remote, or there were no connection attempts at all). Introduced in version 0.9.9-rc1.

9.2. General query settings

9.2.1. SetLimits

Prototype: function SetLimits ( $offset, $limit, $max_matches=1000, $cutoff=0 )

Sets offset into server-side result set ($offset) and amount of matches to return to client starting from that offset ($limit). Can additionally control maximum server-side result set size for current query ($max_matches) and the threshold amount of matches to stop searching at ($cutoff). All parameters must be non-negative integers.

First two parameters to SetLimits() are identical in behavior to MySQL LIMIT clause. They instruct searchd to return at most $limit matches starting from match number $offset. The default offset and limit settings are 0 and 20, that is, to return first 20 matches.

max_matches setting controls how much matches searchd will keep in RAM while searching. All matching documents will be normally processed, ranked, filtered, and sorted even if max_matches is set to 1. But only best N documents are stored in memory at any given moment for performance and RAM usage reasons, and this setting controls that N. Note that there are two places where max_matches limit is enforced. Per-query limit is controlled by this API call, but there also is per-server limit controlled by max_matches setting in the config file. To prevent RAM usage abuse, server will not allow to set per-query limit higher than the per-server limit.

You can't retrieve more than max_matches matches to the client application. The default limit is set to 1000. Normally, you must not have to go over this limit. One thousand records is enough to present to the end user. And if you're thinking about pulling the results to application for further sorting or filtering, that would be much more efficient if performed on Sphinx side.

$cutoff setting is intended for advanced performance control. It tells searchd to forcibly stop search query once $cutoff matches had been found and processed.

9.2.2. SetMaxQueryTime

Prototype: function SetMaxQueryTime ( $max_query_time )

Sets maximum search query time, in milliseconds. Parameter must be a non-negative integer. Default value is 0 which means "do not limit".

Similar to $cutoff setting from SetLimits(), but limits elapsed query time instead of processed matches count. Local search queries will be stopped once that much time has elapsed. Note that if you're performing a search which queries several local indexes, this limit applies to each index separately.

9.2.3. SetOverride


Prototype: function SetOverride ( $attrname, $attrtype, $values )

Sets temporary (per-query) per-document attribute value overrides. Only supports scalar attributes. $values must be a hash that maps document IDs to overridden attribute values. Introduced in version 0.9.9-rc1.

Override feature lets you "temporary" update attribute values for some documents within a single query, leaving all other queries unaffected. This might be useful for personalized data. For example, assume you're implementing a personalized search function that wants to boost the posts that the user's friends recommend. Such data is not just dynamic, but also personal; so you can't simply put it in the index because you don't want everyone's searches affected. Overrides, on the other hand, are local to a single query and invisible to everyone else. So you can, say, setup a "friends_weight" value for every document, defaulting to 0, then temporary override it with 1 for documents 123, 456 and 789 (recommended by exactly the friends of current user), and use that value when ranking.

9.2.4. SetSelect

Prototype: function SetSelect ( $clause )

Sets the select clause, listing specific attributes to fetch, and expressions to compute and fetch. Clause syntax mimics SQL. Introduced in version 0.9.9-rc1.

SetSelect() is very similar to the part of a typical SQL query between SELECT and FROM. It lets you choose what attributes (columns) to fetch, and also what expressions over the columns to compute and fetch. A certain difference from SQL is that expressions must always be aliased to a correct identifier (consisting of letters and digits) using 'AS' keyword. SQL also lets you do that but does not require to. Sphinx enforces aliases so that the computation results can always be returned under a "normal" name in the result set, used in other clauses, etc.

Everything else is basically identical to SQL. Star ('*') is supported. Functions are supported. Arbitrary amount of expressions is supported. Computed expressions can be used for sorting, filtering, and grouping, just as the regular attributes.

Starting with version 0.9.9-rc2, aggregate functions (AVG(), MIN(), MAX(), SUM()) are supported when using GROUP BY.

Expression sorting (Section 5.6, “SPH_SORT_EXPR mode”) and geodistance functions (Section 9.4.5, “SetGeoAnchor”) are now internally implemented using this computed expressions mechanism, using magic names '@expr' and '@geodist' respectively.


$cl->SetSelect ( "*, @weight+(user_karma+ln(pageviews))*0.1 AS myweight" );
$cl->SetSelect ( "exp_years, salary_gbp*{$gbp_usd_rate} AS salary_usd,
   IF(age>40,1,0) AS over40" );
$cl->SetSelect ( "*, AVG(price) AS avgprice" );

9.3. Full-text search query settings

9.3.1. SetMatchMode


Prototype: function SetMatchMode ( $mode )

Sets full-text query matching mode, as described in Section 5.1, “Matching modes”. Parameter must be a constant specifying one of the known modes.

WARNING: (PHP specific) you must not take the matching mode constant name in quotes, that syntax specifies a string and is incorrect:

$cl->SetMatchMode ( "SPH_MATCH_ANY" ); // INCORRECT! will not work as expected
$cl->SetMatchMode ( SPH_MATCH_ANY ); // correct, works OK

9.3.2. SetRankingMode

Prototype: function SetRankingMode ( $ranker, $rankexpr="" )

Sets ranking mode (aka ranker). Only available in SPH_MATCH_EXTENDED matching mode. Parameter must be a constant specifying one of the known rankers.

By default, in the EXTENDED matching mode Sphinx computes two factors which contribute to the final match weight. The major part is a phrase proximity value between the document text and the query. The minor part is so-called BM25 statistical function, which varies from 0 to 1 depending on the keyword frequency within document (more occurrences yield higher weight) and within the whole index (more rare keywords yield higher weight).

However, in some cases you'd want to compute weight differently - or maybe avoid computing it at all for performance reasons because you're sorting the result set by something else anyway. This can be accomplished by setting the appropriate ranking mode. The list of the modes is available in Section 5.4, “Search results ranking”.

$rankexpr argument was added in version 2.0.2-beta. It lets you specify a ranking formula to use with the expression based ranker, that is, when $ranker is set to SPH_RANK_EXPR. In all other cases, $rankexpr is ignored.

9.3.3. SetSortMode

Prototype: function SetSortMode ( $mode, $sortby="" )

Set matches sorting mode, as described in Section 5.6, “Sorting modes”. Parameter must be a constant specifying one of the known modes.

WARNING: (PHP specific) you must not take the matching mode constant name in quotes, that syntax specifies a string and is incorrect:

$cl->SetSortMode ( "SPH_SORT_ATTR_DESC" ); // INCORRECT! will not work as expected
$cl->SetSortMode ( SPH_SORT_ATTR_ASC ); // correct, works OK

9.3.4. SetWeights

Prototype: function SetWeights ( $weights )

Binds per-field weights in the order of appearance in the index. DEPRECATED, use SetFieldWeights() instead.

9.3.5. SetFieldWeights

Prototype: function SetFieldWeights ( $weights )

Binds per-field weights by name. Parameter must be a hash (associative array) mapping string field names to integer weights.

Match ranking can be affected by per-field weights. For instance, see Section 5.4, “Search results ranking” for an explanation how phrase proximity ranking is affected. This call lets you specify what non-default weights to assign to different full-text fields.

The weights must be positive 32-bit integers. The final weight will be a 32-bit integer too. Default weight value is 1. Unknown field names will be silently ignored.

There is no enforced limit on the maximum weight value at the moment. However, beware that if you set it too high you can start hitting 32-bit wraparound issues. For instance, if you set a weight of 10,000,000 and search in extended mode, then maximum possible weight will be equal to 10 million (your weight) by 1 thousand (internal BM25 scaling factor, see Section 5.4, “Search results ranking”) by 1 or more (phrase proximity rank). The result is at least 10 billion that does not fit in 32 bits and will be wrapped around, producing unexpected results.

9.3.6. SetIndexWeights

Prototype: function SetIndexWeights ( $weights )

Sets per-index weights, and enables weighted summing of match weights across different indexes. Parameter must be a hash (associative array) mapping string index names to integer weights. Default is empty array that means to disable weighting summing.

When a match with the same document ID is found in several different local indexes, by default Sphinx simply chooses the match from the index specified last in the query. This is to support searching through partially overlapping index partitions.

However in some cases the indexes are not just partitions, and you might want to sum the weights across the indexes instead of picking one. SetIndexWeights() lets you do that. With summing enabled, final match weight in result set will be computed as a sum of match weight coming from the given index multiplied by respective per-index weight specified in this call. Ie. if the document 123 is found in index A with the weight of 2, and also in index B with the weight of 3, and you called SetIndexWeights ( array ( "A"=>100, "B"=>10 ) ), the final weight return to the client will be 2*100+3*10 = 230.

9.4. Result set filtering settings

9.4.1. SetIDRange

Prototype: function SetIDRange ( $min, $max )

Sets an accepted range of document IDs. Parameters must be integers. Defaults are 0 and 0; that combination means to not limit by range.

After this call, only those records that have document ID between $min and $max (including IDs exactly equal to $min or $max) will be matched.

9.4.2. SetFilter

Prototype: function SetFilter ( $attribute, $values, $exclude=false )

Adds new integer values set filter.

On this call, additional new filter is added to the existing list of filters. $attribute must be a string with attribute name. $values must be a plain array containing integer values. $exclude must be a boolean value; it controls whether to accept the matching documents (default mode, when $exclude is false) or reject them.

Only those documents where $attribute column value stored in the index matches any of the values from $values array will be matched (or rejected, if $exclude is true).

9.4.3. SetFilterRange

Prototype: function SetFilterRange ( $attribute, $min, $max, $exclude=false )

Adds new integer range filter.

On this call, additional new filter is added to the existing list of filters. $attribute must be a string with attribute name. $min and $max must be integers that define the acceptable attribute values range (including the boundaries). $exclude must be a boolean value; it controls whether to accept the matching documents (default mode, when $exclude is false) or reject them.

Only those documents where $attribute column value stored in the index is between $min and $max (including values that are exactly equal to $min or $max) will be matched (or rejected, if $exclude is true).

9.4.4. SetFilterFloatRange

Prototype: function SetFilterFloatRange ( $attribute, $min, $max, $exclude=false )

Adds new float range filter.

On this call, additional new filter is added to the existing list of filters. $attribute must be a string with attribute name. $min and $max must be floats that define the acceptable attribute values range (including the boundaries). $exclude must be a boolean value; it controls whether to accept the matching documents (default mode, when $exclude is false) or reject them.

Only those documents where $attribute column value stored in the index is between $min and $max (including values that are exactly equal to $min or $max) will be matched (or rejected, if $exclude is true).

9.4.5. SetGeoAnchor

Prototype: function SetGeoAnchor ( $attrlat, $attrlong, $lat, $long )

Sets anchor point for and geosphere distance (geodistance) calculations, and enable them.

$attrlat and $attrlong must be strings that contain the names of latitude and longitude attributes, respectively. $lat and $long are floats that specify anchor point latitude and longitude, in radians.

Once an anchor point is set, you can use magic "@geodist" attribute name in your filters and/or sorting expressions. Sphinx will compute geosphere distance between the given anchor point and a point specified by latitude and longitude attributes from each full-text match, and attach this value to the resulting match. The latitude and longitude values both in SetGeoAnchor and the index attribute data are expected to be in radians. The result will be returned in meters, so geodistance value of 1000.0 means 1 km. 1 mile is approximately 1609.344 meters.

9.4.6. SetFilterString

Prototype: function SetFilterString ( $attribute, $value, $exclude=false )

Adds new string value filter.

On this call, additional new filter is added to the existing list of filters. $attribute must be a string with attribute name. $value must be a string. $exclude must be a boolean value; it controls whether to accept the matching documents (default mode, when $exclude is false) or reject them.

Only those documents where $attribute column value stored in the index matches string value from $value will be matched (or rejected, if $exclude is true).

9.5. GROUP BY settings

9.5.1. SetGroupBy

Prototype: function SetGroupBy ( $attribute, $func, $groupsort="@group desc" )

Sets grouping attribute, function, and groups sorting mode; and enables grouping (as described in Section 5.7, “Grouping (clustering) search results ”).

$attribute is a string that contains group-by attribute name. $func is a constant that chooses a function applied to the attribute value in order to compute group-by key. $groupsort is a clause that controls how the groups will be sorted. Its syntax is similar to that described in Section 5.6, “SPH_SORT_EXTENDED mode”.

Grouping feature is very similar in nature to GROUP BY clause from SQL. Results produces by this function call are going to be the same as produced by the following pseudo code:

SELECT ... GROUP BY $func($attribute) ORDER BY $groupsort

Note that it's $groupsort that affects the order of matches in the final result set. Sorting mode (see Section 9.3.3, “SetSortMode”) affect the ordering of matches within group, ie. what match will be selected as the best one from the group. So you can for instance order the groups by matches count and select the most relevant match within each group at the same time.

Starting with version 0.9.9-rc2, aggregate functions (AVG(), MIN(), MAX(), SUM()) are supported through SetSelect() API call when using GROUP BY.

Starting with version 2.0.1-beta, grouping on string attributes is supported, with respect to current collation.

9.5.2. SetGroupDistinct

Prototype: function SetGroupDistinct ( $attribute )

Sets attribute name for per-group distinct values count calculations. Only available for grouping queries.

$attribute is a string that contains the attribute name. For each group, all values of this attribute will be stored (as RAM limits permit), then the amount of distinct values will be calculated and returned to the client. This feature is similar to COUNT(DISTINCT) clause in standard SQL; so these Sphinx calls:

$cl->SetGroupBy ( "category", SPH_GROUPBY_ATTR, "@count desc" );
$cl->SetGroupDistinct ( "vendor" );

can be expressed using the following SQL clauses:

SELECT id, weight, all-attributes,
    COUNT(DISTINCT vendor) AS @distinct,
    COUNT(*) AS @count
FROM products
GROUP BY category

In the sample pseudo code shown just above, SetGroupDistinct() call corresponds to COUNT(DISINCT vendor) clause only. GROUP BY, ORDER BY, and COUNT(*) clauses are all an equivalent of SetGroupBy() settings. Both queries will return one matching row for each category. In addition to indexed attributes, matches will also contain total per-category matches count, and the count of distinct vendor IDs within each category.

9.6. Querying

9.6.1. Query

Prototype: function Query ( $query, $index="*", $comment="" )

Connects to searchd server, runs given search query with current settings, obtains and returns the result set.

$query is a query string. $index is an index name (or names) string. Returns false and sets GetLastError() message on general error. Returns search result set on success. Additionally, the contents of $comment are sent to the query log, marked in square brackets, just before the search terms, which can be very useful for debugging. Currently, the comment is limited to 128 characters.

Default value for $index is "*" that means to query all local indexes. Characters allowed in index names include Latin letters (a-z), numbers (0-9) and underscore (_); everything else is considered a separator. Note that index name should not start with underscore character. Therefore, all of the following samples calls are valid and will search the same two indexes:

$cl->Query ( "test query", "main delta" );
$cl->Query ( "test query", "main;delta" );
$cl->Query ( "test query", "main, delta" );

Index specification order matters. If document with identical IDs are found in two or more indexes, weight and attribute values from the very last matching index will be used for sorting and returning to client (unless explicitly overridden with SetIndexWeights()). Therefore, in the example above, matches from "delta" index will always win over matches from "main".

On success, Query() returns a result set that contains some of the found matches (as requested by SetLimits()) and additional general per-query statistics. The result set is a hash (PHP specific; other languages might utilize other structures instead of hash) with the following keys and values:


Hash which maps found document IDs to another small hash containing document weight and attribute values (or an array of the similar small hashes if SetArrayResult() was enabled).


Total amount of matches retrieved on server (ie. to the server side result set) by this query. You can retrieve up to this amount of matches from server for this query text with current query settings.


Total amount of matching documents in index (that were found and processed on server).


Hash which maps query keywords (case-folded, stemmed, and otherwise processed) to a small hash with per-keyword statistics ("docs", "hits").


Query error message reported by searchd (string, human readable). Empty if there were no errors.


Query warning message reported by searchd (string, human readable). Empty if there were no warnings.

It should be noted that Query() carries out the same actions as AddQuery() and RunQueries() without the intermediate steps; it is analogous to a single AddQuery() call, followed by a corresponding RunQueries(), then returning the first array element of matches (from the first, and only, query.)

9.6.2. AddQuery

Prototype: function AddQuery ( $query, $index="*", $comment="" )

Adds additional query with current settings to multi-query batch. $query is a query string. $index is an index name (or names) string. Additionally if provided, the contents of $comment are sent to the query log, marked in square brackets, just before the search terms, which can be very useful for debugging. Currently, this is limited to 128 characters. Returns index to results array returned from RunQueries().

Batch queries (or multi-queries) enable searchd to perform internal optimizations if possible. They also reduce network connection overheads and search process creation overheads in all cases. They do not result in any additional overheads compared to simple queries. Thus, if you run several different queries from your web page, you should always consider using multi-queries.

For instance, running the same full-text query but with different sorting or group-by settings will enable searchd to perform expensive full-text search and ranking operation only once, but compute multiple group-by results from its output.

This can be a big saver when you need to display not just plain search results but also some per-category counts, such as the amount of products grouped by vendor. Without multi-query, you would have to run several queries which perform essentially the same search and retrieve the same matches, but create result sets differently. With multi-query, you simply pass all these queries in a single batch and Sphinx optimizes the redundant full-text search internally.

AddQuery() internally saves full current settings state along with the query, and you can safely change them afterwards for subsequent AddQuery() calls. Already added queries will not be affected; there's actually no way to change them at all. Here's an example:

$cl->SetSortMode ( SPH_SORT_RELEVANCE );
$cl->AddQuery ( "hello world", "documents" );

$cl->SetSortMode ( SPH_SORT_ATTR_DESC, "price" );
$cl->AddQuery ( "ipod", "products" );

$cl->AddQuery ( "harry potter", "books" );

$results = $cl->RunQueries ();

With the code above, 1st query will search for "hello world" in "documents" index and sort results by relevance, 2nd query will search for "ipod" in "products" index and sort results by price, and 3rd query will search for "harry potter" in "books" index while still sorting by price. Note that 2nd SetSortMode() call does not affect the first query (because it's already added) but affects both other subsequent queries.

Additionally, any filters set up before an AddQuery() will fall through to subsequent queries. So, if SetFilter() is called before the first query, the same filter will be in place for the second (and subsequent) queries batched through AddQuery() unless you call ResetFilters() first. Alternatively, you can add additional filters as well.

This would also be true for grouping options and sorting options; no current sorting, filtering, and grouping settings are affected by this call; so subsequent queries will reuse current query settings.

AddQuery() returns an index into an array of results that will be returned from RunQueries() call. It is simply a sequentially increasing 0-based integer, ie. first call will return 0, second will return 1, and so on. Just a small helper so you won't have to track the indexes manually if you need then.

9.6.3. RunQueries

Prototype: function RunQueries ()

Connect to searchd, runs a batch of all queries added using AddQuery(), obtains and returns the result sets. Returns false and sets GetLastError() message on general error (such as network I/O failure). Returns a plain array of result sets on success.

Each result set in the returned array is exactly the same as the result set returned from Query().

Note that the batch query request itself almost always succeeds - unless there's a network error, blocking index rotation in progress, or another general failure which prevents the whole request from being processed.

However individual queries within the batch might very well fail. In this case their respective result sets will contain non-empty "error" message, but no matches or query statistics. In the extreme case all queries within the batch could fail. There still will be no general error reported, because API was able to successfully connect to searchd, submit the batch, and receive the results - but every result set will have a specific error message.

9.6.4. ResetFilters

Prototype: function ResetFilters ()

Clears all currently set filters.

This call is only normally required when using multi-queries. You might want to set different filters for different queries in the batch. To do that, you should call ResetFilters() and add new filters using the respective calls.

9.6.5. ResetGroupBy

Prototype: function ResetGroupBy ()

Clears all currently group-by settings, and disables group-by.

This call is only normally required when using multi-queries. You can change individual group-by settings using SetGroupBy() and SetGroupDistinct() calls, but you can not disable group-by using those calls. ResetGroupBy() fully resets previous group-by settings and disables group-by mode in the current state, so that subsequent AddQuery() calls can perform non-grouping searches.

9.7. Additional functionality

9.7.1. BuildExcerpts

Prototype: function BuildExcerpts ( $docs, $index, $words, $opts=array() )

Excerpts (snippets) builder function. Connects to searchd, asks it to generate excerpts (snippets) from given documents, and returns the results.

$docs is a plain array of strings that carry the documents' contents. $index is an index name string. Different settings (such as charset, morphology, wordforms) from given index will be used. $words is a string that contains the keywords to highlight. They will be processed with respect to index settings. For instance, if English stemming is enabled in the index, "shoes" will be highlighted even if keyword is "shoe". Starting with version 0.9.9-rc1, keywords can contain wildcards, that work similarly to star-syntax available in queries. $opts is a hash which contains additional optional highlighting parameters:


A string to insert before a keyword match. Starting with version 1.10-beta, a %PASSAGE_ID% macro can be used in this string. The first match of the macro is replaced with an incrementing passage number within a current snippet. Numbering starts at 1 by default but can be overridden with "start_passage_id" option. In a multi-document call, %PASSAGE_ID% would restart at every given document. Default is "<b>".


A string to insert after a keyword match. Starting with version 1.10-beta, a %PASSAGE_ID% macro can be used in this string. Default is "</b>".


A string to insert between snippet chunks (passages). Default is " ... ".


Maximum snippet size, in symbols (codepoints). Integer, default is 256.


How much words to pick around each matching keywords block. Integer, default is 5.


Whether to highlight exact query phrase matches only instead of individual keywords. Boolean, default is false.


Whether to additionally break passages by phrase boundary characters, as configured in index settings with phrase_boundary directive. Boolean, default is false.


Whether to sort the extracted passages in order of relevance (decreasing weight), or in order of appearance in the document (increasing position). Boolean, default is false.


Added in version 1.10-beta. Whether to handle $words as a query in extended syntax, or as a bag of words (default behavior). For instance, in query mode ("one two" | "three four") will only highlight and include those occurrences "one two" or "three four" when the two words from each pair are adjacent to each other. In default mode, any single occurrence of "one", "two", "three", or "four" would be highlighted. Boolean, default is false.


Added in version 1.10-beta. Ignores the snippet length limit until it includes all the keywords. Boolean, default is false.


Added in version 1.10-beta. Limits the maximum number of passages that can be included into the snippet. Integer, default is 0 (no limit).


Added in version 1.10-beta. Limits the maximum number of words that can be included into the snippet. Note the limit applies to any words, and not just the matched keywords to highlight. For example, if we are highlighting "Mary" and a passage "Mary had a little lamb" is selected, then it contributes 5 words to this limit, not just 1. Integer, default is 0 (no limit).


Added in version 1.10-beta. Specifies the starting value of %PASSAGE_ID% macro (that gets detected and expanded in before_match, after_match strings). Integer, default is 1.


Added in version 1.10-beta. Whether to handle $docs as data to extract snippets from (default behavior), or to treat it as file names, and load data from specified files on the server side. Starting with version 2.0.1-beta, up to dist_threads worker threads per request will be created to parallelize the work when this flag is enabled. Boolean, default is false. Starting with version 2.0.2-beta, building of the snippets could be parallelized between remote agents. Just set the 'dist_threads' param in the config to the value greater than 1, and then invoke the snippets generation over the distributed index, which contain only one(!) local agent and several remotes. Starting with version 2.1.1-beta, the snippets_file_prefix option is also in the game and the final filename is calculated by concatenation of the prefix with given name. Otherwords, when snippets_file_prefix is '/var/data' and filename is 'text.txt' the sphinx will try to generate the snippets from the file '/var/datatext.txt', which is exactly '/var/data' + 'text.txt'.


Added in version 2.0.2-beta. It works only with distributed snippets generation with remote agents. The source files for snippets could be distributed among different agents, and the main daemon will merge together all non-erroneous results. So, if one agent of the distributed index has 'file1.txt', another has 'file2.txt' and you call for the snippets with both these files, the sphinx will merge results from the agents together, so you will get the snippets from both 'file1.txt' and 'file2.txt'. Boolean, default is false.

If the "load_files" is also set, the request will return the error in case if any of the files is not available anywhere. Otherwise (if "load_files" is not set) it will just return the empty strings for all absent files. The master instance reset this flag when distributes the snippets among agents. So, for agents the absence of a file is not critical error, but for the master it might be so. If you want to be sure that all snippets are actually created, set both "load_files_scattered" and "load_files". If the absence of some snippets caused by some agents is not critical for you - set just "load_files_scattered", leaving "load_files" not set.


Added in version 1.10-beta. HTML stripping mode setting. Defaults to "index", which means that index settings will be used. The other values are "none" and "strip", that forcibly skip or apply stripping irregardless of index settings; and "retain", that retains HTML markup and protects it from highlighting. The "retain" mode can only be used when highlighting full documents and thus requires that no snippet size limits are set. String, allowed values are "none", "strip", "index", and "retain".


Added in version 1.10-beta. Allows empty string to be returned as highlighting result when a snippet could not be generated (no keywords match, or no passages fit the limit). By default, the beginning of original text would be returned instead of an empty string. Boolean, default is false.


Added in version 2.0.1-beta. Ensures that passages do not cross a sentence, paragraph, or zone boundary (when used with an index that has the respective indexing settings enabled). String, allowed values are "sentence", "paragraph", and "zone".


Added in version 2.0.1-beta. Emits an HTML tag with an enclosing zone name before each passage. Boolean, default is false.

Snippets extraction algorithm currently favors better passages (with closer phrase matches), and then passages with keywords not yet in snippet. Generally, it will try to highlight the best match with the query, and it will also to highlight all the query keywords, as made possible by the limits. In case the document does not match the query, beginning of the document trimmed down according to the limits will be return by default. Starting with 1.10-beta, you can also return an empty snippet instead case by setting "allow_empty" option to true.

Returns false on failure. Returns a plain array of strings with excerpts (snippets) on success.

9.7.2. UpdateAttributes

Prototype: function UpdateAttributes ( $index, $attrs, $values, $mva=false, $ignorenonexistent=false )

Instantly updates given attribute values in given documents. Returns number of actually updated documents (0 or more) on success, or -1 on failure.

$index is a name of the index (or indexes) to be updated. $attrs is a plain array with string attribute names, listing attributes that are updated. $values is a hash where key is document ID, and value is a plain array of new attribute values. Optional boolean parameter mva points that there is update of MVA attributes. In this case the $values must be a dict with int key (document ID) and list of lists of int values (new MVA attribute values). Optional boolean parameter $ignorenonexistent (added in version 2.1.1-beta) points that the update will silently ignore any warnings about trying to update a column which is not exists in current index schema.

$index can be either a single index name or a list, like in Query(). Unlike Query(), wildcard is not allowed and all the indexes to update must be specified explicitly. The list of indexes can include distributed index names. Updates on distributed indexes will be pushed to all agents.

The updates only work with docinfo=extern storage strategy. They are very fast because they're working fully in RAM, but they can also be made persistent: updates are saved on disk on clean searchd shutdown initiated by SIGTERM signal. With additional restrictions, updates are also possible on MVA attributes; refer to mva_updates_pool directive for details.

Usage example:

$cl->UpdateAttributes ( "test1", array("group_id"), array(1=>array(456)) );
$cl->UpdateAttributes ( "products", array ( "price", "amount_in_stock" ),
    array ( 1001=>array(123,5), 1002=>array(37,11), 1003=>(25,129) ) );

The first sample statement will update document 1 in index "test1", setting "group_id" to 456. The second one will update documents 1001, 1002 and 1003 in index "products". For document 1001, the new price will be set to 123 and the new amount in stock to 5; for document 1002, the new price will be 37 and the new amount will be 11; etc.

9.7.3. BuildKeywords

Prototype: function BuildKeywords ( $query, $index, $hits )

Extracts keywords from query using tokenizer settings for given index, optionally with per-keyword occurrence statistics. Returns an array of hashes with per-keyword information.

$query is a query to extract keywords from. $index is a name of the index to get tokenizing settings and keyword occurrence statistics from. $hits is a boolean flag that indicates whether keyword occurrence statistics are required.

Usage example:

$keywords = $cl->BuildKeywords ( "this.is.my query", "test1", false );

9.7.4. EscapeString

Prototype: function EscapeString ( $string )

Escapes characters that are treated as special operators by the query language parser. Returns an escaped string.

$string is a string to escape.

This function might seem redundant because it's trivial to implement in any calling application. However, as the set of special characters might change over time, it makes sense to have an API call that is guaranteed to escape all such characters at all times.

Usage example:

$escaped = $cl->EscapeString ( "escaping-sample@query/string" );

9.7.5. Status

Prototype: function Status ()

Queries searchd status, and returns an array of status variable name and value pairs.

Usage example:

$status = $cl->Status ();
foreach ( $status as $row )
    print join ( ": ", $row ) . "\n";

9.7.6. FlushAttributes

Prototype: function FlushAttributes ()

Forces searchd to flush pending attribute updates to disk, and blocks until completion. Returns a non-negative internal "flush tag" on success. Returns -1 and sets an error message on error. Introduced in version 1.10-beta.

Attribute values updated using UpdateAttributes() API call are only kept in RAM until a so-called flush (which writes the current, possibly updated attribute values back to disk). FlushAttributes() call lets you enforce a flush. The call will block until searchd finishes writing the data to disk, which might take seconds or even minutes depending on the total data size (.spa file size). All the currently updated indexes will be flushed.

Flush tag should be treated as an ever growing magic number that does not mean anything. It's guaranteed to be non-negative. It is guaranteed to grow over time, though not necessarily in a sequential fashion; for instance, two calls that return 10 and then 1000 respectively are a valid situation. If two calls to FlushAttrs() return the same tag, it means that there were no actual attribute updates in between them, and therefore current flushed state remained the same (for all indexes).

Usage example:

$status = $cl->FlushAttributes ();
if ( $status<0 )
    print "ERROR: " . $cl->GetLastError();

9.8. Persistent connections

Persistent connections allow to use single network connection to run multiple commands that would otherwise require reconnects.

9.8.1. Open

Prototype: function Open ()

Opens persistent connection to the server.

9.8.2. Close

Prototype: function Close ()

Closes previously opened persistent connection.

Chapter 10. MySQL storage engine (SphinxSE)

10.1. SphinxSE overview

SphinxSE is MySQL storage engine which can be compiled into MySQL server 5.x using its pluggable architecture. It is not available for MySQL 4.x series. It also requires MySQL 5.0.22 or higher in 5.0.x series, or MySQL 5.1.12 or higher in 5.1.x series.

Despite the name, SphinxSE does not actually store any data itself. It is actually a built-in client which allows MySQL server to talk to searchd, run search queries, and obtain search results. All indexing and searching happen outside MySQL.

Obvious SphinxSE applications include:

  • easier porting of MySQL FTS applications to Sphinx;

  • allowing Sphinx use with programming languages for which native APIs are not available yet;

  • optimizations when additional Sphinx result set processing on MySQL side is required (eg. JOINs with original document tables, additional MySQL-side filtering, etc).

10.2. Installing SphinxSE

You will need to obtain a copy of MySQL sources, prepare those, and then recompile MySQL binary. MySQL sources (mysql-5.x.yy.tar.gz) could be obtained from dev.mysql.com Web site.

For some MySQL versions, there are delta tarballs with already prepared source versions available from Sphinx Web site. After unzipping those over original sources MySQL would be ready to be configured and built with Sphinx support.

If such tarball is not available, or does not work for you for any reason, you would have to prepare sources manually. You will need to GNU Autotools framework (autoconf, automake and libtool) installed to do that.

10.2.1. Compiling MySQL 5.0.x with SphinxSE

  1. copy sphinx.5.0.yy.diff patch file into MySQL sources directory and run

    patch -p1 < sphinx.5.0.yy.diff

    If there's no .diff file exactly for the specific version you need to build, try applying .diff with closest version numbers. It is important that the patch should apply with no rejects.

  2. in MySQL sources directory, run

    sh BUILD/autorun.sh

  3. in MySQL sources directory, create sql/sphinx directory in and copy all files in mysqlse directory from Sphinx sources there. Example:

    cp -R /root/builds/sphinx-0.9.7/mysqlse /root/builds/mysql-5.0.24/sql/sphinx

  4. configure MySQL and enable Sphinx engine:

    ./configure --with-sphinx-storage-engine

  5. build and install MySQL:

    make install

10.2.2. Compiling MySQL 5.1.x with SphinxSE

  1. in MySQL sources directory, create storage/sphinx directory in and copy all files in mysqlse directory from Sphinx sources there. Example:

    cp -R /root/builds/sphinx-0.9.7/mysqlse /root/builds/mysql-5.1.14/storage/sphinx

  2. in MySQL sources directory, run

    sh BUILD/autorun.sh

  3. configure MySQL and enable Sphinx engine:

    ./configure --with-plugins=sphinx

  4. build and install MySQL:

    make install

10.2.3. Checking SphinxSE installation

To check whether SphinxSE has been successfully compiled into MySQL, launch newly built servers, run mysql client and issue SHOW ENGINES query. You should see a list of all available engines. Sphinx should be present and "Support" column should contain "YES":

mysql> show engines;
| Engine     | Support  | Comment                                                     |
| MyISAM     | DEFAULT  | Default engine as of MySQL 3.23 with great performance      |
| SPHINX     | YES      | Sphinx storage engine                                       |
13 rows in set (0.00 sec)

10.3. Using SphinxSE

To search via SphinxSE, you would need to create special ENGINE=SPHINX "search table", and then SELECT from it with full text query put into WHERE clause for query column.

Let's begin with an example create statement and search query:

    weight      INTEGER NOT NULL,
    query       VARCHAR(3072) NOT NULL,
    group_id    INTEGER,
) ENGINE=SPHINX CONNECTION="sphinx://localhost:9312/test";

SELECT * FROM t1 WHERE query='test it;mode=any';

First 3 columns of search table must have a types of INTEGER UNSINGED or BIGINT for the 1st column (document id), INTEGER or BIGINT for the 2nd column (match weight), and VARCHAR or TEXT for the 3rd column (your query), respectively. This mapping is fixed; you can not omit any of these three required columns, or move them around, or change types. Also, query column must be indexed; all the others must be kept unindexed. Columns' names are ignored so you can use arbitrary ones.

Additional columns must be either INTEGER, TIMESTAMP, BIGINT, VARCHAR, or FLOAT. They will be bound to attributes provided in Sphinx result set by name, so their names must match attribute names specified in sphinx.conf. If there's no such attribute name in Sphinx search results, column will have NULL values.

Special "virtual" attributes names can also be bound to SphinxSE columns. _sph_ needs to be used instead of @ for that. For instance, to obtain the values of @groupby, @count, or @distinct virtual attributes, use _sph_groupby, _sph_count or _sph_distinct column names, respectively.

CONNECTION string parameter can be used to specify default searchd host, port and indexes for queries issued using this table. If no connection string is specified in CREATE TABLE, index name "*" (ie. search all indexes) and localhost:9312 are assumed. Connection string syntax is as follows:


You can change the default connection string later:


You can also override all these parameters per-query.

As seen in example, both query text and search options should be put into WHERE clause on search query column (ie. 3rd column); the options are separated by semicolons; and their names from values by equality sign. Any number of options can be specified. Available options are:

  • query - query text;

  • mode - matching mode. Must be one of "all", "any", "phrase", "boolean", or "extended". Default is "all";

  • sort - match sorting mode. Must be one of "relevance", "attr_desc", "attr_asc", "time_segments", or "extended". In all modes besides "relevance" attribute name (or sorting clause for "extended") is also required after a colon:

    ... WHERE query='test;sort=attr_asc:group_id';
    ... WHERE query='test;sort=extended:@weight desc, group_id asc';

  • offset - offset into result set, default is 0;

  • limit - amount of matches to retrieve from result set, default is 20;

  • index - names of the indexes to search:

    ... WHERE query='test;index=test1;';
    ... WHERE query='test;index=test1,test2,test3;';

  • minid, maxid - min and max document ID to match;

  • weights - comma-separated list of weights to be assigned to Sphinx full-text fields:

    ... WHERE query='test;weights=1,2,3;';

  • filter, !filter - comma-separated attribute name and a set of values to match:

    # only include groups 1, 5 and 19
    ... WHERE query='test;filter=group_id,1,5,19;';
    # exclude groups 3 and 11
    ... WHERE query='test;!filter=group_id,3,11;';

  • range, !range - comma-separated (integer or bigint) Sphinx attribute name, and min and max values to match:

    # include groups from 3 to 7, inclusive
    ... WHERE query='test;range=group_id,3,7;';
    # exclude groups from 5 to 25
    ... WHERE query='test;!range=group_id,5,25;';

  • floatrange, !floatrange - comma-separated (floating point) Sphinx attribute name, and min and max values to match:

    # filter by a float size
    ... WHERE query='test;floatrange=size,2,3;';
    # pick all results within 1000 meter from geoanchor
    ... WHERE query='test;floatrange=@geodist,0,1000;';

  • maxmatches - per-query max matches value, as in max_matches parameter to SetLimits() API call:

    ... WHERE query='test;maxmatches=2000;';

  • cutoff - maximum allowed matches, as in cutoff parameter to SetLimits() API call:

    ... WHERE query='test;cutoff=10000;';

  • maxquerytime - maximum allowed query time (in milliseconds), as in SetMaxQueryTime() API call:

    ... WHERE query='test;maxquerytime=1000;';

  • groupby - group-by function and attribute, corresponding to SetGroupBy() API call:

    ... WHERE query='test;groupby=day:published_ts;';
    ... WHERE query='test;groupby=attr:group_id;';

  • groupsort - group-by sorting clause:

    ... WHERE query='test;groupsort=@count desc;';

  • distinct - an attribute to compute COUNT(DISTINCT) for when doing group-by, as in SetGroupDistinct() API call:

    ... WHERE query='test;groupby=attr:country_id;distinct=site_id';

  • indexweights - comma-separated list of index names and weights to use when searching through several indexes:

    ... WHERE query='test;indexweights=idx_exact,2,idx_stemmed,1;';

  • fieldweights - comma-separated list of per-field weights that can be used by the ranker:

    ... WHERE query='test;fieldweights=title,10,abstract,3,content,1;';

  • comment - a string to mark this query in query log (mapping to $comment parameter in Query() API call):

    ... WHERE query='test;comment=marker001;';

  • select - a string with expressions to compute (mapping to SetSelect() API call):

    ... WHERE query='test;select=2*a+3*b as myexpr;';

  • host, port - remote searchd host name and TCP port, respectively:

    ... WHERE query='test;host=sphinx-test.loc;port=7312;';

  • ranker - a ranking function to use with "extended" matching mode, as in SetRankingMode() API call (the only mode that supports full query syntax). Known values are "proximity_bm25", "bm25", "none", "wordcount", "proximity", "matchany", "fieldmask", "sph04" (starting with 1.10-beta), "expr:EXPRESSION" (starting with 2.0.4-release) syntax to support expression-based ranker (where EXPRESSION should be replaced with your specific ranking formula), and "export:EXPRESSION" (starting with 2.1.1-beta):

    ... WHERE query='test;mode=extended;ranker=bm25;';
    ... WHERE query='test;mode=extended;ranker=expr:sum(lcs);';

    The "export" ranker works exactly like ranker=expr, but it stores the per-document factor values, while ranker=expr discards them after computing the final WEIGHT() value. Note that ranker=export is meant to be used but rarely, only to train a ML (machine learning) function or to define your own ranking function by hand, and never in actual production. When using this ranker, you'll probably want to examine the output of the RANKFACTORS() function (added in version 2.1.1-beta) that produces a string with all the field level factors for each document.

            FROM myindex
            WHERE MATCH('dog')
            OPTION ranker=export('100*bm25')

    would produce something like

    *************************** 1. row ***************************
               id: 555617
        published: 1110067331
       channel_id: 1059819
            title: 7
          content: 428
         weight(): 69900
    rankfactors(): bm25=699, bm25a=0.666478, field_mask=2,
    doc_word_count=1, field1=(lcs=1, hit_count=4, word_count=1,
    tf_idf=1.038127, min_idf=0.259532, max_idf=0.259532, sum_idf=0.259532,
    min_hit_pos=120, min_best_span_pos=120, exact_hit=0,
    max_window_hits=1), word1=(tf=4, idf=0.259532)
    *************************** 2. row ***************************
               id: 555313
        published: 1108438365
       channel_id: 1058561
            title: 8
          content: 249
         weight(): 68500
    rankfactors(): bm25=685, bm25a=0.675213, field_mask=3,
    doc_word_count=1, field0=(lcs=1, hit_count=1, word_count=1,
    tf_idf=0.259532, min_idf=0.259532, max_idf=0.259532, sum_idf=0.259532,
    min_hit_pos=8, min_best_span_pos=8, exact_hit=0, max_window_hits=1),
    field1=(lcs=1, hit_count=2, word_count=1, tf_idf=0.519063,
    min_idf=0.259532, max_idf=0.259532, sum_idf=0.259532, min_hit_pos=36,
    min_best_span_pos=36, exact_hit=0, max_window_hits=1), word1=(tf=3,
  • geoanchor - geodistance anchor, as in SetGeoAnchor() API call. Takes 4 parameters which are latitude and longitude attribute names, and anchor point coordinates respectively:

    ... WHERE query='test;geoanchor=latattr,lonattr,0.123,0.456';

One very important note that it is much more efficient to allow Sphinx to perform sorting, filtering and slicing the result set than to raise max matches count and use WHERE, ORDER BY and LIMIT clauses on MySQL side. This is for two reasons. First, Sphinx does a number of optimizations and performs better than MySQL on these tasks. Second, less data would need to be packed by searchd, transferred and unpacked by SphinxSE.

Starting with version 0.9.9-rc1, additional query info besides result set could be retrieved with SHOW ENGINE SPHINX STATUS statement:

| Type   | Name  | Status                                          |
| SPHINX | stats | total: 25, total found: 25, time: 126, words: 2 |
| SPHINX | words | sphinx:591:1256 soft:11076:15945                |
2 rows in set (0.00 sec)

This information can also be accessed through status variables. Note that this method does not require super-user privileges.

mysql> SHOW STATUS LIKE 'sphinx_%';
| Variable_name      | Value                            |
| sphinx_total       | 25                               |
| sphinx_total_found | 25                               |
| sphinx_time        | 126                              |
| sphinx_word_count  | 2                                |
| sphinx_words       | sphinx:591:1256 soft:11076:15945 |
5 rows in set (0.00 sec)

You could perform JOINs on SphinxSE search table and tables using other engines. Here's an example with "documents" from example.sql:

mysql> SELECT content, date_added FROM test.documents docs
-> JOIN t1 ON (docs.id=t1.id)
-> WHERE query="one document;mode=any";
| content                             | docdate             |
| this is my test document number two | 2006-06-17 14:04:28 |
| this is my test document number one | 2006-06-17 14:04:28 |
2 rows in set (0.00 sec)

| Type   | Name  | Status                                      |
| SPHINX | stats | total: 2, total found: 2, time: 0, words: 2 |
| SPHINX | words | one:1:2 document:2:2                        |
2 rows in set (0.00 sec)

10.4. Building snippets (excerpts) via MySQL

Starting with version 0.9.9-rc2, SphinxSE also includes a UDF function that lets you create snippets through MySQL. The functionality is fully similar to BuildExcerprts API call but accessible through MySQL+SphinxSE.

The binary that provides the UDF is named sphinx.so and should be automatically built and installed to proper location along with SphinxSE itself. If it does not get installed automatically for some reason, look for sphinx.so in the build directory and copy it to the plugins directory of your MySQL instance. After that, register the UDF using the following statement:

CREATE FUNCTION sphinx_snippets RETURNS STRING SONAME 'sphinx.so';

Function name must be sphinx_snippets, you can not use an arbitrary name. Function arguments are as follows:

Prototype: function sphinx_snippets ( document, index, words, [options] );

Document and words arguments can be either strings or table columns. Options must be specified like this: 'value' AS option_name. For a list of supported options, refer to BuildExcerprts() API call. The only UDF-specific additional option is named 'sphinx' and lets you specify searchd location (host and port).

Usage examples:

SELECT sphinx_snippets('hello world doc', 'main', 'world',
    'sphinx://' AS sphinx, true AS exact_phrase,
    '[b]' AS before_match, '[/b]' AS after_match)
FROM documents;

SELECT title, sphinx_snippets(text, 'index', 'mysql php') AS text
    FROM sphinx, documents
    WHERE query='mysql php' AND sphinx.id=documents.id;

Chapter 11. Reporting bugs

Unfortunately, Sphinx is not yet 100% bug free (even though we're working hard towards that), so you might occasionally run into some issues.

Reporting as much as possible about each bug is very important - because to fix it, we need to be able either to reproduce and fix the bug, or to deduce what's causing it from the information that you provide. So here are some instructions on how to do that.


Nothing special to say here. Here is the <a href="http://sphinxsearch.com/bugs">link</a>. Create a new ticket and describe your bug in details so both you and developers can save their time.


In case of crashes we sometimes can get enough info to fix from backtrace.

Sphinx tries to write crash backtrace to its log file. It may look like this:


This is an example of a good backtrace - we can see mangled function names here.

But sometimes backtrace may look like this:


Developers can get nothing useful from those cryptic numbers. They're ordinary humans and want to see function names. To help them you need to provide symbols (function and variable names). If you've installed sphinx by building from the sources, run the following command over your binary:

nm -n indexer > indexer.sym

Attach this file to bug report along with backtrace. You should however ensure that the binary is not stripped. Our official binary packages should be fine. (That, or we have the symbols stored.) However, if you manually build Sphinx from the source tarball, do not run strip utility on that binary, and/or do not let your build/packaging system do that!

Uploading your data

To fix your bug developers often need to reproduce it on their machines. To do this they need your sphinx.conf, index files, binlog (if present), sometimes data to index (like SQL tables or XMLpipe2 data files) and queries.

Attach your data to ticket. In case it's too big to attach ask developers and they give you an address to write-only FTP created exactly for such puproses.

Chapter 12. sphinx.conf options reference

Table of Contents

12.1. Data source configuration options
12.1.1. type
12.1.2. sql_host
12.1.3. sql_port
12.1.4. sql_user
12.1.5. sql_pass
12.1.6. sql_db
12.1.7. sql_sock
12.1.8. mysql_connect_flags
12.1.9. mysql_ssl_cert, mysql_ssl_key, mysql_ssl_ca
12.1.10. odbc_dsn
12.1.11. sql_query_pre
12.1.12. sql_query
12.1.13. sql_joined_field
12.1.14. sql_query_range
12.1.15. sql_range_step
12.1.16. sql_query_killlist
12.1.17. sql_attr_uint
12.1.18. sql_attr_bool
12.1.19. sql_attr_bigint
12.1.20. sql_attr_timestamp
12.1.21. sql_attr_float
12.1.22. sql_attr_multi
12.1.23. sql_attr_string
12.1.24. sql_attr_json
12.1.25. sql_column_buffers
12.1.26. sql_field_string
12.1.27. sql_file_field
12.1.28. sql_query_post
12.1.29. sql_query_post_index
12.1.30. sql_ranged_throttle
12.1.31. xmlpipe_command
12.1.32. xmlpipe_field
12.1.33. xmlpipe_field_string
12.1.34. xmlpipe_attr_uint
12.1.35. xmlpipe_attr_bigint
12.1.36. xmlpipe_attr_bool
12.1.37. xmlpipe_attr_timestamp
12.1.38. xmlpipe_attr_float
12.1.39. xmlpipe_attr_multi
12.1.40. xmlpipe_attr_multi_64
12.1.41. xmlpipe_attr_string
12.1.42. xmlpipe_attr_json
12.1.43. xmlpipe_fixup_utf8
12.1.44. mssql_winauth
12.1.45. unpack_zlib
12.1.46. unpack_mysqlcompress
12.1.47. unpack_mysqlcompress_maxsize
12.1.48. csvpipe_delimiter
12.2. Index configuration options
12.2.1. type
12.2.2. source
12.2.3. path
12.2.4. docinfo
12.2.5. mlock
12.2.6. morphology
12.2.7. dict
12.2.8. index_sp
12.2.9. index_zones
12.2.10. min_stemming_len
12.2.11. stopwords
12.2.12. wordforms
12.2.13. embedded_limit
12.2.14. exceptions
12.2.15. min_word_len
12.2.16. charset_table
12.2.17. ignore_chars
12.2.18. min_prefix_len
12.2.19. min_infix_len
12.2.20. max_substring_len
12.2.21. prefix_fields
12.2.22. infix_fields
12.2.23. ngram_len
12.2.24. ngram_chars
12.2.25. phrase_boundary
12.2.26. phrase_boundary_step
12.2.27. html_strip
12.2.28. html_index_attrs
12.2.29. html_remove_elements
12.2.30. local
12.2.31. agent
12.2.32. agent_persistent
12.2.33. agent_blackhole
12.2.34. agent_connect_timeout
12.2.35. agent_query_timeout
12.2.36. preopen
12.2.37. inplace_enable
12.2.38. inplace_hit_gap
12.2.39. inplace_docinfo_gap
12.2.40. inplace_reloc_factor
12.2.41. inplace_write_factor
12.2.42. index_exact_words
12.2.43. overshort_step
12.2.44. stopword_step
12.2.45. hitless_words
12.2.46. expand_keywords
12.2.47. blend_chars
12.2.48. blend_mode
12.2.49. rt_mem_limit
12.2.50. rt_field
12.2.51. rt_attr_uint
12.2.52. rt_attr_bool
12.2.53. rt_attr_bigint
12.2.54. rt_attr_float
12.2.55. rt_attr_multi
12.2.56. rt_attr_multi_64
12.2.57. rt_attr_timestamp
12.2.58. rt_attr_string
12.2.59. rt_attr_json
12.2.60. ha_strategy
12.2.61. bigram_freq_words
12.2.62. bigram_index
12.2.63. index_field_lengths
12.2.64. regexp_filter
12.2.65. stopwords_unstemmed
12.2.66. global_idf
12.2.67. rlp_context
12.2.68. ondisk_attrs
12.3. indexer program configuration options
12.3.1. mem_limit
12.3.2. max_iops
12.3.3. max_iosize
12.3.4. max_xmlpipe2_field
12.3.5. write_buffer
12.3.6. max_file_field_buffer
12.3.7. on_file_field_error
12.3.8. lemmatizer_cache
12.4. searchd program configuration options
12.4.1. listen
12.4.2. log
12.4.3. query_log
12.4.4. query_log_format
12.4.5. read_timeout
12.4.6. client_timeout
12.4.7. max_children
12.4.8. pid_file
12.4.9. seamless_rotate
12.4.10. preopen_indexes
12.4.11. unlink_old
12.4.12. attr_flush_period
12.4.13. max_packet_size
12.4.14. mva_updates_pool
12.4.15. max_filters
12.4.16. max_filter_values
12.4.17. listen_backlog
12.4.18. read_buffer
12.4.19. read_unhinted
12.4.20. max_batch_queries
12.4.21. subtree_docs_cache
12.4.22. subtree_hits_cache
12.4.23. workers
12.4.24. dist_threads
12.4.25. binlog_path
12.4.26. binlog_flush
12.4.27. binlog_max_log_size
12.4.28. snippets_file_prefix
12.4.29. collation_server
12.4.30. collation_libc_locale
12.4.31. mysql_version_string
12.4.32. rt_flush_period
12.4.33. thread_stack
12.4.34. expansion_limit
12.4.35. watchdog
12.4.36. prefork_rotation_throttle
12.4.37. sphinxql_state
12.4.38. ha_ping_interval
12.4.39. ha_period_karma
12.4.40. persistent_connections_limit
12.4.41. rt_merge_iops
12.4.42. rt_merge_maxiosize
12.4.43. predicted_time_costs
12.4.44. shutdown_timeout
12.4.45. ondisk_attrs_default
12.4.46. query_log_min_msec
12.4.47. agent_connect_timeout
12.4.48. agent_query_timeout
12.4.49. agent_retry_count
12.4.50. agent_retry_delay
12.5. Common section configuration options
12.5.1. lemmatizer_base
12.5.2. on_json_attr_error
12.5.3. json_autoconv_numbers
12.5.4. json_autoconv_keynames
12.5.5. rlp_root
12.5.6. rlp_environment
12.5.7. rlp_max_batch_size
12.5.8. rlp_max_batch_docs
12.5.9. plugin_dir

12.1. Data source configuration options

12.1.1. type

Data source type. Mandatory, no default value. Known types are mysql, pgsql, mssql, xmlpipe2, tsvpipe, csvpipe and odbc.

All other per-source options depend on source type selected by this option. Names of the options used for SQL sources (ie. MySQL, PostgreSQL, MS SQL) start with "sql_"; names of the ones used for xmlpipe2 or tsvpipe, csvpipe start with "xmlpipe_" and "tsvpipe_", "csvpipe_" correspondingly. All source types are conditional; they might or might not be supported depending on your build settings, installed client libraries, etc. mssql type is currently only available on Windows. odbc type is available both on Windows natively and on Linux through UnixODBC library.


type = mysql

12.1.2. sql_host

SQL server host to connect to. Mandatory, no default value. Applies to SQL source types (mysql, pgsql, mssql) only.

In the simplest case when Sphinx resides on the same host with your MySQL or PostgreSQL installation, you would simply specify "localhost". Note that MySQL client library chooses whether to connect over TCP/IP or over UNIX socket based on the host name. Specifically "localhost" will force it to use UNIX socket (this is the default and generally recommended mode) and "" will force TCP/IP usage. Refer to MySQL manual for more details.


sql_host = localhost

12.1.3. sql_port

SQL server IP port to connect to. Optional, default is 3306 for mysql source type and 5432 for pgsql type. Applies to SQL source types (mysql, pgsql, mssql) only. Note that it depends on sql_host setting whether this value will actually be used.


sql_port = 3306

12.1.4. sql_user

SQL user to use when connecting to sql_host. Mandatory, no default value. Applies to SQL source types (mysql, pgsql, mssql) only.


sql_user = test

12.1.5. sql_pass

SQL user password to use when connecting to sql_host. Mandatory, no default value. Applies to SQL source types (mysql, pgsql, mssql) only.


sql_pass = mysecretpassword

12.1.6. sql_db

SQL database (in MySQL terms) to use after the connection and perform further queries within. Mandatory, no default value. Applies to SQL source types (mysql, pgsql, mssql) only.


sql_db = test

12.1.7. sql_sock

UNIX socket name to connect to for local SQL servers. Optional, default value is empty (use client library default settings). Applies to SQL source types (mysql, pgsql, mssql) only.

On Linux, it would typically be /var/lib/mysql/mysql.sock. On FreeBSD, it would typically be /tmp/mysql.sock. Note that it depends on sql_host setting whether this value will actually be used.


sql_sock = /tmp/mysql.sock

12.1.8. mysql_connect_flags

MySQL client connection flags. Optional, default value is 0 (do not set any flags). Applies to mysql source type only.

This option must contain an integer value with the sum of the flags. The value will be passed to mysql_real_connect() verbatim. The flags are enumerated in mysql_com.h include file. Flags that are especially interesting in regard to indexing, with their respective values, are as follows:

  • CLIENT_COMPRESS = 32; can use compression protocol

  • CLIENT_SSL = 2048; switch to SSL after handshake

  • CLIENT_SECURE_CONNECTION = 32768; new 4.1 authentication

For instance, you can specify 2080 (2048+32) to use both compression and SSL, or 32768 to use new authentication only. Initially, this option was introduced to be able to use compression when the indexer and mysqld are on different hosts. Compression on 1 Gbps links is most likely to hurt indexing time though it reduces network traffic, both in theory and in practice. However, enabling compression on 100 Mbps links may improve indexing time significantly (upto 20-30% of the total indexing time improvement was reported). Your mileage may vary.


mysql_connect_flags = 32 # enable compression

12.1.9. mysql_ssl_cert, mysql_ssl_key, mysql_ssl_ca

SSL certificate settings to use for connecting to MySQL server. Optional, default values are empty strings (do not use SSL). Applies to mysql source type only.

These directives let you set up secure SSL connection between indexer and MySQL. The details on creating the certificates and setting up MySQL server can be found in MySQL documentation.


mysql_ssl_cert = /etc/ssl/client-cert.pem
mysql_ssl_key = /etc/ssl/client-key.pem
mysql_ssl_ca = /etc/ssl/cacert.pem

12.1.10. odbc_dsn

ODBC DSN to connect to. Mandatory, no default value. Applies to odbc source type only.

ODBC DSN (Data Source Name) specifies the credentials (host, user, password, etc) to use when connecting to ODBC data source. The format depends on specific ODBC driver used.


odbc_dsn = Driver={Oracle ODBC Driver};Dbq=myDBName;Uid=myUsername;Pwd=myPassword

12.1.11. sql_query_pre

Pre-fetch query, or pre-query. Multi-value, optional, default is empty list of queries. Applies to SQL source types (mysql, pgsql, mssql) only.

Multi-value means that you can specify several pre-queries. They are executed before the main fetch query, and they will be executed exactly in order of appearance in the configuration file. Pre-query results are ignored.

Pre-queries are useful in a lot of ways. They are used to setup encoding, mark records that are going to be indexed, update internal counters, set various per-connection SQL server options and variables, and so on.

Perhaps the most frequent pre-query usage is to specify the encoding that the server will use for the rows it returns. Note that Sphinx accepts only UTF-8 texts. Two MySQL specific examples of setting the encoding are:

sql_query_pre = SET CHARACTER_SET_RESULTS=utf8
sql_query_pre = SET NAMES utf8

Also specific to MySQL sources, it is useful to disable query cache (for indexer connection only) in pre-query, because indexing queries are not going to be re-run frequently anyway, and there's no sense in caching their results. That could be achieved with:

sql_query_pre = SET SESSION query_cache_type=OFF


sql_query_pre = SET NAMES utf8
sql_query_pre = SET SESSION query_cache_type=OFF

12.1.12. sql_query

Main document fetch query. Mandatory, no default value. Applies to SQL source types (mysql, pgsql, mssql) only.

There can be only one main query. This is the query which is used to retrieve documents from SQL server. You can specify up to 32 full-text fields (formally, upto SPH_MAX_FIELDS from sphinx.h), and an arbitrary amount of attributes. All of the columns that are neither document ID (the first one) nor attributes will be full-text indexed.

Document ID MUST be the very first field, and it MUST BE UNIQUE UNSIGNED POSITIVE (NON-ZERO, NON-NEGATIVE) INTEGER NUMBER. It can be either 32-bit or 64-bit, depending on how you built Sphinx; by default it builds with 32-bit IDs support but --enable-id64 option to configure allows to build with 64-bit document and word IDs support.


sql_query = \
    SELECT id, group_id, UNIX_TIMESTAMP(date_added) AS date_added, \
        title, content \
    FROM documents

12.1.13. sql_joined_field

Joined/payload field fetch query. Multi-value, optional, default is empty list of queries. Applies to SQL source types (mysql, pgsql, mssql) only.

sql_joined_field lets you use two different features: joined fields, and payloads (payload fields). It's syntax is as follows:

sql_joined_field = FIELD-NAME 'from'  ( 'query' | 'payload-query' \
    | 'ranged-query' ); QUERY [ ; RANGE-QUERY ]


  • FIELD-NAME is a joined/payload field name;

  • QUERY is an SQL query that must fetch values to index.

  • RANGE-QUERY is an optional SQL query that fetches a range of values to index. (Added in version 2.0.1-beta.)

Joined fields let you avoid JOIN and/or GROUP_CONCAT statements in the main document fetch query (sql_query). This can be useful when SQL-side JOIN is slow, or needs to be offloaded on Sphinx side, or simply to emulate MySQL-specific GROUP_CONCAT functionality in case your database server does not support it.

The query must return exactly 2 columns: document ID, and text to append to a joined field. Document IDs can be duplicate, but they must be in ascending order. All the text rows fetched for a given ID will be concatenated together, and the concatenation result will be indexed as the entire contents of a joined field. Rows will be concatenated in the order returned from the query, and separating whitespace will be inserted between them. For instance, if joined field query returns the following rows:

( 1, 'red' )
( 1, 'right' )
( 1, 'hand' )
( 2, 'mysql' )
( 2, 'sphinx' )

then the indexing results would be equivalent to that of adding a new text field with a value of 'red right hand' to document 1 and 'mysql sphinx' to document 2.

Joined fields are only indexed differently. There are no other differences between joined fields and regular text fields.

Starting with 2.0.1-beta, ranged queries can be used when a single query is not efficient enough or does not work because of the database driver limitations. It works similar to the ranged queries in the main indexing loop, see Section 3.8, “Ranged queries”. The range will be queried for and fetched upfront once, then multiple queries with different $start and $end substitutions will be run to fetch the actual data.

Payloads let you create a special field in which, instead of keyword positions, so-called user payloads are stored. Payloads are custom integer values attached to every keyword. They can then be used in search time to affect the ranking.

The payload query must return exactly 3 columns: document ID; keyword; and integer payload value. Document IDs can be duplicate, but they must be in ascending order. Payloads must be unsigned integers within 24-bit range, ie. from 0 to 16777215. For reference, payloads are currently internally stored as in-field keyword positions, but that is not guaranteed and might change in the future.

Currently, the only method to account for payloads is to use SPH_RANK_PROXIMITY_BM25 ranker. On indexes with payload fields, it will automatically switch to a variant that matches keywords in those fields, computes a sum of matched payloads multiplied by field weights, and adds that sum to the final rank.


sql_joined_field = \
    tagstext from query; \
    SELECT docid, CONCAT('tag',tagid) FROM tags ORDER BY docid ASC

sql_joined_field = bigint tag from ranged-query; \
    SELECT id, tag FROM tags WHERE id>=$start AND id<=$end ORDER BY id ASC; \
    SELECT MIN(id), MAX(id) FROM tags

12.1.14. sql_query_range

Range query setup. Optional, default is empty. Applies to SQL source types (mysql, pgsql, mssql) only.

Setting this option enables ranged document fetch queries (see Section 3.8, “Ranged queries”). Ranged queries are useful to avoid notorious MyISAM table locks when indexing lots of data. (They also help with other less notorious issues, such as reduced performance caused by big result sets, or additional resources consumed by InnoDB to serialize big read transactions.)

The query specified in this option must fetch min and max document IDs that will be used as range boundaries. It must return exactly two integer fields, min ID first and max ID second; the field names are ignored.

When ranged queries are enabled, sql_query will be required to contain $start and $end macros (because it obviously would be a mistake to index the whole table many times over). Note that the intervals specified by $start..$end will not overlap, so you should not remove document IDs that are exactly equal to $start or $end from your query. The example in Section 3.8, “Ranged queries”) illustrates that; note how it uses greater-or-equal and less-or-equal comparisons.


sql_query_range = SELECT MIN(id),MAX(id) FROM documents

12.1.15. sql_range_step

Range query step. Optional, default is 1024. Applies to SQL source types (mysql, pgsql, mssql) only.

Only used when ranged queries are enabled. The full document IDs interval fetched by sql_query_range will be walked in this big steps. For example, if min and max IDs fetched are 12 and 3456 respectively, and the step is 1000, indexer will call sql_query several times with the following substitutions:

  • $start=12, $end=1011

  • $start=1012, $end=2011

  • $start=2012, $end=3011

  • $start=3012, $end=3456


sql_range_step = 1000

12.1.16. sql_query_killlist

Kill-list query. Optional, default is empty (no query). Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 0.9.9-rc1.

This query is expected to return a number of 1-column rows, each containing just the document ID. The returned document IDs are stored within an index. Kill-list for a given index suppresses results from other indexes, depending on index order in the query. The intended use is to help implement deletions and updates on existing indexes without rebuilding (actually even touching them), and especially to fight phantom results problem.

Let us dissect an example. Assume we have two indexes, 'main' and 'delta'. Assume that documents 2, 3, and 5 were deleted since last reindex of 'main', and documents 7 and 11 were updated (ie. their text contents were changed). Assume that a keyword 'test' occurred in all these mentioned documents when we were indexing 'main'; still occurs in document 7 as we index 'delta'; but does not occur in document 11 any more. We now reindex delta and then search through both these indexes in proper (least to most recent) order:

$res = $cl->Query ( "test", "main delta" );

First, we need to properly handle deletions. The result set should not contain documents 2, 3, or 5. Second, we also need to avoid phantom results. Unless we do something about it, document 11 will appear in search results! It will be found in 'main' (but not 'delta'). And it will make it to the final result set unless something stops it.

Kill-list, or K-list for short, is that something. Kill-list attached to 'delta' will suppress the specified rows from all the preceding indexes, in this case just 'main'. So to get the expected results, we should put all the updated and deleted document IDs into it.

Note that in the distributed index setup, K-lists are local to every node in the cluster. They are not get transmitted over the network when sending queries. (Because that might be too much of an impact when the K-list is huge.) You will need to setup a separate per-server K-lists in that case.


sql_query_killlist = \
    SELECT id FROM documents WHERE updated_ts>=@last_reindex UNION \
    SELECT id FROM documents_deleted WHERE deleted_ts>=@last_reindex

12.1.17. sql_attr_uint

Unsigned integer attribute declaration. Multi-value (there might be multiple attributes declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only.

The column value should fit into 32-bit unsigned integer range. Values outside this range will be accepted but wrapped around. For instance, -1 will be wrapped around to 2^32-1 or 4,294,967,295.

You can specify bit count for integer attributes by appending ':BITCOUNT' to attribute name (see example below). Attributes with less than default 32-bit size, or bitfields, perform slower. But they require less RAM when using extern storage: such bitfields are packed together in 32-bit chunks in .spa attribute data file. Bit size settings are ignored if using inline storage.


sql_attr_uint = group_id
sql_attr_uint = forum_id:9 # 9 bits for forum_id

12.1.18. sql_attr_bool

Boolean attribute declaration. Multi-value (there might be multiple attributes declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only. Equivalent to sql_attr_uint declaration with a bit count of 1.


sql_attr_bool = is_deleted # will be packed to 1 bit

12.1.19. sql_attr_bigint

64-bit signed integer attribute declaration. Multi-value (there might be multiple attributes declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only. Note that unlike sql_attr_uint, these values are signed. Introduced in version 0.9.9-rc1.


sql_attr_bigint = my_bigint_id

12.1.20. sql_attr_timestamp

UNIX timestamp attribute declaration. Multi-value (there might be multiple attributes declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only.

Timestamps can store date and time in the range of Jan 01, 1970 to Jan 19, 2038 with a precision of one second. The expected column value should be a timestamp in UNIX format, ie. 32-bit unsigned integer number of seconds elapsed since midnight, January 01, 1970, GMT. Timestamps are internally stored and handled as integers everywhere. But in addition to working with timestamps as integers, it's also legal to use them along with different date-based functions, such as time segments sorting mode, or day/week/month/year extraction for GROUP BY.

Note that DATE or DATETIME column types in MySQL can not be directly used as timestamp attributes in Sphinx; you need to explicitly convert such columns using UNIX_TIMESTAMP function (if data is in range).

Note timestamps can not represent dates before January 01, 1970, and UNIX_TIMESTAMP() in MySQL will not return anything expected. If you only needs to work with dates, not times, consider TO_DAYS() function in MySQL instead.


# sql_query = ... UNIX_TIMESTAMP(added_datetime) AS added_ts ...
sql_attr_timestamp = added_ts

12.1.21. sql_attr_float

Floating point attribute declaration. Multi-value (there might be multiple attributes declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only.

The values will be stored in single precision, 32-bit IEEE 754 format. Represented range is approximately from 1e-38 to 1e+38. The amount of decimal digits that can be stored precisely is approximately 7. One important usage of the float attributes is storing latitude and longitude values (in radians), for further usage in query-time geosphere distance calculations.


sql_attr_float = lat_radians
sql_attr_float = long_radians

12.1.22. sql_attr_multi

Multi-valued attribute (MVA) declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only.

Plain attributes only allow to attach 1 value per each document. However, there are cases (such as tags or categories) when it is desired to attach multiple values of the same attribute and be able to apply filtering or grouping to value lists.

The declaration format is as follows (backslashes are for clarity only; everything can be declared in a single line as well):

sql_attr_multi = ATTR-TYPE ATTR-NAME 'from' SOURCE-TYPE \
    [;QUERY] \


  • ATTR-TYPE is 'uint', 'bigint' or 'timestamp'

  • SOURCE-TYPE is 'field', 'query', or 'ranged-query'

  • QUERY is SQL query used to fetch all ( docid, attrvalue ) pairs

  • RANGE-QUERY is SQL query used to fetch min and max ID values, similar to 'sql_query_range'


sql_attr_multi = uint tag from query; SELECT id, tag FROM tags
sql_attr_multi = bigint tag from ranged-query; \
    SELECT id, tag FROM tags WHERE id>=$start AND id<=$end; \
    SELECT MIN(id), MAX(id) FROM tags

12.1.23. sql_attr_string

String attribute declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 1.10-beta.

String attributes can store arbitrary strings attached to every document. There's a fixed size limit of 4 MB per value. Also, searchd will currently cache all the values in RAM, which is an additional implicit limit.

Starting from 2.0.1-beta string attributes can be used for sorting and grouping(ORDER BY, GROUP BY, WITHIN GROUP ORDER BY). Note that attributes declared using sql_attr_string will not be full-text indexed; you can use sql_field_string directive for that.


sql_attr_string = title # will be stored but will not be indexed

12.1.24. sql_attr_json

JSON attribute declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 2.1.1-beta.

When indexing JSON attributes, Sphinx expects a text field with JSON formatted data. As of 2.2.1-beta JSON attributes supports arbitrary JSON data with no limitation in nested levels or types.

    "id": 1,
    "gid": 2,
    "title": "some title",
    "tags": [
			"one": "two",
			"three": [4, 5]

These attributes allow Sphinx to work with documents without a fixed set of attribute columns. When you filter on a key of a JSON attribute, documents that don't include the key will simply be ignored.

You can read more on JSON attributes in http://sphinxsearch.com/blog/2013/08/08/full-json-support-in-trunk/.


sql_attr_json = properties

12.1.25. sql_column_buffers

Per-column buffer sizes. Optional, default is empty (deduce the sizes automatically). Applies to odbc, mssql source types only. Introduced in version 2.0.1-beta.

ODBC and MS SQL drivers sometimes can not return the maximum actual column size to be expected. For instance, NVARCHAR(MAX) columns always report their length as 2147483647 bytes to indexer even though the actually used length is likely considerably less. However, the receiving buffers still need to be allocated upfront, and their sizes have to be determined. When the driver does not report the column length at all, Sphinx allocates default 1 KB buffers for each non-char column, and 1 MB buffers for each char column. Driver-reported column length also gets clamped by an upper limit of 8 MB, so in case the driver reports (almost) a 2 GB column length, it will be clamped and a 8 MB buffer will be allocated instead for that column. These hard-coded limits can be overridden using the sql_column_buffers directive, either in order to save memory on actually shorter columns, or overcome the 8 MB limit on actually longer columns. The directive values must be a comma-separated lists of selected column names and sizes:

sql_column_buffers = <colname>=<size>[K|M] [, ...]


sql_query = SELECT id, mytitle, mycontent FROM documents
sql_column_buffers = mytitle=64K, mycontent=10M

12.1.26. sql_field_string

Combined string attribute and full-text field declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 1.10-beta.

sql_attr_string only stores the column value but does not full-text index it. In some cases it might be desired to both full-text index the column and store it as attribute. sql_field_string lets you do exactly that. Both the field and the attribute will be named the same.


sql_field_string = title # will be both indexed and stored

12.1.27. sql_file_field

File based field declaration. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 1.10-beta.

This directive makes indexer interpret field contents as a file name, and load and index the referred file. Files larger than max_file_field_buffer in size are skipped. Any errors during the file loading (IO errors, missed limits, etc) will be reported as indexing warnings and will not early terminate the indexing. No content will be indexed for such files.


sql_file_field = my_file_path # load and index files referred to by my_file_path

12.1.28. sql_query_post

Post-fetch query. Optional, default value is empty. Applies to SQL source types (mysql, pgsql, mssql) only.

This query is executed immediately after sql_query completes successfully. When post-fetch query produces errors, they are reported as warnings, but indexing is not terminated. It's result set is ignored. Note that indexing is not yet completed at the point when this query gets executed, and further indexing still may fail. Therefore, any permanent updates should not be done from here. For instance, updates on helper table that permanently change the last successfully indexed ID should not be run from post-fetch query; they should be run from post-index query instead.


sql_query_post = DROP TABLE my_tmp_table

12.1.29. sql_query_post_index

Post-index query. Optional, default value is empty. Applies to SQL source types (mysql, pgsql, mssql) only.

This query is executed when indexing is fully and successfully completed. If this query produces errors, they are reported as warnings, but indexing is not terminated. It's result set is ignored. $maxid macro can be used in its text; it will be expanded to maximum document ID which was actually fetched from the database during indexing. If no documents were indexed, $maxid will be expanded to 0.


sql_query_post_index = REPLACE INTO counters ( id, val ) \
    VALUES ( 'max_indexed_id', $maxid )

12.1.30. sql_ranged_throttle

Ranged query throttling period, in milliseconds. Optional, default is 0 (no throttling). Applies to SQL source types (mysql, pgsql, mssql) only.

Throttling can be useful when indexer imposes too much load on the database server. It causes the indexer to sleep for given amount of milliseconds once per each ranged query step. This sleep is unconditional, and is performed before the fetch query.


sql_ranged_throttle = 1000 # sleep for 1 sec before each query step

12.1.31. xmlpipe_command

Shell command that invokes xmlpipe2 stream producer. Mandatory. Applies to xmlpipe2 source types only.

Specifies a command that will be executed and which output will be parsed for documents. Refer to Section 3.9, “xmlpipe2 data source” for specific format description.


xmlpipe_command = cat /home/sphinx/test.xml

12.1.32. xmlpipe_field

xmlpipe field declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Refer to Section 3.9, “xmlpipe2 data source”.


xmlpipe_field = subject
xmlpipe_field = content

12.1.33. xmlpipe_field_string

xmlpipe field and string attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Refer to Section 3.9, “xmlpipe2 data source”. Introduced in version 1.10-beta.

Makes the specified XML element indexed as both a full-text field and a string attribute. Equivalent to <sphinx:field name="field" attr="string"/> declaration within the XML file.


xmlpipe_field_string = subject

12.1.34. xmlpipe_attr_uint

xmlpipe integer attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Syntax fully matches that of sql_attr_uint.


xmlpipe_attr_uint = author_id

12.1.35. xmlpipe_attr_bigint

xmlpipe signed 64-bit integer attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Syntax fully matches that of sql_attr_bigint.


xmlpipe_attr_bigint = my_bigint_id

12.1.36. xmlpipe_attr_bool

xmlpipe boolean attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Syntax fully matches that of sql_attr_bool.


xmlpipe_attr_bool = is_deleted # will be packed to 1 bit

12.1.37. xmlpipe_attr_timestamp

xmlpipe UNIX timestamp attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Syntax fully matches that of sql_attr_timestamp.


xmlpipe_attr_timestamp = published

12.1.38. xmlpipe_attr_float

xmlpipe floating point attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Syntax fully matches that of sql_attr_float.


xmlpipe_attr_float = lat_radians
xmlpipe_attr_float = long_radians

12.1.39. xmlpipe_attr_multi

xmlpipe MVA attribute declaration. Multi-value, optional. Applies to xmlpipe2 source type only.

This setting declares an MVA attribute tag in xmlpipe2 stream. The contents of the specified tag will be parsed and a list of integers that will constitute the MVA will be extracted, similar to how sql_attr_multi parses SQL column contents when 'field' MVA source type is specified.


xmlpipe_attr_multi = taglist

12.1.40. xmlpipe_attr_multi_64

xmlpipe MVA attribute declaration. Declares the BIGINT (signed 64-bit integer) MVA attribute. Multi-value, optional. Applies to xmlpipe2 source type only.

This setting declares an MVA attribute tag in xmlpipe2 stream. The contents of the specified tag will be parsed and a list of integers that will constitute the MVA will be extracted, similar to how sql_attr_multi parses SQL column contents when 'field' MVA source type is specified.


xmlpipe_attr_multi_64 = taglist

12.1.41. xmlpipe_attr_string

xmlpipe string declaration. Multi-value, optional. Applies to xmlpipe2 source type only. Introduced in version 1.10-beta.

This setting declares a string attribute tag in xmlpipe2 stream. The contents of the specified tag will be parsed and stored as a string value.


xmlpipe_attr_string = subject

12.1.42. xmlpipe_attr_json

JSON attribute declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Introduced in version 2.1.1-beta.

This directive is used to declare that the contents of a given XML tag are to be treated as a JSON document and stored into a Sphinx index for later use. Refer to Section 12.1.24, “sql_attr_json” for more details on the JSON attributes.


xmlpipe_attr_json = properties

12.1.43. xmlpipe_fixup_utf8

Perform Sphinx-side UTF-8 validation and filtering to prevent XML parser from choking on non-UTF-8 documents. Optional, default is 0. Applies to xmlpipe2 source type only.

Under certain occasions it might be hard or even impossible to guarantee that the incoming XMLpipe2 document bodies are in perfectly valid and conforming UTF-8 encoding. For instance, documents with national single-byte encodings could sneak into the stream. libexpat XML parser is fragile, meaning that it will stop processing in such cases. UTF8 fixup feature lets you avoid that. When fixup is enabled, Sphinx will preprocess the incoming stream before passing it to the XML parser and replace invalid UTF-8 sequences with spaces.


xmlpipe_fixup_utf8 = 1

12.1.44. mssql_winauth

MS SQL Windows authentication flag. Boolean, optional, default value is 0 (false). Applies to mssql source type only. Introduced in version 0.9.9-rc1.

Whether to use currently logged in Windows account credentials for authentication when connecting to MS SQL Server. Note that when running searchd as a service, account user can differ from the account you used to install the service.


mssql_winauth = 1

12.1.45. unpack_zlib

Columns to unpack using zlib (aka deflate, aka gunzip). Multi-value, optional, default value is empty list of columns. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 0.9.9-rc1.

Columns specified using this directive will be unpacked by indexer using standard zlib algorithm (called deflate and also implemented by gunzip). When indexing on a different box than the database, this lets you offload the database, and save on network traffic. The feature is only available if zlib and zlib-devel were both available during build time.


unpack_zlib = col1
unpack_zlib = col2

12.1.46. unpack_mysqlcompress

Columns to unpack using MySQL UNCOMPRESS() algorithm. Multi-value, optional, default value is empty list of columns. Applies to SQL source types (mysql, pgsql, mssql) only. Introduced in version 0.9.9-rc1.

Columns specified using this directive will be unpacked by indexer using modified zlib algorithm used by MySQL COMPRESS() and UNCOMPRESS() functions. When indexing on a different box than the database, this lets you offload the database, and save on network traffic. The feature is only available if zlib and zlib-devel were both available during build time.


unpack_mysqlcompress = body_compressed
unpack_mysqlcompress = description_compressed

12.1.47. unpack_mysqlcompress_maxsize

Buffer size for UNCOMPRESS()ed data. Optional, default value is 16M. Introduced in version 0.9.9-rc1.

When using unpack_mysqlcompress, due to implementation intricacies it is not possible to deduce the required buffer size from the compressed data. So the buffer must be preallocated in advance, and unpacked data can not go over the buffer size. This option lets you control the buffer size, both to limit indexer memory use, and to enable unpacking of really long data fields if necessary.


unpack_mysqlcompress_maxsize = 1M

12.1.48. csvpipe_delimiter

csvpipe source fields delimiter. Optional, default value is ','. Introduced in version 2.2.1-beta.


csvpipe_delimiter = ;

12.2. Index configuration options

12.2.1. type

Index type. Known values are 'plain', 'distributed', 'rt' and 'template'. Optional, default is 'plain' (plain local index).

Sphinx supports several different types of indexes. Versions 0.9.x supported two index types: plain local indexes that are stored and processed on the local machine; and distributed indexes, that involve not only local searching but querying remote searchd instances over the network as well (see Section 5.8, “Distributed searching”). Version 1.10-beta also adds support for so-called real-time indexes (or RT indexes for short) that are also stored and processed locally, but additionally allow for on-the-fly updates of the full-text index (see Chapter 4, Real-time indexes). Note that attributes can be updated on-the-fly using either plain local indexes or RT ones. In 2.2.1-beta template indexes was introduced. They are actually a pseudo-indexes because they do not store any data. That means they do not create any files on your hard drive. But you can use them for keywords and snippets generation, which may be useful in some cases.

Index type setting lets you choose the needed type. By default, plain local index type will be assumed.


type = distributed

12.2.2. source

Adds document source to local index. Multi-value, mandatory.

Specifies document source to get documents from when the current index is indexed. There must be at least one source. There may be multiple sources, without any restrictions on the source types: ie. you can pull part of the data from MySQL server, part from PostgreSQL, part from the filesystem using xmlpipe2 wrapper.

However, there are some restrictions on the source data. First, document IDs must be globally unique across all sources. If that condition is not met, you might get unexpected search results. Second, source schemas must be the same in order to be stored within the same index.

No source ID is stored automatically. Therefore, in order to be able to tell what source the matched document came from, you will need to store some additional information yourself. Two typical approaches include:

  1. mangling document ID and encoding source ID in it:

    source src1
        sql_query = SELECT id*10+1, ... FROM table1
    source src2
        sql_query = SELECT id*10+2, ... FROM table2

  2. storing source ID simply as an attribute:

    source src1
        sql_query = SELECT id, 1 AS source_id FROM table1
        sql_attr_uint = source_id
    source src2
        sql_query = SELECT id, 2 AS source_id FROM table2
        sql_attr_uint = source_id


source = srcpart1
source = srcpart2
source = srcpart3

12.2.3. path

Index files path and file name (without extension). Mandatory.

Path specifies both directory and file name, but without extension. indexer will append different extensions to this path when generating final names for both permanent and temporary index files. Permanent data files have several different extensions starting with '.sp'; temporary files' extensions start with '.tmp'. It's safe to remove .tmp* files is if indexer fails to remove them automatically.

For reference, different index files store the following data:

  • .spa stores document attributes (used in extern docinfo storage mode only);

  • .spd stores matching document ID lists for each word ID;

  • .sph stores index header information;

  • .spi stores word lists (word IDs and pointers to .spd file);

  • .spk stores kill-lists;

  • .spm stores MVA data;

  • .spp stores hit (aka posting, aka word occurrence) lists for each word ID;

  • .sps stores string attribute data.

  • .spe stores skip-lists to speed up doc-list filtering


path = /var/data/test1

12.2.4. docinfo

Document attribute values (docinfo) storage mode. Optional, default is 'extern'. Known values are 'none', 'extern' and 'inline'.

Docinfo storage mode defines how exactly docinfo will be physically stored on disk and RAM. "none" means that there will be no docinfo at all (ie. no attributes). Normally you need not to set "none" explicitly because Sphinx will automatically select "none" when there are no attributes configured. "inline" means that the docinfo will be stored in the .spd file, along with the document ID lists. "extern" means that the docinfo will be stored separately (externally) from document ID lists, in a special .spa file.

Basically, externally stored docinfo must be kept in RAM when querying. for performance reasons. So in some cases "inline" might be the only option. However, such cases are infrequent, and docinfo defaults to "extern". Refer to Section 3.3, “Attributes” for in-depth discussion and RAM usage estimates.


docinfo = inline

12.2.5. mlock

Memory locking for cached data. Optional, default is 0 (do not call mlock()).

For search performance, searchd preloads a copy of .spa and .spi files in RAM, and keeps that copy in RAM at all times. But if there are no searches on the index for some time, there are no accesses to that cached copy, and OS might decide to swap it out to disk. First queries to such "cooled down" index will cause swap-in and their latency will suffer.

Setting mlock option to 1 makes Sphinx lock physical RAM used for that cached data using mlock(2) system call, and that prevents swapping (see man 2 mlock for details). mlock(2) is a privileged call, so it will require searchd to be either run from root account, or be granted enough privileges otherwise. If mlock() fails, a warning is emitted, but index continues working.


mlock = 1

12.2.6. morphology

A list of morphology preprocessors (stemmers or lemmatizers) to apply. Optional, default is empty (do not apply any preprocessor).

Morphology preprocessors can be applied to the words being indexed to replace different forms of the same word with the base, normalized form. For instance, English stemmer will normalize both "dogs" and "dog" to "dog", making search results for both searches the same.

There are 3 different morphology preprocessors that Sphinx implements: lemmatizers, stemmers, and phonetic algorithms.

  • Lemmatizer reduces a keyword form to a so-called lemma, a proper normal form, or in other words, a valid natural language root word. For example, "running" could be reduced to "run", the infinitive verb form, and "octopi" would be reduced to "octopus", the singular noun form. Note that sometimes a word form can have multiple corresponding root words. For instance, by looking at "dove" it is not possible to tell whether this is a past tense of "dive" the verb as in "He dove into a pool.", or "dove" the noun as in "White dove flew over the cuckoo's nest." In this case lemmatizer can generate all the possible root forms.

  • Stemmer reduces a keyword form to a so-called stem by removing and/or replacing certain well-known suffixes. The resulting stem is however notguaranteed to be a valid word on itself. For instance, with a Porter English stemmers "running" would still reduce to "run", which is fine, but "business" would reduce to "busi", which is not a word, and "octopi" would not reduce at all. Stemmers are essentially (much) simpler but still pretty good replacements of full-blown lemmatizers.

  • Phonetic algorithms replace the words with specially crafted phonetic codes that are equal even when the words original are different, but phonetically close.

The morphology processors that come with our own built-in Sphinx implementations are:

  • English, Russian, and German lemmatizers;

  • English, Russian, Arabic, and Czech stemmers;

  • SoundEx and MetaPhone phonetic algorithms.

You can also link with libstemmer library for even more stemmers (see details below). With libstemmer, Sphinx also supports morphological processing for more than 15 other languages. Binary packages should come prebuilt with libstemmer support, too.

Lemmatizer support was added in version 2.1.1-beta, starting with a Russian lemmatizer. English and German lemmatizers were then added in version 2.2.1-beta.

Lemmatizers require a dictionary that needs to be additionally downloaded from the Sphinx website. That dictionary needs to be installed in a directory specified by lemmatizer_base directive. Also, there is a lemmatizer_cache directive that lets you speed up lemmatizing (and therefore indexing) by spending more RAM for, basically, an uncompressed cache of a dictionary.

Chinese segmentation using Rosette Linguistics Platform was added in 2.2.1-beta. It is a much more precise but slower way (compared to n-grams) to segment Chinese documents. charset_table must contain all Chinese characters except Chinese punctuation marks because incoming documents are first processed by sphinx tokenizer and then the result is processed by RLP. Sphinx performs per-token language detection on the incoming documents. If token language is identified as Chinese, it will only be processed the RLP, even if multiple morphology processors are specified. Otherwise, it will be processed by all the morphology processors specified in the "morphology" option. Rosette Linguistics Platform must be installed and configured and sphinx must be built with a --with-rlp switch. See also rlp_root, rlp_environment and rlp_context options. A batched version of RLP segmentation is also available (rlp_chinese_batched). It provides the same functionality as the basic rlp_chinese segmentation, but enables batching documents before processing them by the RLP. Processing several documents at once can result in a substantial indexing speedup if the documents are small (for example, less than 1k). See also rlp_max_batch_size and rlp_max_batch_docs options.

Additional stemmers provided by Snowball project libstemmer library can be enabled at compile time using --with-libstemmer configure option. Built-in English and Russian stemmers should be faster than their libstemmer counterparts, but can produce slightly different results, because they are based on an older version.

Soundex implementation matches that of MySQL. Metaphone implementation is based on Double Metaphone algorithm and indexes the primary code.

Built-in values that are available for use in morphology option are as follows:

  • none - do not perform any morphology processing;

  • lemmatize_ru - apply Russian lemmatizer and pick a single root form (added in 2.1.1-beta);

  • lemmatize_en - apply English lemmatizer and pick a single root form (added in 2.2.1-beta);

  • lemmatize_de - apply German lemmatizer and pick a single root form (added in 2.2.1-beta);

  • lemmatize_ru_all - apply Russian lemmatizer and index all possible root forms (added in 2.1.1-beta);

  • lemmatize_en_all - apply English lemmatizer and index all possible root forms (added in 2.2.1-beta);

  • lemmatize_de_all - apply German lemmatizer and index all possible root forms (added in 2.2.1-beta);

  • stem_en - apply Porter's English stemmer;

  • stem_ru - apply Porter's Russian stemmer;

  • stem_enru - apply Porter's English and Russian stemmers;

  • stem_cz - apply Czech stemmer;

  • stem_ar - apply Arabic stemmer (added in 2.1.1-beta);

  • soundex - replace keywords with their SOUNDEX code;

  • metaphone - replace keywords with their METAPHONE code.

  • rlp_chinese - apply Chinese text segmentation using Rosette Linguistics Platform

  • rlp_chinese_batched - apply Chinese text segmentation using Rosette Linguistics Platform with document batching

Additional values provided by libstemmer are in 'libstemmer_XXX' format, where XXX is libstemmer algorithm codename (refer to libstemmer_c/libstemmer/modules.txt for a complete list).

Several stemmers can be specified (comma-separated). They will be applied to incoming words in the order they are listed, and the processing will stop once one of the stemmers actually modifies the word. Also when wordforms feature is enabled the word will be looked up in word forms dictionary first, and if there is a matching entry in the dictionary, stemmers will not be applied at all. Or in other words, wordforms can be used to implement stemming exceptions.


morphology = stem_en, libstemmer_sv

12.2.7. dict

The keywords dictionary type. Known values are 'crc' and 'keywords'. 'crc' is DEPRECATED. Use 'keywords' instead. Optional, default is 'keywords'. Introduced in version 2.0.1-beta.

CRC dictionary mode (dict=crc) is the default dictionary type in Sphinx, and the only one available until version 2.0.1-beta. Keywords dictionary mode (dict=keywords) was added in 2.0.1-beta, primarily to (greatly) reduce indexing impact and enable substring searches on huge collections. They also eliminate the chance of CRC32 collisions. In 2.0.1-beta, that mode was only supported for disk indexes. Starting with 2.0.2-beta, RT indexes are also supported.

CRC dictionaries never store the original keyword text in the index. Instead, keywords are replaced with their control sum value (either CRC32 or FNV64, depending whether Sphinx was built with --enable-id64) both when searching and indexing, and that value is used internally in the index.

That approach has two drawbacks. First, in CRC32 case there is a chance of control sum collision between several pairs of different keywords, growing quadratically with the number of unique keywords in the index. (FNV64 case is unaffected in practice, as a chance of a single FNV64 collision in a dictionary of 1 billion entries is approximately 1:16, or 6.25 percent. And most dictionaries will be much more compact that a billion keywords, as a typical spoken human language has in the region of 1 to 10 million word forms.) Second, and more importantly, substring searches are not directly possible with control sums. Sphinx alleviated that by pre-indexing all the possible substrings as separate keywords (see Section 12.2.18, “min_prefix_len”, Section 12.2.19, “min_infix_len” directives). That actually has an added benefit of matching substrings in the quickest way possible. But at the same time pre-indexing all substrings grows the index size a lot (factors of 3-10x and even more would not be unusual) and impacts the indexing time respectively, rendering substring searches on big indexes rather impractical.

Keywords dictionary, introduced in 2.0.1-beta, fixes both these drawbacks. It stores the keywords in the index and performs search-time wildcard expansion. For example, a search for a 'test*' prefix could internally expand to 'test|tests|testing' query based on the dictionary contents. That expansion is fully transparent to the application, except that the separate per-keyword statistics for all the actually matched keywords would now also be reported.

Version 2.1.1-beta introduced extended wildcards support, now special symbols like '?' and '%' are supported along with substring (infix) search (e.g. "t?st*", "run%", "*abc*"). Note, however, these wildcards work only with dict=keywords, and not elsewhere.

Indexing with keywords dictionary should be 1.1x to 1.3x slower compared to regular, non-substring indexing - but times faster compared to substring indexing (either prefix or infix). Index size should only be slightly bigger that than of the regular non-substring index, with a 1..10% percent total difference. Regular keyword searching time must be very close or identical across all three discussed index kinds (CRC non-substring, CRC substring, keywords). Substring searching time can vary greatly depending on how many actual keywords match the given substring (in other words, into how many keywords does the search term expand). The maximum number of keywords matched is restricted by the expansion_limit directive.

Essentially, keywords and CRC dictionaries represent the two different trade-off substring searching decisions. You can choose to either sacrifice indexing time and index size in favor of top-speed worst-case searches (CRC dictionary), or only slightly impact indexing time but sacrifice worst-case searching time when the prefix expands into very many keywords (keywords dictionary).


dict = keywords

12.2.8. index_sp

Whether to detect and index sentence and paragraph boundaries. Optional, default is 0 (do not detect and index). Introduced in version 2.0.1-beta.

This directive enables sentence and paragraph boundary indexing. It's required for the SENTENCE and PARAGRAPH operators to work. Sentence boundary detection is based on plain text analysis, so you only need to set index_sp = 1 to enable it. Paragraph detection is however based on HTML markup, and happens in the HTML stripper. So to index paragraph locations you also need to enable the stripper by specifying html_strip = 1. Both types of boundaries are detected based on a few built-in rules enumerated just below.

Sentence boundary detection rules are as follows.

  • Question and exclamation signs (? and !) are always a sentence boundary.

  • Trailing dot (.) is a sentence boundary, except:

    • When followed by a letter. That's considered a part of an abbreviation (as in "S.T.A.L.K.E.R" or "Goldman Sachs S.p.A.").

    • When followed by a comma. That's considered an abbreviation followed by a comma (as in "Telecom Italia S.p.A., founded in 1994").

    • When followed by a space and a small letter. That's considered an abbreviation within a sentence (as in "News Corp. announced in February").

    • When preceded by a space and a capital letter, and followed by a space. That's considered a middle initial (as in "John D. Doe").

Paragraph boundaries are inserted at every block-level HTML tag. Namely, those are (as taken from HTML 4 standard) ADDRESS, BLOCKQUOTE, CAPTION, CENTER, DD, DIV, DL, DT, H1, H2, H3, H4, H5, LI, MENU, OL, P, PRE, TABLE, TBODY, TD, TFOOT, TH, THEAD, TR, and UL.

Both sentences and paragraphs increment the keyword position counter by 1.


index_sp = 1

12.2.9. index_zones

A list of in-field HTML/XML zones to index. Optional, default is empty (do not index zones). Introduced in version 2.0.1-beta.

Zones can be formally defined as follows. Everything between an opening and a matching closing tag is called a span, and the aggregate of all spans corresponding sharing the same tag name is called a zone. For instance, everything between the occurrences of <H1> and </H1> in the document field belongs to H1 zone.

Zone indexing, enabled by index_zones directive, is an optional extension of the HTML stripper. So it will also require that the stripper is enabled (with html_strip = 1). The value of the index_zones should be a comma-separated list of those tag names and wildcards (ending with a star) that should be indexed as zones.

Zones can nest and overlap arbitrarily. The only requirement is that every opening tag has a matching tag. You can also have an arbitrary number of both zones (as in unique zone names, such as H1) and spans (all the occurrences of those H1 tags) in a document. Once indexed, zones can then be used for matching with the ZONE operator, see Section 5.3, “Extended query syntax”.


index_zones = h*, th, title

Earlier versions than 2.1.1-beta only provided this feature for plain index files; currently, RT index files also provide it.

12.2.10. min_stemming_len

Minimum word length at which to enable stemming. Optional, default is 1 (stem everything). Introduced in version 0.9.9-rc1.

Stemmers are not perfect, and might sometimes produce undesired results. For instance, running "gps" keyword through Porter stemmer for English results in "gp", which is not really the intent. min_stemming_len feature lets you suppress stemming based on the source word length, ie. to avoid stemming too short words. Keywords that are shorter than the given threshold will not be stemmed. Note that keywords that are exactly as long as specified will be stemmed. So in order to avoid stemming 3-character keywords, you should specify 4 for the value. For more finely grained control, refer to wordforms feature.


min_stemming_len = 4

12.2.11. stopwords

Stopword files list (space separated). Optional, default is empty.

Stopwords are the words that will not be indexed. Typically you'd put most frequent words in the stopwords list because they do not add much value to search results but consume a lot of resources to process.

You can specify several file names, separated by spaces. All the files will be loaded. Stopwords file format is simple plain text. The encoding must be UTF-8. File data will be tokenized with respect to charset_table settings, so you can use the same separators as in the indexed data.

The stemmers will normally be applied when parsing stopwords file. That might however lead to undesired results. Starting with 2.1.1-beta, you can turn that off with stopwords_unstemmed.

Starting with version 2.1.1-beta small enough files are stored in the index header, see Section 12.2.13, “embedded_limit” for details.

While stopwords are not indexed, they still do affect the keyword positions. For instance, assume that "the" is a stopword, that document 1 contains the line "in office", and that document 2 contains "in the office". Searching for "in office" as for exact phrase will only return the first document, as expected, even though "the" in the second one is stopped. That behavior can be tweaked through the stopword_step directive.

Stopwords files can either be created manually, or semi-automatically. indexer provides a mode that creates a frequency dictionary of the index, sorted by the keyword frequency, see --buildstops and --buildfreqs switch in Section 7.1, “indexer command reference”. Top keywords from that dictionary can usually be used as stopwords.


stopwords = /usr/local/sphinx/data/stopwords.txt
stopwords = stopwords-ru.txt stopwords-en.txt

12.2.12. wordforms

Word forms dictionary. Optional, default is empty.

Word forms are applied after tokenizing the incoming text by charset_table rules. They essentially let you replace one word with another. Normally, that would be used to bring different word forms to a single normal form (eg. to normalize all the variants such as "walks", "walked", "walking" to the normal form "walk"). It can also be used to implement stemming exceptions, because stemming is not applied to words found in the forms list.

Starting with version 2.1.1-beta small enough files are stored in the index header, see Section 12.2.13, “embedded_limit” for details.

Dictionaries are used to normalize incoming words both during indexing and searching. Therefore, to pick up changes in wordforms file it's required to rotate index.

Word forms support in Sphinx is designed to support big dictionaries well. They moderately affect indexing speed: for instance, a dictionary with 1 million entries slows down indexing about 1.5 times. Searching speed is not affected at all. Additional RAM impact is roughly equal to the dictionary file size, and dictionaries are shared across indexes: ie. if the very same 50 MB wordforms file is specified for 10 different indexes, additional searchd RAM usage will be about 50 MB.

Dictionary file should be in a simple plain text format. Each line should contain source and destination word forms, in UTF-8 encoding, separated by "greater" sign. Rules from the charset_table will be applied when the file is loaded. So basically it's as case sensitive as your other full-text indexed data, ie. typically case insensitive. Here's the file contents sample:

walks > walk
walked > walk
walking > walk

There is a bundled spelldump utility that helps you create a dictionary file in the format Sphinx can read from source .dict and .aff dictionary files in ispell or MySpell format (as bundled with OpenOffice).

Starting with version 0.9.9-rc1, you can map several source words to a single destination word. Because the work happens on tokens, not the source text, differences in whitespace and markup are ignored.

Starting with version 2.1.1-beta, you can use "=>" instead of ">". Comments (starting with "#" are also allowed. Finally, if a line starts with a tilde ("~") the wordform will be applied after morphology, instead of before.

core 2 duo > c2d
e6600 > c2d
core 2duo => c2d # Some people write '2duo' together...

Stating with version 2.2.4, you can specify multiple destination tokens:

s02e02 > season 2 episode 2
s3 e3 > season 3 episode 3


wordforms = /usr/local/sphinx/data/wordforms.txt
wordforms = /usr/local/sphinx/data/alternateforms.txt
wordforms = /usr/local/sphinx/private/dict*.txt

Starting with version 2.1.1-beta you can specify several files and not only just one. Masks can be used as a pattern, and all matching files will be processed in simple ascending order. (If multi-byte codepages are used, and file names can include foreign characters, the resulting order may not be exactly alphabetic.) If a same wordform definition is found in several files, the latter one is used, and it overrides previous definitions.

12.2.13. embedded_limit

Embedded exceptions, wordforms, or stopwords file size limit. Optional, default is 16K. Added in version 2.1.1-beta.

Before 2.1.1-beta, the contents of exceptions, wordforms, or stopwords files were always kept in the files. Only the file names were stored into the index. Starting with 2.1.1-beta, indexer can either save the file name, or embed the file contents directly into the index. Files sized under embedded_limit get stored into the index. For bigger files, only the file names are stored. This also simplifies moving index files to a different machine; you may get by just copying a single file.

With smaller files, such embedding reduces the number of the external files on which the index depends, and helps maintenance. But at the same time it makes no sense to embed a 100 MB wordforms dictionary into a tiny delta index. So there needs to be a size threshold, and embedded_limit is that threshold.


embedded_limit = 32K

12.2.14. exceptions

Tokenizing exceptions file. Optional, default is empty.

Exceptions allow to map one or more tokens (including tokens with characters that would normally be excluded) to a single keyword. They are similar to wordforms in that they also perform mapping, but have a number of important differences.

Starting with version 2.1.1-beta small enough files are stored in the index header, see Section 12.2.13, “embedded_limit” for details.

Short summary of the differences is as follows:

  • exceptions are case sensitive, wordforms are not;

  • exceptions can use special characters that are not in charset_table, wordforms fully obey charset_table;

  • exceptions can underperform on huge dictionaries, wordforms handle millions of entries well.

The expected file format is also plain text, with one line per exception, and the line format is as follows:

map-from-tokens => map-to-token

Example file:

at & t => at&t
AT&T => AT&T
Standarten   Fuehrer => standartenfuhrer
Standarten Fuhrer => standartenfuhrer
MS Windows => ms windows
Microsoft Windows => ms windows
C++ => cplusplus
c++ => cplusplus
C plus plus => cplusplus

All tokens here are case sensitive: they will not be processed by charset_table rules. Thus, with the example exceptions file above, "at&t" text will be tokenized as two keywords "at" and "t", because of lowercase letters. On the other hand, "AT&T" will match exactly and produce single "AT&T" keyword.

Note that this map-to keyword is a) always interpreted as a single word, and b) is both case and space sensitive! In our sample, "ms windows" query will not match the document with "MS Windows" text. The query will be interpreted as a query for two keywords, "ms" and "windows". And what "MS Windows" gets mapped to is a single keyword "ms windows", with a space in the middle. On the other hand, "standartenfuhrer" will retrieve documents with "Standarten Fuhrer" or "Standarten Fuehrer" contents (capitalized exactly like this), or any capitalization variant of the keyword itself, eg. "staNdarTenfUhreR". (It won't catch "standarten fuhrer", however: this text does not match any of the listed exceptions because of case sensitivity, and gets indexed as two separate keywords.)

Whitespace in the map-from tokens list matters, but its amount does not. Any amount of the whitespace in the map-form list will match any other amount of whitespace in the indexed document or query. For instance, "AT & T" map-from token will match "AT    &  T" text, whatever the amount of space in both map-from part and the indexed text. Such text will therefore be indexed as a special "AT&T" keyword, thanks to the very first entry from the sample.

Exceptions also allow to capture special characters (that are exceptions from general charset_table rules; hence the name). Assume that you generally do not want to treat '+' as a valid character, but still want to be able search for some exceptions from this rule such as 'C++'. The sample above will do just that, totally independent of what characters are in the table and what are not.

Exceptions are applied to raw incoming document and query data during indexing and searching respectively. Therefore, to pick up changes in the file it's required to reindex and restart searchd.


exceptions = /usr/local/sphinx/data/exceptions.txt

12.2.15. min_word_len

Minimum indexed word length. Optional, default is 1 (index everything).

Only those words that are not shorter than this minimum will be indexed. For instance, if min_word_len is 4, then 'the' won't be indexed, but 'they' will be.


min_word_len = 4

12.2.16. charset_table

Accepted characters table, with case folding rules. Optional, default value are latin and cyrillic characters.

charset_table is the main workhorse of Sphinx tokenizing process, ie. the process of extracting keywords from document text or query text. It controls what characters are accepted as valid and what are not, and how the accepted characters should be transformed (eg. should the case be removed or not).

You can think of charset_table as of a big table that has a mapping for each and every of 100K+ characters in Unicode. By default, every character maps to 0, which means that it does not occur within keywords and should be treated as a separator. Once mentioned in the table, character is mapped to some other character (most frequently, either to itself or to a lowercase letter), and is treated as a valid keyword part.

The expected value format is a commas-separated list of mappings. Two simplest mappings simply declare a character as valid, and map a single character to another single character, respectively. But specifying the whole table in such form would result in bloated and barely manageable specifications. So there are several syntax shortcuts that let you map ranges of characters at once. The complete list is as follows:


Single char mapping, declares source char 'A' as allowed to occur within keywords and maps it to destination char 'a' (but does not declare 'a' as allowed).


Range mapping, declares all chars in source range as allowed and maps them to the destination range. Does not declare destination range as allowed. Also checks ranges' lengths (the lengths must be equal).


Stray char mapping, declares a character as allowed and maps it to itself. Equivalent to a->a single char mapping.


Stray range mapping, declares all characters in range as allowed and maps them to themselves. Equivalent to a..z->a..z range mapping.


Checkerboard range map. Maps every pair of chars to the second char. More formally, declares odd characters in range as allowed and maps them to the even ones; also declares even characters as allowed and maps them to themselves. For instance, A..Z/2 is equivalent to A->B, B->B, C->D, D->D, ..., Y->Z, Z->Z. This mapping shortcut is helpful for a number of Unicode blocks where uppercase and lowercase letters go in such interleaved order instead of contiguous chunks.

Control characters with codes from 0 to 31 are always treated as separators. Characters with codes 32 to 127, ie. 7-bit ASCII characters, can be used in the mappings as is. To avoid configuration file encoding issues, 8-bit ASCII characters and Unicode characters must be specified in U+xxx form, where 'xxx' is hexadecimal codepoint number. This form can also be used for 7-bit ASCII characters to encode special ones: eg. use U+20 to encode space, U+2E to encode dot, U+2C to encode comma.

Starting with 2.2.3-beta, aliases "english" and "russian" are allowed at control character mapping.


# default are English and Russian letters
charset_table = 0..9, A..Z->a..z, _, a..z, \
    U+410..U+42F->U+430..U+44F, U+430..U+44F, U+401->U+451, U+451
# english charset defined with alias
charset_table = 0..9, english, _

12.2.17. ignore_chars

Ignored characters list. Optional, default is empty.

Useful in the cases when some characters, such as soft hyphenation mark (U+00AD), should be not just treated as separators but rather fully ignored. For example, if '-' is simply not in the charset_table, "abc-def" text will be indexed as "abc" and "def" keywords. On the contrary, if '-' is added to ignore_chars list, the same text will be indexed as a single "abcdef" keyword.

The syntax is the same as for charset_table, but it's only allowed to declare characters, and not allowed to map them. Also, the ignored characters must not be present in charset_table.


ignore_chars = U+AD

12.2.18. min_prefix_len

Minimum word prefix length to index. Optional, default is 0 (do not index prefixes).

Prefix indexing allows to implement wildcard searching by 'wordstart*' wildcards. When mininum prefix length is set to a positive number, indexer will index all the possible keyword prefixes (ie. word beginnings) in addition to the keywords themselves. Too short prefixes (below the minimum allowed length) will not be indexed.

For instance, indexing a keyword "example" with min_prefix_len=3 will result in indexing "exa", "exam", "examp", "exampl" prefixes along with the word itself. Searches against such index for "exam" will match documents that contain "example" word, even if they do not contain "exam" on itself. However, indexing prefixes will make the index grow significantly (because of many more indexed keywords), and will degrade both indexing and searching times.

There's no automatic way to rank perfect word matches higher in a prefix index, but there's a number of tricks to achieve that. First, you can setup two indexes, one with prefix indexing and one without it, search through both, and use SetIndexWeights() call to combine weights. Second, you can rewriteyour extended-mode queries:

$cl->Query ( "( keyword | keyword* ) other keywords" );


min_prefix_len = 3

12.2.19. min_infix_len

Minimum infix prefix length to index. Optional, default is 0 (do not index infixes).

Infix indexing allows to implement wildcard searching by 'start*', '*end', and '*middle*' wildcards. When minimum infix length is set to a positive number, indexer will index all the possible keyword infixes (ie. substrings) in addition to the keywords themselves. Too short infixes (below the minimum allowed length) will not be indexed. For instance, indexing a keyword "test" with min_infix_len=2 will result in indexing "te", "es", "st", "tes", "est" infixes along with the word itself. Searches against such index for "es" will match documents that contain "test" word, even if they do not contain "es" on itself. However, indexing infixes will make the index grow significantly (because of many more indexed keywords), and will degrade both indexing and searching times.

There's no automatic way to rank perfect word matches higher in an infix index, but the same tricks as with prefix indexes can be applied.


min_infix_len = 3

12.2.20. max_substring_len

Maximum substring (either prefix or infix) length to index. Optional, default is 0 (do not limit indexed substrings). Applies to dict=crc only.

By default, substring (either prefix or infix) indexing in the dict=crc mode will index all the possible substrings as separate keywords. That might result in an overly large index. So the max_substring_len directive lets you limit the impact of substring indexing by skipping too-long substrings (which, chances are, will never get searched for anyway).

For example, a test index of 10,000 blog posts takes this much disk space depending on the settings:

  • 6.4 MB baseline (no substrings)
  • 24.3 MB (3.8x) with min_prefix_len = 3
  • 22.2 MB (3.5x) with min_prefix_len = 3, max_substring_len = 8
  • 19.3 MB (3.0x) with min_prefix_len = 3, max_substring_len = 6
  • 94.3 MB (14.7x) with min_infix_len = 3
  • 84.6 MB (13.2x) with min_infix_len = 3, max_substring_len = 8
  • 70.7 MB (11.0x) with min_infix_len = 3, max_substring_len = 6

So in this test limiting the max substring length saved us 10-15% on the index size.

There is no performance impact associated with substring length when using dict=keywords mode, so this directive is not applicable and intentionally forbidden in that case. If required, you can still limit the length of a substring that you search for in the application code.


max_substring_len = 12

12.2.21. prefix_fields

The list of full-text fields to limit prefix indexing to. Optional, default is empty (index all fields in prefix mode).

Because prefix indexing impacts both indexing and searching performance, it might be desired to limit it to specific full-text fields only: for instance, to provide prefix searching through URLs, but not through page contents. prefix_fields specifies what fields will be prefix-indexed; all other fields will be indexed in normal mode. The value format is a comma-separated list of field names.


prefix_fields = url, domain

12.2.22. infix_fields

The list of full-text fields to limit infix indexing to. Optional, default is empty (index all fields in infix mode).

Similar to prefix_fields, but lets you limit infix-indexing to given fields.


infix_fields = url, domain

12.2.23. ngram_len

N-gram lengths for N-gram indexing. Optional, default is 0 (disable n-gram indexing). Known values are 0 and 1 (other lengths to be implemented).

N-grams provide basic CJK (Chinese, Japanese, Korean) support for unsegmented texts. The issue with CJK searching is that there could be no clear separators between the words. Ideally, the texts would be filtered through a special program called segmenter that would insert separators in proper locations. However, segmenters are slow and error prone, and it's common to index contiguous groups of N characters, or n-grams, instead.

When this feature is enabled, streams of CJK characters are indexed as N-grams. For example, if incoming text is "ABCDEF" (where A to F represent some CJK characters) and length is 1, in will be indexed as if it was "A B C D E F". (With length equal to 2, it would produce "AB BC CD DE EF"; but only 1 is supported at the moment.) Only those characters that are listed in ngram_chars table will be split this way; other ones will not be affected.

Note that if search query is segmented, ie. there are separators between individual words, then wrapping the words in quotes and using extended mode will result in proper matches being found even if the text was not segmented. For instance, assume that the original query is BC DEF. After wrapping in quotes on the application side, it should look like "BC" "DEF" (with quotes). This query will be passed to Sphinx and internally split into 1-grams too, resulting in "B C" "D E F" query, still with quotes that are the phrase matching operator. And it will match the text even though there were no separators in the text.

Even if the search query is not segmented, Sphinx should still produce good results, thanks to phrase based ranking: it will pull closer phrase matches (which in case of N-gram CJK words can mean closer multi-character word matches) to the top.


ngram_len = 1

12.2.24. ngram_chars

N-gram characters list. Optional, default is empty.

To be used in conjunction with in ngram_len, this list defines characters, sequences of which are subject to N-gram extraction. Words comprised of other characters will not be affected by N-gram indexing feature. The value format is identical to charset_table.


ngram_chars = U+3000..U+2FA1F

12.2.25. phrase_boundary

Phrase boundary characters list. Optional, default is empty.

This list controls what characters will be treated as phrase boundaries, in order to adjust word positions and enable phrase-level search emulation through proximity search. The syntax is similar to charset_table. Mappings are not allowed and the boundary characters must not overlap with anything else.

On phrase boundary, additional word position increment (specified by phrase_boundary_step) will be added to current word position. This enables phrase-level searching through proximity queries: words in different phrases will be guaranteed to be more than phrase_boundary_step distance away from each other; so proximity search within that distance will be equivalent to phrase-level search.

Phrase boundary condition will be raised if and only if such character is followed by a separator; this is to avoid abbreviations such as S.T.A.L.K.E.R or URLs being treated as several phrases.


phrase_boundary = ., ?, !, U+2026 # horizontal ellipsis

12.2.26. phrase_boundary_step

Phrase boundary word position increment. Optional, default is 0.

On phrase boundary, current word position will be additionally incremented by this number. See phrase_boundary for details.


phrase_boundary_step = 100

12.2.27. html_strip

Whether to strip HTML markup from incoming full-text data. Optional, default is 0. Known values are 0 (disable stripping) and 1 (enable stripping).

Both HTML tags and entities and considered markup and get processed.

HTML tags are removed, their contents (i.e., everything between <P> and </P>) are left intact by default. You can choose to keep and index attributes of the tags (e.g., HREF attribute in an A tag, or ALT in an IMG one). Several well-known inline tags are completely removed, all other tags are treated as block level and replaced with whitespace. For example, 'te<B>st</B>' text will be indexed as a single keyword 'test', however, 'te<P>st</P>' will be indexed as two keywords 'te' and 'st'. Known inline tags are as follows: A, B, I, S, U, BASEFONT, BIG, EM, FONT, IMG, LABEL, SMALL, SPAN, STRIKE, STRONG, SUB, SUP, TT.

HTML entities get decoded and replaced with corresponding UTF-8 characters. Stripper supports both numeric forms (such as &#239;) and text forms (such as &oacute; or &nbsp;). All entities as specified by HTML4 standard are supported.

Stripping should work with properly formed HTML and XHTML, but, just as most browsers, may produce unexpected results on malformed input (such as HTML with stray <'s or unclosed >'s).

Only the tags themselves, and also HTML comments, are stripped. To strip the contents of the tags too (eg. to strip embedded scripts), see html_remove_elements option. There are no restrictions on tag names; ie. everything that looks like a valid tag start, or end, or a comment will be stripped.


html_strip = 1

12.2.28. html_index_attrs

A list of markup attributes to index when stripping HTML. Optional, default is empty (do not index markup attributes).

Specifies HTML markup attributes whose contents should be retained and indexed even though other HTML markup is stripped. The format is per-tag enumeration of indexable attributes, as shown in the example below.


html_index_attrs = img=alt,title; a=title;

12.2.29. html_remove_elements

A list of HTML elements for which to strip contents along with the elements themselves. Optional, default is empty string (do not strip contents of any elements).

This feature allows to strip element contents, ie. everything that is between the opening and the closing tags. It is useful to remove embedded scripts, CSS, etc. Short tag form for empty elements (ie. <br />) is properly supported; ie. the text that follows such tag will not be removed.

The value is a comma-separated list of element (tag) names whose contents should be removed. Tag names are case insensitive.


html_remove_elements = style, script

12.2.30. local

Local index declaration in the distributed index. Multi-value, optional, default is empty.

This setting is used to declare local indexes that will be searched when given distributed index is searched. Many local indexes can be declared per each distributed index. Any local index can also be mentioned several times in different distributed indexes.

Note that by default all local indexes will be searched sequentially, utilizing only 1 CPU or core. To parallelize processing of the local parts in the distributed index, you should use dist_threads directive, see Section 12.4.24, “dist_threads”.

Before dist_threads, there also was a legacy solution to configure searchd to query itself instead of using local indexes (refer to Section 12.2.31, “agent” for the details). However, that creates redundant CPU and network load, and dist_threads is now strongly suggested instead.


local = chunk1
local = chunk2

12.2.31. agent

Remote agent declaration in the distributed index. Multi-value, optional, default is empty.

agent directive declares remote agents that are searched every time when the enclosing distributed index is searched. The agents are, essentially, pointers to networked indexes. Prior to version 2.1.1-beta, the value format was:

agent = address:index-list

Starting with 2.1.1-beta, the value can additionally specify multiple alternatives (agent mirrors) for either the address only, or the address and index list:

agent = address1 [ | address2 [...] ]:index-list
agent = address1:index-list [ | address2:index-list [...] ]

In both cases the address specification must be one of the following:

address = hostname:port # eg. server2:9312
address = /absolute/unix/socket/path # eg. /var/run/sphinx2.sock

Where hostname is the remote host name, port is the remote TCP port number, index-list is a comma-separated list of index names, and square braces [] designate an optional clause.

In other words, you can point every single agent to one or more remote indexes, residing on one or more networked servers. There are absolutely no restrictions on the pointers. To point out a couple important things, the host can be localhost, and the remote index can be a distributed index in turn, all that is legal. That enables a bunch of very different usage modes:

  • sharding over multiple agent servers, and creating an arbitrary cluster topology;

  • sharding over multiple agent servers, mirrored for HA/LB (High Availability and Load Balancing) purposes (starting with 2.1.1-beta);

  • sharding within localhost, to utilize multiple cores (historical and not recommended in versions 1.x and above, use multiple local indexes and dist_threads directive instead);

All agents are searched in parallel. An index list is passed verbatim to the remote agent. How exactly that list is searched within the agent (ie. sequentially or in parallel too) depends solely on the agent configuration (ie. dist_threads directive). Master has no remote control over that.

Starting with 2.2.9-release, the value can additionally enumerate per agent options such as:
agent = address1:index-list[[ha_strategy=value] | [conn=value] | [blackhole=value]]


# config on box2
# sharding an index over 3 servers
agent = box2:9312:chunk2
agent = box3:9312:chunk3

# config on box2
# sharding an index over 3 servers
agent = box1:9312:chunk2
agent = box3:9312:chunk3

# config on box3
# sharding an index over 3 servers
agent = box1:9312:chunk2
agent = box2:9312:chunk3

# per agent options
agent = box1:9312:chunk1[ha_strategy=nodeads]
agent = box2:9312:chunk2[conn=pconn]
agent = test:9312:any[blackhole=1]

Agent mirrors

New syntax added in 2.1.1-beta lets you define so-called agent mirrors that can be used interchangeably when processing a search query. Master server keeps track of mirror status (alive or dead) and response times, and does automatic failover and load balancing based on that. For example, this line:

agent = box1:9312|box2:9312|box3:9312:chunk2

Declares that box1:9312, box2:9312, and box3:9312 all have an index called chunk2, and can be used as interchangeable mirrors. If any single of those servers go down, the queries will be distributed between the other two. When it gets back up, master will detect that and begin routing queries to all three boxes again.

Another way to define the mirrors is to explicitly specify the index list for every mirror:

agent = box1:9312:box1chunk2|box2:9312:box2chunk2

This works essentially the same as the previous example, but different index names will be used when querying different severs: box1chunk2 when querying box1:9312, and box2chunk when querying box2:9312.

By default, all queries are routed to the best of the mirrors. The best one is picked based on the recent statistics, as controlled by the ha_period_karma config directive. Master stores a number of metrics (total query count, error count, response time, etc) recently observed for every agent. It groups those by time spans, and karma is that time span length. The best agent mirror is then determined dynamically based on the last 2 such time spans. Specific algorithm that will be used to pick a mirror can be configured ha_strategy directive.

The karma period is in seconds and defaults to 60 seconds. Master stores upto 15 karma spans with per-agent statistics for instrumentation purposes (see SHOW AGENT STATUS statement). However, only the last 2 spans out of those are ever used for HA/LB logic.

When there are no queries, master sends a regular ping command every ha_ping_interval milliseconds in order to have some statistics and at least check, whether the remote host is still alive. ha_ping_interval defaults to 1000 msec. Setting it to 0 disables pings and statistics will only be accumulated based on actual queries.


# sharding index over 4 servers total
# in just 2 chunks but with 2 failover mirrors for each chunk
# box1, box2 carry chunk1 as local
# box3, box4 carry chunk2 as local

# config on box1, box2
agent = box3:9312|box4:9312:chunk2

# config on box3, box4
agent = box1:9312|box2:9312:chunk1

12.2.32. agent_persistent

Persistently connected remote agent declaration. Multi-value, optional, default is empty. Introduced in version 2.1.1-beta.

agent_persistent directive syntax matches that of the agent directive. The only difference is that the master will not open a new connection to the agent for every query and then close it. Rather, it will keep a connection open and attempt to reuse for the subsequent queries. The maximal number of such persistent connections per one agent host is limited by persistent_connections_limit option of searchd section.

Note, that you have to set the last one in something greater than 0 if you want to use persistent agent connections. Otherwise - when persistent_connections_limit is not defined, it assumes the zero num of persistent connections, and 'agent_persistent' acts exactly as simple 'agent'.

Persistent master-agent connections reduce TCP port pressure, and save on connection handshakes. As of time of this writing, they are supported only in workers=threads mode. In other modes, simple non-persistent connections (i.e., one connection per operation) will be used, and a warning will show up in the console.


agent_persistent = remotebox:9312:index2

12.2.33. agent_blackhole

Remote blackhole agent declaration in the distributed index. Multi-value, optional, default is empty. Introduced in version 0.9.9-rc1.

agent_blackhole lets you fire-and-forget queries to remote agents. That is useful for debugging (or just testing) production clusters: you can setup a separate debugging/testing searchd instance, and forward the requests to this instance from your production master (aggregator) instance without interfering with production work. Master searchd will attempt to connect and query blackhole agent normally, but it will neither wait nor process any responses. Also, all network errors on blackhole agents will be ignored. The value format is completely identical to regular agent directive.


agent_blackhole = testbox:9312:testindex1,testindex2

12.2.34. agent_connect_timeout

Remote agent connection timeout, in milliseconds. Optional, default is 1000 (ie. 1 second).

When connecting to remote agents, searchd will wait at most this much time for connect() call to complete successfully. If the timeout is reached but connect() does not complete, and retries are enabled, retry will be initiated.


agent_connect_timeout = 300

12.2.35. agent_query_timeout

Remote agent query timeout, in milliseconds. Optional, default is 3000 (ie. 3 seconds). Added in version 2.1.1-beta.

After connection, searchd will wait at most this much time for remote queries to complete. This timeout is fully separate from connection timeout; so the maximum possible delay caused by a remote agent equals to the sum of agent_connection_timeout and agent_query_timeout. Queries will not be retried if this timeout is reached; a warning will be produced instead.


agent_query_timeout = 10000 # our query can be long, allow up to 10 sec

12.2.36. preopen

Whether to pre-open all index files, or open them per each query. Optional, default is 0 (do not preopen).

This option tells searchd that it should pre-open all index files on startup (or rotation) and keep them open while it runs. Currently, the default mode is not to pre-open the files (this may change in the future). Preopened indexes take a few (currently 2) file descriptors per index. However, they save on per-query open() calls; and also they are invulnerable to subtle race conditions that may happen during index rotation under high load. On the other hand, when serving many indexes (100s to 1000s), it still might be desired to open the on per-query basis in order to save file descriptors.

This directive does not affect indexer in any way, it only affects searchd.


preopen = 1

12.2.37. inplace_enable

Whether to enable in-place index inversion. Optional, default is 0 (use separate temporary files). Introduced in version 0.9.9-rc1.

inplace_enable greatly reduces indexing disk footprint, at a cost of slightly slower indexing (it uses around 2x less disk, but yields around 90-95% the original performance).

Indexing involves two major phases. The first phase collects, processes, and partially sorts documents by keyword, and writes the intermediate result to temporary files (.tmp*). The second phase fully sorts the documents, and creates the final index files. Thus, rebuilding a production index on the fly involves around 3x peak disk footprint: 1st copy for the intermediate temporary files, 2nd copy for newly constructed copy, and 3rd copy for the old index that will be serving production queries in the meantime. (Intermediate data is comparable in size to the final index.) That might be too much disk footprint for big data collections, and inplace_enable allows to reduce it. When enabled, it reuses the temporary files, outputs the final data back to them, and renames them on completion. However, this might require additional temporary data chunk relocation, which is where the performance impact comes from.

This directive does not affect searchd in any way, it only affects indexer.


inplace_enable = 1

12.2.38. inplace_hit_gap

In-place inversion fine-tuning option. Controls preallocated hitlist gap size. Optional, default is 0. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


inplace_hit_gap = 1M

12.2.39. inplace_docinfo_gap

In-place inversion fine-tuning option. Controls preallocated docinfo gap size. Optional, default is 0. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


inplace_docinfo_gap = 1M

12.2.40. inplace_reloc_factor

In-place inversion fine-tuning option. Controls relocation buffer size within indexing memory arena. Optional, default is 0.1. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


inplace_reloc_factor = 0.1

12.2.41. inplace_write_factor

In-place inversion fine-tuning option. Controls in-place write buffer size within indexing memory arena. Optional, default is 0.1. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


inplace_write_factor = 0.1

12.2.42. index_exact_words

Whether to index the original keywords along with the stemmed/remapped versions. Optional, default is 0 (do not index). Introduced in version 0.9.9-rc1.

When enabled, index_exact_words forces indexer to put the raw keywords in the index along with the stemmed versions. That, in turn, enables exact form operator in the query language to work. This impacts the index size and the indexing time. However, searching performance is not impacted at all.


index_exact_words = 1

12.2.43. overshort_step

Position increment on overshort (less that min_word_len) keywords. Optional, allowed values are 0 and 1, default is 1. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


overshort_step = 1

12.2.44. stopword_step

Position increment on stopwords. Optional, allowed values are 0 and 1, default is 1. Introduced in version 0.9.9-rc1.

This directive does not affect searchd in any way, it only affects indexer.


stopword_step = 1

12.2.45. hitless_words

Hitless words list. Optional, allowed values are 'all', or a list file name. Introduced in version 1.10-beta.

By default, Sphinx full-text index stores not only a list of matching documents for every given keyword, but also a list of its in-document positions (aka hitlist). Hitlists enables phrase, proximity, strict order and other advanced types of searching, as well as phrase proximity ranking. However, hitlists for specific frequent keywords (that can not be stopped for some reason despite being frequent) can get huge and thus slow to process while querying. Also, in some cases we might only care about boolean keyword matching, and never need position-based searching operators (such as phrase matching) nor phrase ranking.

hitless_words lets you create indexes that either do not have positional information (hitlists) at all, or skip it for specific keywords.

Hitless index will generally use less space than the respective regular index (about 1.5x can be expected). Both indexing and searching should be faster, at a cost of missing positional query and ranking support. When searching, positional queries (eg. phrase queries) will be automatically converted to respective non-positional (document-level) or combined queries. For instance, if keywords "hello" and "world" are hitless, "hello world" phrase query will be converted to (hello & world) bag-of-words query, matching all documents that mention either of the keywords but not necessarily the exact phrase. And if, in addition, keywords "simon" and "says" are not hitless, "simon says hello world" will be converted to ("simon says" & hello & world) query, matching all documents that contain "hello" and "world" anywhere in the document, and also "simon says" as an exact phrase.


hitless_words = all

12.2.46. expand_keywords

Expand keywords with exact forms and/or stars when possible. Optional, default is 0 (do not expand keywords). Introduced in version 1.10-beta.

Queries against indexes with expand_keywords feature enabled are internally expanded as follows. If the index was built with prefix or infix indexing enabled, every keyword gets internally replaced with a disjunction of keyword itself and a respective prefix or infix (keyword with stars). If the index was built with both stemming and index_exact_words enabled, exact form is also added. Here's an example that shows how internal expansion works when all of the above (infixes, stemming, and exact words) are combined:

running -> ( running | *running* | =running )

Expanded queries take naturally longer to complete, but can possibly improve the search quality, as the documents with exact form matches should be ranked generally higher than documents with stemmed or infix matches.

Note that the existing query syntax does not allow to emulate this kind of expansion, because internal expansion works on keyword level and expands keywords within phrase or quorum operators too (which is not possible through the query syntax).

This directive does not affect indexer in any way, it only affects searchd.


expand_keywords = 1

12.2.47. blend_chars

Blended characters list. Optional, default is empty. Introduced in version 1.10-beta.

Blended characters are indexed both as separators and valid characters. For instance, assume that & is configured as blended and AT&T occurs in an indexed document. Three different keywords will get indexed, namely "at&t", treating blended characters as valid, plus "at" and "t", treating them as separators.

Positions for tokens obtained by replacing blended characters with whitespace are assigned as usual, so regular keywords will be indexed just as if there was no blend_chars specified at all. An additional token that mixes blended and non-blended characters will be put at the starting position. For instance, if the field contents are "AT&T company" occurs in the very beginning of the text field, "at" will be given position 1, "t" position 2, "company" position 3, and "AT&T" will also be given position 1 ("blending" with the opening regular keyword). Thus, querying for either AT&T or just AT will match that document, and querying for "AT T" as a phrase also match it. Last but not least, phrase query for "AT&T company" will also match it, despite the position

Blended characters can overlap with special characters used in query syntax (think of T-Mobile or @twitter). Where possible, query parser will automatically handle blended character as blended. For instance, "hello @twitter" within quotes (a phrase operator) would handle @-sign as blended, because @-syntax for field operator is not allowed within phrases. Otherwise, the character would be handled as an operator. So you might want to escape the keywords.

Starting with version 2.0.1-beta, blended characters can be remapped, so that multiple different blended characters could be normalized into just one base form. This is useful when indexing multiple alternative Unicode codepoints with equivalent glyphs.


blend_chars = +, &, U+23
blend_chars = +, &->+ # 2.0.1 and above

12.2.48. blend_mode

Blended tokens indexing mode. Optional, default is trim_none. Introduced in version 2.0.1-beta.

By default, tokens that mix blended and non-blended characters get indexed in there entirety. For instance, when both at-sign and an exclamation are in blend_chars, "@dude!" will get result in two tokens indexed: "@dude!" (with all the blended characters) and "dude" (without any). Therefore "@dude" query will not match it.

blend_mode directive adds flexibility to this indexing behavior. It takes a comma-separated list of options.

blend_mode = option [, option [, ...]]
option = trim_none | trim_head | trim_tail | trim_both | skip_pure

Options specify token indexing variants. If multiple options are specified, multiple variants of the same token will be indexed. Regular keywords (resulting from that token by replacing blended with whitespace) are always be indexed.


Index the entire token.


Trim heading blended characters, and index the resulting token.


Trim trailing blended characters, and index the resulting token.


Trim both heading and trailing blended characters, and index the resulting token.


Do not index the token if it's purely blended, that is, consists of blended characters only.

Returning to the "@dude!" example above, setting blend_mode = trim_head, trim_tail will result in two tokens being indexed, "@dude" and "dude!". In this particular example, trim_both would have no effect, because trimming both blended characters results in "dude" which is already indexed as a regular keyword. Indexing "@U.S.A." with trim_both (and assuming that dot is blended two) would result in "U.S.A" being indexed. Last but not least, skip_pure enables you to fully ignore sequences of blended characters only. For example, "one @@@ two" would be indexed exactly as "one two", and match that as a phrase. That is not the case by default because a fully blended token gets indexed and offsets the second keyword position.

Default behavior is to index the entire token, equivalent to blend_mode = trim_none.


blend_mode = trim_tail, skip_pure

12.2.49. rt_mem_limit

RAM chunk size limit. Optional, default is 128M. Introduced in version 1.10-beta.

RT index keeps some data in memory (so-called RAM chunk) and also maintains a number of on-disk indexes (so-called disk chunks). This directive lets you control the RAM chunk size. Once there's too much data to keep in RAM, RT index will flush it to disk, activate a newly created disk chunk, and reset the RAM chunk.

The limit is pretty strict; RT index should never allocate more memory than it's limited to. The memory is not preallocated either, hence, specifying 512 MB limit and only inserting 3 MB of data should result in allocating 3 MB, not 512 MB.


rt_mem_limit = 512M

12.2.50. rt_field

Full-text field declaration. Multi-value, mandatory Introduced in version 1.10-beta.

Full-text fields to be indexed are declared using rt_field directive. The names must be unique. The order is preserved; and so field values in INSERT statements without an explicit list of inserted columns will have to be in the same order as configured.


rt_field = author
rt_field = title
rt_field = content

12.2.51. rt_attr_uint

Unsigned integer attribute declaration. Multi-value (an arbitrary number of attributes is allowed), optional. Declares an unsigned 32-bit attribute. Introduced in version 1.10-beta.


rt_attr_uint = gid

12.2.52. rt_attr_bool

Boolean attribute declaration. Multi-value (there might be multiple attributes declared), optional. Declares a 1-bit unsigned integer attribute. Introduced in version 2.1.2-release.


rt_attr_bool = available

12.2.53. rt_attr_bigint

BIGINT attribute declaration. Multi-value (an arbitrary number of attributes is allowed), optional. Declares a signed 64-bit attribute. Introduced in version 1.10-beta.


rt_attr_bigint = guid

12.2.54. rt_attr_float

Floating point attribute declaration. Multi-value (an arbitrary number of attributes is allowed), optional. Declares a single precision, 32-bit IEEE 754 format float attribute. Introduced in version 1.10-beta.


rt_attr_float = gpa

12.2.55. rt_attr_multi

Multi-valued attribute (MVA) declaration. Declares the UNSIGNED INTEGER (unsigned 32-bit) MVA attribute. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to RT indexes only.


rt_attr_multi = my_tags

12.2.56. rt_attr_multi_64

Multi-valued attribute (MVA) declaration. Declares the BIGINT (signed 64-bit) MVA attribute. Multi-value (ie. there may be more than one such attribute declared), optional. Applies to RT indexes only.


rt_attr_multi_64 = my_wide_tags

12.2.57. rt_attr_timestamp

Timestamp attribute declaration. Multi-value (an arbitrary number of attributes is allowed), optional. Introduced in version 1.10-beta.


rt_attr_timestamp = date_added

12.2.58. rt_attr_string

String attribute declaration. Multi-value (an arbitrary number of attributes is allowed), optional. Introduced in version 1.10-beta.


rt_attr_string = author

12.2.59. rt_attr_json

JSON attribute declaration. Multi-value (ie. there may be more than one such attribute declared), optional. Introduced in version 2.1.1-beta.

Refer to Section 12.1.24, “sql_attr_json” for more details on the JSON attributes.


rt_attr_json = properties

12.2.60. ha_strategy

Agent mirror selection strategy, for load balancing. Optional, default is random. Added in 2.1.1-beta.

The strategy used for mirror selection, or in other words, choosing a specific agent mirror in a distributed index. Essentially, this directive controls how exactly master does the load balancing between the configured mirror agent nodes. As of 2.1.1-beta, the following strategies are implemented:

Simple random balancing

ha_strategy = random

The default balancing mode. Simple linear random distribution among the mirrors. That is, equal selection probability are assigned to every mirror. Kind of similar to round-robin (RR), but unlike RR, does not impose a strict selection order.

Adaptive randomized balancing

The default simple random strategy does not take mirror status, error rate, and, most importantly, actual response latencies into account. So to accommodate for heterogeneous clusters and/or temporary spikes in agent node load, we have a group of balancing strategies that dynamically adjusts the probabilities based on the actual query latencies observed by the master.

The adaptive strategies based on latency-weighted probabilities basically work as follows:

  • latency stats are accumulated, in blocks of ha_period_karma seconds;

  • once per karma period, latency-weighted probabilities get recomputed;

  • once per request (including ping requests), "dead or alive" flag is adjusted.

Currently (as of 2.1.1-beta), we begin with equal probabilities (or percentages, for brevity), and on every step, scale them by the inverse of the latencies observed during the last "karma" period, and then renormalize them. For example, if during the first 60 seconds after the master startup 4 mirrors had latencies of 10, 5, 30, and 3 msec/query respectively, the first adjustment step would go as follow:

  • initial percentages: 0.25, 0.25, 0.25, 0.2%;

  • observed latencies: 10 ms, 5 ms, 30 ms, 3 ms;

  • inverse latencies: 0.1, 0.2, 0.0333, 0.333;

  • scaled percentages: 0.025, 0.05, 0.008333, 0.0833;

  • renormalized percentages: 0.15, 0.30, 0.05, 0.50.

Meaning that the 1st mirror would have a 15% chance of being chosen during the next karma period, the 2nd one a 30% chance, the 3rd one (slowest at 30 ms) only a 5% chance, and the 4th and the fastest one (at 3 ms) a 50% chance. Then, after that period, the second adjustment step would update those chances again, and so on.

The rationale here is, once the observed latencies stabilize, the latency weighted probabilities stabilize as well. So all these adjustment iterations are supposed to converge at a point where the average latencies are (roughly) equal over all mirrors.

ha_strategy = nodeads

Latency-weighted probabilities, but dead mirrors are excluded from the selection. "Dead" mirror is defined as a mirror that resulted in multiple hard errors (eg. network failure, or no answer, etc) in a row.

ha_strategy = noerrors

Latency-weighted probabilities, but mirrors with worse errors/success ratio are excluded from the selection.

Round-robin balancing

ha_strategy = roundrobin

Simple round-robin selection, that is, selecting the 1st mirror in the list, then the 2nd one, then the 3rd one, etc, and then repeating the process once the last mirror in the list is reached. Unlike with the randomized strategies, RR imposes a strict querying order (1, 2, 3, .., N-1, N, 1, 2, 3, ... and so on) and guarantees that no two subsequent queries will be sent to the same mirror.

12.2.61. bigram_freq_words

A list of keywords considered "frequent" when indexing bigrams. Optional, default is empty. Added in 2.1.1-beta.

Bigram indexing is a feature to accelerate phrase searches. When indexing, it stores a document list for either all or some of the adjacent words pairs into the index. Such a list can then be used at searching time to significantly accelerate phrase or sub-phrase matching.

Some of the bigram indexing modes (see Section 12.2.62, “bigram_index”) require to define a list of frequent keywords. These are not to be confused with stopwords! Stopwords are completely eliminated when both indexing and searching. Frequent keywords are only used by bigrams to determine whether to index a current word pair or not.

bigram_freq_words lets you define a list of such keywords.


bigram_freq_words = the, a, you, i

12.2.62. bigram_index

Bigram indexing mode. Optional, default is none. Added in 2.1.1-beta.

Bigram indexing is a feature to accelerate phrase searches. When indexing, it stores a document list for either all or some of the adjacent words pairs into the index. Such a list can then be used at searching time to significantly accelerate phrase or sub-phrase matching.

bigram_index controls the selection of specific word pairs. The known modes are:

  • all, index every single word pair. (NB: probably totally not worth it even on a moderately sized index, but added anyway for the sake of completeness.)

  • first_freq, only index word pairs where the first word is in a list of frequent words (see Section 12.2.61, “bigram_freq_words”). For example, with bigram_freq_words = the, in, i, a, indexing "alone in the dark" text will result in "in the" and "the dark" pairs being stored as bigrams, because they begin with a frequent keyword (either "in" or "the" respectively), but "alone in" would not be indexed, because "in" is a second word in that pair.

  • both_freq, only index word pairs where both words are frequent. Continuing with the same example, in this mode indexing "alone in the dark" would only store "in the" (the very worst of them all from searching perspective) as a bigram, but none of the other word pairs.

For most usecases, both_freq would be the best mode, but your mileage may vary.


bigram_freq_words = both_freq

12.2.63. index_field_lengths

Enables computing and storing of field lengths (both per-document and average per-index values) into the index. Optional, default is 0 (do not compute and store). Added in 2.1.1-beta.

When index_field_lengths is set to 1, indexer will 1) create a respective length attribute for every full-text field, sharing the same name but with _len suffix; 2) compute a field length (counted in keywords) for every document and store in to a respective attribute; 3) compute the per-index averages. The lengths attributes will have a special TOKENCOUNT type, but their values are in fact regular 32-bit integers, and their values are generally accessible.

BM25A() and BM25F() functions in the expression ranker are based on these lengths and require index_field_lengths to be enabled. Historically, Sphinx used a simplified, stripped-down variant of BM25 that, unlike the complete function, did not account for document length. (We later realized that it should have been called BM15 from the start.) Starting with 2.1.1-beta, we added support for both a complete variant of BM25, and its extension towards multiple fields, called BM25F. They require per-document length and per-field lengths, respectively. Hence the additional directive.


index_field_lengths = 1

12.2.64. regexp_filter

Regular expressions (regexps) to filter the fields and queries with. Optional, multi-value, default is an empty list of regexps. Added in 2.1.1-beta.

In certain applications (like product search) there can be many different ways to call a model, or a product, or a property, and so on. For instance, 'iphone 3gs' and 'iphone 3 gs' (or even 'iphone3 gs') are very likely to mean the same product. Or, for a more tricky example, '13-inch', '13 inch', '13"', and '13in' in a laptop screen size descriptions do mean the same.

Regexps provide you with a mechanism to specify a number of rules specific to your application to handle such cases. In the first 'iphone 3gs' example, you could possibly get away with a wordforms files tailored to handle a handful of iPhone models. However even in a comparatively simple second '13-inch' example there is just way too many individual forms and you are better off specifying rules that would normalize both '13-inch' and '13in' to something identical.

Regular expressions listed in regexp_filter are applied in the order they are listed. That happens at the earliest stage possible, before any other processing, even before tokenization. That is, regexps are applied to the raw source fields when indeixng, and to the raw search query text when searching.

We use the RE2 engine to implement regexps. So when building from the source, the library must be installed in the system and Sphinx must be configured built with a --with-re2 switch. Binary packages should come with RE2 builtin.


# index '13-inch' as '13inch'
regexp_filter = \b(\d+)\" => \1inch

# index 'blue' or 'red' as 'color'
regexp_filter = (blue|red) => color

12.2.65. stopwords_unstemmed

Whether to apply stopwords before or after stemming. Optional, default is 0 (apply stopword filter after stemming). Added in 2.1.1-beta.

By default, stopwords are stemmed themselves, and applied to tokens after stemming (or any other morphology processing). In other words, by default, a token is stopped when stem(token) == stem(stopword). That can lead to unexpected results when a token gets (erroneously) stemmed to a stopped root. For example, 'Andes' gets stemmed to 'and' by our current stemmer implementation, so when 'and' is a stopword, 'Andes' is also stopped.

stopwords_unstemmed directive fixes that issue. When it's enabled, stopwords are applied before stemming (and therefore to the original word forms), and the tokens are stopped when token == stopword.


stopwords_unstemmed = 1

12.2.66. global_idf

The path to a file with global (cluster-wide) keyword IDFs. Optional, default is empty (use local IDFs). Added in 2.1.1-beta.

On a multi-index cluster, per-keyword frequencies are quite likely to differ across different indexes. That means that when the ranking function uses TF-IDF based values, such as BM25 family of factors, the results might be ranked slightly different depending on what cluster node they reside.

The easiest way to fix that issue is to create and utilize a global frequency dictionary, or a global IDF file for short. This directive lets you specify the location of that file. It it suggested (but not required) to use a .idf extension. When the IDF file is specified for a given index and and OPTION global_idf is set to 1, the engine will use the keyword frequencies and collection documents count from the global_idf file, rather than just the local index. That way, IDFs and the values that depend on them will stay consistent across the cluster.

IDF files can be shared across multiple indexes. Only a single copy of an IDF file will be loaded by searchd, even when many indexes refer to that file. Should the contents of an IDF file change, the new contents can be loaded with a SIGHUP.

You can build an .idf file using indextool utility, by dumping dictionaries using --dumpdict switch first, then converting those to .idf format using --buildidf, then merging all .idf files across cluser using --mergeidf. Refer to Section 7.4, “indextool command reference” for more information.


global_idf = /usr/local/sphinx/var/global.idf

12.2.67. rlp_context

RLP context configuration file. Mandatory if RLP is used. Added in 2.2.1-beta.


rlp_context = /home/myuser/RLP/rlp-context.xml

12.2.68. ondisk_attrs

Allows for fine-grain control over how attributes are loaded into memory when using indexes with external storage. It is now possible (since version 2.2.1-beta) to keep attributes on disk. Although, the daemon does map them to memory and the OS loads small chunks of data on demand. This allows use of docinfo = extern instead of docinfo = inline, but still leaves plenty of free memory for cases when you have large collections of pooled attributes (string/JSON/MVA) or when you're using many indexes per daemon that don't consume memory. It is not possible to update attributes left on disk when this option is enabled and the constraint of 4Gb of entries per pool is still in effect.

Note that this option also affects RT indexes. When it is enabled, all atribute updates will be disabled, and also all disk chunks of RT indexes will behave described above. However inserting and deleting of docs from RT indexes is still possible with enabled ondisk_attrs.

Possible values:

  • 0 - disabled and default value, all attributes are loaded in memory (the normal behaviour of docinfo = extern)
  • 1 - all attributes stay on disk. Daemon loads no files (spa, spm, sps). This is the most memory conserving mode, however it is also the slowest as the whole doc-id-list and block index doesn't load.
  • pool - only pooled attributes stay on disk. Pooled attributes are string, MVA, and JSON attributes (sps, spm files). Scalar attributes stored in docinfo (spa file) load as usual.

This option does not affect indexing in any way, it only requires daemon restart.


ondisk_attrs = pool #keep pooled attributes on disk

12.3. indexer program configuration options

12.3.1. mem_limit

Indexing RAM usage limit. Optional, default is 128M.

Enforced memory usage limit that the indexer will not go above. Can be specified in bytes, or kilobytes (using K postfix), or megabytes (using M postfix); see the example. This limit will be automatically raised if set to extremely low value causing I/O buffers to be less than 8 KB; the exact lower bound for that depends on the indexed data size. If the buffers are less than 256 KB, a warning will be produced.

Maximum possible limit is 2047M. Too low values can hurt indexing speed, but 256M to 1024M should be enough for most if not all datasets. Setting this value too high can cause SQL server timeouts. During the document collection phase, there will be periods when the memory buffer is partially sorted and no communication with the database is performed; and the database server can timeout. You can resolve that either by raising timeouts on SQL server side or by lowering mem_limit.


mem_limit = 256M
# mem_limit = 262144K # same, but in KB
# mem_limit = 268435456 # same, but in bytes

12.3.2. max_iops

Maximum I/O operations per second, for I/O throttling. Optional, default is 0 (unlimited).

I/O throttling related option. It limits maximum count of I/O operations (reads or writes) per any given second. A value of 0 means that no limit is imposed.

indexer can cause bursts of intensive disk I/O during indexing, and it might desired to limit its disk activity (and keep something for other programs running on the same machine, such as searchd). I/O throttling helps to do that. It works by enforcing a minimum guaranteed delay between subsequent disk I/O operations performed by indexer. Modern SATA HDDs are able to perform up to 70-100+ I/O operations per second (that's mostly limited by disk heads seek time). Limiting indexing I/O to a fraction of that can help reduce search performance degradation caused by indexing.


max_iops = 40

12.3.3. max_iosize

Maximum allowed I/O operation size, in bytes, for I/O throttling. Optional, default is 0 (unlimited).

I/O throttling related option. It limits maximum file I/O operation (read or write) size for all operations performed by indexer. A value of 0 means that no limit is imposed. Reads or writes that are bigger than the limit will be split in several smaller operations, and counted as several operation by max_iops setting. At the time of this writing, all I/O calls should be under 256 KB (default internal buffer size) anyway, so max_iosize values higher than 256 KB must not affect anything.


max_iosize = 1048576

12.3.4. max_xmlpipe2_field

Maximum allowed field size for XMLpipe2 source type, bytes. Optional, default is 2 MB.


max_xmlpipe2_field = 8M

12.3.5. write_buffer

Write buffer size, bytes. Optional, default is 1 MB.

Write buffers are used to write both temporary and final index files when indexing. Larger buffers reduce the number of required disk writes. Memory for the buffers is allocated in addition to mem_limit. Note that several (currently up to 4) buffers for different files will be allocated, proportionally increasing the RAM usage.


write_buffer = 4M

12.3.6. max_file_field_buffer

Maximum file field adaptive buffer size, bytes. Optional, default is 8 MB, minimum is 1 MB.

File field buffer is used to load files referred to from sql_file_field columns. This buffer is adaptive, starting at 1 MB at first allocation, and growing in 2x steps until either file contents can be loaded, or maximum buffer size, specified by max_file_field_buffer directive, is reached.

Thus, if there are no file fields are specified, no buffer is allocated at all. If all files loaded during indexing are under (for example) 2 MB in size, but max_file_field_buffer value is 128 MB, peak buffer usage would still be only 2 MB. However, files over 128 MB would be entirely skipped.


max_file_field_buffer = 128M

12.3.7. on_file_field_error

How to handle IO errors in file fields. Optional, default is ignore_field. Introduced in version 2.0.2-beta.

When there is a problem indexing a file referenced by a file field (Section 12.1.27, “sql_file_field”), indexer can either index the document, assuming empty content in this particular field, or skip the document, or fail indexing entirely. on_file_field_error directive controls that behavior. The values it takes are:

  • ignore_field, index the current document without field;

  • skip_document, skip the current document but continue indexing;

  • fail_index, fail indexing with an error message.

The problems that can arise are: open error, size error (file too big), and data read error. Warning messages on any problem will be given at all times, irregardless of the phase and the on_file_field_error setting.

Note that with on_file_field_error = skip_document documents will only be ignored if problems are detected during an early check phase, and not during the actual file parsing phase. indexer will open every referenced file and check its size before doing any work, and then open it again when doing actual parsing work. So in case a file goes away between these two open attempts, the document will still be indexed.


on_file_field_error = skip_document

12.3.8. lemmatizer_cache

Lemmatizer cache size. Optional, default is 256K. Added in version 2.1.1-beta.

Our lemmatizer implementation (see Section 12.2.6, “morphology” for a discussion of what lemmatizers are) uses a compressed dictionary format that enables a space/speed tradeoff. It can either perform lemmatization off the compressed data, using more CPU but less RAM, or it can decompress and precache the dictionary either partially or fully, thus using less CPU but more RAM. And the lemmatizer_cache directive lets you control how much RAM exactly can be spent for that uncompressed dictionary cache.

Currently, the only available dictionary is ru.pak, the Russian one. The compressed dictionary is approximately 10 MB in size. Note that the dictionary stays in memory at all times, too. The default cache size is 256 KB. The accepted cache sizes are 0 to 2047 MB. It's safe to raise the cache size too high; the lemmatizer will only use the needed memory. For instance, the entire Russian dictionary decompresses to approximately 110 MB; and thus setting lemmatizer_cache anywhere higher than that will not affect the memory use: even when 1024 MB is allowed for the cache, if only 110 MB is needed, it will only use those 110 MB.

On our benchmarks, the total indexing time with different cache sizes was as follows:

  • 9.07 sec, morphology = lemmatize_ru, lemmatizer_cache = 0
  • 8.60 sec, morphology = lemmatize_ru, lemmatizer_cache = 256K
  • 8.33 sec, morphology = lemmatize_ru, lemmatizer_cache = 8M
  • 7.95 sec, morphology = lemmatize_ru, lemmatizer_cache = 128M
  • 6.85 sec, morphology = stem_ru (baseline)

Your mileage may vary, but a simple rule of thumb would be to either go with the small default 256 KB cache when pressed for memory, or spend 128 MB extra RAM and cache the entire dictionary for maximum indexing performance.


lemmatizer_cache = 256M # cache it all

12.4. searchd program configuration options

12.4.1. listen

This setting lets you specify IP address and port, or Unix-domain socket path, that searchd will listen on. Introduced in version 0.9.9-rc1.

The informal grammar for listen setting is:

listen = ( address ":" port | port | path ) [ ":" protocol ]

I.e. you can specify either an IP address (or hostname) and port number, or just a port number, or Unix socket path. If you specify port number but not the address, searchd will listen on all network interfaces. Unix path is identified by a leading slash.

Starting with version 0.9.9-rc2, you can also specify a protocol handler (listener) to be used for connections on this socket. Supported protocol values are 'sphinx' (Sphinx 0.9.x API protocol) and 'mysql41' (MySQL protocol used since 4.1 upto at least 5.1). More details on MySQL protocol support can be found in Section 5.10, “MySQL protocol support and SphinxQL” section.


listen = localhost
listen = localhost:5000
listen =
listen = /var/run/sphinx.s
listen = 9312
listen = localhost:9306:mysql41

There can be multiple listen directives, searchd will listen for client connections on all specified ports and sockets. If no listen directives are found then the server will listen on all available interfaces using the default SphinxAPI port 9312. Starting with 1.10-beta, it will also listen on default SphinxQL port 9306. Both port numbers are assigned by IANA (see http://www.iana.org/assignments/port-numbers for details) and should therefore be available.

Unix-domain sockets are not supported on Windows.

12.4.2. log

Log file name. Optional, default is 'searchd.log'. All searchd run time events will be logged in this file.

Also you can use the 'syslog' as the file name. In this case the events will be sent to syslog daemon. To use the syslog option the sphinx must be configured '--with-syslog' on building.


log = /var/log/searchd.log

12.4.3. query_log

Query log file name. Optional, default is empty (do not log queries). All search queries will be logged in this file. The format is described in Section 5.9, “searchd query log formats”.

In case of 'plain' format, you can use the 'syslog' as the path to the log file. In this case all search queries will be sent to syslog daemon with LOG_INFO priority, prefixed with '[query]' instead of timestamp. To use the syslog option the sphinx must be configured '--with-syslog' on building.


query_log = /var/log/query.log

12.4.4. query_log_format

Query log format. Optional, allowed values are 'plain' and 'sphinxql', default is 'plain'. Introduced in version 2.0.1-beta.

Starting with version 2.0.1-beta, two different log formats are supported. The default one logs queries in a custom text format. The new one logs valid SphinxQL statements. This directive allows to switch between the two formats on search daemon startup. The log format can also be altered on the fly, using SET GLOBAL query_log_format=sphinxql syntax. Refer to Section 5.9, “searchd query log formats” for more discussion and format details.


query_log_format = sphinxql

12.4.5. read_timeout

Network client request read timeout, in seconds. Optional, default is 5 seconds. searchd will forcibly close the client connections which fail to send a query within this timeout.


read_timeout = 1

12.4.6. client_timeout

Maximum time to wait between requests (in seconds) when using persistent connections. Optional, default is five minutes.


client_timeout = 3600

12.4.7. max_children

Maximum amount of children to fork (or in other words, concurrent searches to run in parallel). Optional, default is 0 (unlimited).

Useful to control server load. There will be no more than this much concurrent searches running, at all times. When the limit is reached, additional incoming clients are dismissed with temporarily failure (SEARCHD_RETRY) status code and a message stating that the server is maxed out.


max_children = 10

12.4.8. pid_file

searchd process ID file name. Mandatory.

PID file will be re-created (and locked) on startup. It will contain head daemon process ID while the daemon is running, and it will be unlinked on daemon shutdown. It's mandatory because Sphinx uses it internally for a number of things: to check whether there already is a running instance of searchd; to stop searchd; to notify it that it should rotate the indexes. Can also be used for different external automation scripts.


pid_file = /var/run/searchd.pid

12.4.9. seamless_rotate

Prevents searchd stalls while rotating indexes with huge amounts of data to precache. Optional, default is 1 (enable seamless rotation). On Windows systems seamless rotation is disabled by default.

Indexes may contain some data that needs to be precached in RAM. At the moment, .spa, .spi and .spm files are fully precached (they contain attribute data, MVA data, and keyword index, respectively.) Without seamless rotate, rotating an index tries to use as little RAM as possible and works as follows:

  1. new queries are temporarily rejected (with "retry" error code);

  2. searchd waits for all currently running queries to finish;

  3. old index is deallocated and its files are renamed;

  4. new index files are renamed and required RAM is allocated;

  5. new index attribute and dictionary data is preloaded to RAM;

  6. searchd resumes serving queries from new index.

However, if there's a lot of attribute or dictionary data, then preloading step could take noticeable time - up to several minutes in case of preloading 1-5+ GB files.

With seamless rotate enabled, rotation works as follows:

  1. new index RAM storage is allocated;

  2. new index attribute and dictionary data is asynchronously preloaded to RAM;

  3. on success, old index is deallocated and both indexes' files are renamed;

  4. on failure, new index is deallocated;

  5. at any given moment, queries are served either from old or new index copy.

Seamless rotate comes at the cost of higher peak memory usage during the rotation (because both old and new copies of .spa/.spi/.spm data need to be in RAM while preloading new copy). Average usage stays the same.


seamless_rotate = 1

12.4.10. preopen_indexes

Whether to forcibly preopen all indexes on startup. Optional, default is 1 (preopen everything).

Starting with 2.0.1-beta, the default value for this option is now 1 (foribly preopen all indexes). In prior versions, it used to be 0 (use per-index settings).

When set to 1, this directive overrides and enforces preopen on all indexes. They will be preopened, no matter what is the per-index preopen setting. When set to 0, per-index settings can take effect. (And they default to 0.)

Pre-opened indexes avoid races between search queries and rotations that can cause queries to fail occasionally. They also make searchd use more file handles. In most scenarios it's therefore preferred and recommended to preopen indexes.


preopen_indexes = 1

12.4.11. unlink_old

Whether to unlink .old index copies on successful rotation. Optional, default is 1 (do unlink).


unlink_old = 0

12.4.12. attr_flush_period

When calling UpdateAttributes() to update document attributes in real-time, changes are first written to the in-memory copy of attributes (docinfo must be set to extern). Then, once searchd shuts down normally (via SIGTERM being sent), the changes are written to disk. Introduced in version 0.9.9-rc1.

Starting with 0.9.9-rc1, it is possible to tell searchd to periodically write these changes back to disk, to avoid them being lost. The time between those intervals is set with attr_flush_period, in seconds.

It defaults to 0, which disables the periodic flushing, but flushing will still occur at normal shut-down.


attr_flush_period = 900 # persist updates to disk every 15 minutes

12.4.13. max_packet_size

Maximum allowed network packet size. Limits both query packets from clients, and response packets from remote agents in distributed environment. Only used for internal sanity checks, does not directly affect RAM use or performance. Optional, default is 8M. Introduced in version 0.9.9-rc1.


max_packet_size = 32M

12.4.14. mva_updates_pool

Shared pool size for in-memory MVA updates storage. Optional, default size is 1M. Introduced in version 0.9.9-rc1.

This setting controls the size of the shared storage pool for updated MVA values. Specifying 0 for the size disable MVA updates at all. Once the pool size limit is hit, MVA update attempts will result in an error. However, updates on regular (scalar) attributes will still work. Due to internal technical difficulties, currently it is not possible to store (flush) any updates on indexes where MVA were updated; though this might be implemented in the future. In the meantime, MVA updates are intended to be used as a measure to quickly catchup with latest changes in the database until the next index rebuild; not as a persistent storage mechanism.


mva_updates_pool = 16M

12.4.15. max_filters

Maximum allowed per-query filter count. Only used for internal sanity checks, does not directly affect RAM use or performance. Optional, default is 256. Introduced in version 0.9.9-rc1.


max_filters = 1024

12.4.16. max_filter_values

Maximum allowed per-filter values count. Only used for internal sanity checks, does not directly affect RAM use or performance. Optional, default is 4096. Introduced in version 0.9.9-rc1.


max_filter_values = 16384

12.4.17. listen_backlog

TCP listen backlog. Optional, default is 5.

Windows builds currently (as of 0.9.9) can only process the requests one by one. Concurrent requests will be enqueued by the TCP stack on OS level, and requests that can not be enqueued will immediately fail with "connection refused" message. listen_backlog directive controls the length of the connection queue. Non-Windows builds should work fine with the default value.


listen_backlog = 20

12.4.18. read_buffer

Per-keyword read buffer size. Optional, default is 256K.

For every keyword occurrence in every search query, there are two associated read buffers (one for document list and one for hit list). This setting lets you control their sizes, increasing per-query RAM use, but possibly decreasing IO time.


read_buffer = 1M

12.4.19. read_unhinted

Unhinted read size. Optional, default is 32K.

When querying, some reads know in advance exactly how much data is there to be read, but some currently do not. Most prominently, hit list size in not currently known in advance. This setting lest you control how much data to read in such cases. It will impact hit list IO time, reducing it for lists larger than unhinted read size, but raising it for smaller lists. It will not affect RAM use because read buffer will be already allocated. So it should be not greater than read_buffer.


read_unhinted = 32K

12.4.20. max_batch_queries

Limits the amount of queries per batch. Optional, default is 32.

Makes searchd perform a sanity check of the amount of the queries submitted in a single batch when using multi-queries. Set it to 0 to skip the check.


max_batch_queries = 256

12.4.21. subtree_docs_cache

Max common subtree document cache size, per-query. Optional, default is 0 (disabled).

Limits RAM usage of a common subtree optimizer (see Section 5.11, “Multi-queries”). At most this much RAM will be spent to cache document entries per each query. Setting the limit to 0 disables the optimizer.


subtree_docs_cache = 8M

12.4.22. subtree_hits_cache

Max common subtree hit cache size, per-query. Optional, default is 0 (disabled).

Limits RAM usage of a common subtree optimizer (see Section 5.11, “Multi-queries”). At most this much RAM will be spent to cache keyword occurrences (hits) per each query. Setting the limit to 0 disables the optimizer.


subtree_hits_cache = 16M

12.4.23. workers

Multi-processing mode (MPM). Optional; allowed values are none, fork, prefork, and threads. Default is threads. Introduced in version 1.10-beta.

Lets you choose how searchd processes multiple concurrent requests. The possible values are:


All requests will be handled serially, one-by-one. Prior to 1.x, this was the only mode available on Windows.


A new child process will be forked to handle every incoming request.


On startup, searchd will pre-fork a number of worker processes, and pass the incoming requests to one of those children.


A new thread will be created to handle every incoming request. This is the only mode compatible with RT indexing backend. This is a default value.

Historically, searchd used fork-based model, which generally performs OK but spends a noticeable amount of CPU in fork() system call when there's a high amount of (tiny) requests per second. Prefork mode was implemented to alleviate that; with prefork, worker processes are basically only created on startup and re-created on index rotation, somewhat reducing fork() call pressure.

Threads mode was implemented along with RT backend and is required to use RT indexes. (Regular disk-based indexes work in all the available modes.)


workers = threads

12.4.24. dist_threads

Max local worker threads to use for parallelizable requests (searching a distributed index; building a batch of snippets). Optional, default is 0, which means to disable in-request parallelism. Introduced in version 1.10-beta.

Distributed index can include several local indexes. dist_threads lets you easily utilize multiple CPUs/cores for that (previously existing alternative was to specify the indexes as remote agents, pointing searchd to itself and paying some network overheads).

When set to a value N greater than 1, this directive will create up to N threads for every query, and schedule the specific searches within these threads. For example, if there are 7 local indexes to search and dist_threads is set to 2, then 2 parallel threads would be created: one that sequentially searches 4 indexes, and another one that searches the other 3 indexes.

In case of CPU bound workload, setting dist_threads to 1x the number of cores is advised (creating more threads than cores will not improve query time). In case of mixed CPU/disk bound workload it might sometimes make sense to use more (so that all cores could be utilizes even when there are threads that wait for I/O completion).

Note that dist_threads does not require threads MPM. You can perfectly use it with fork or prefork MPMs too.

Starting with version 2.0.1-beta, building a batch of snippets with load_files flag enabled can also be parallelized. Up to dist_threads threads are be created to process those files. That speeds up snippet extraction when the total amount of document data to process is significant (hundreds of megabytes).


index dist_test
    type = distributed
    local = chunk1
    local = chunk2
    local = chunk3
    local = chunk4

# ...

dist_threads = 4

12.4.25. binlog_path

Binary log (aka transaction log) files path. Optional, default is build-time configured data directory. Introduced in version 1.10-beta.

Binary logs are used for crash recovery of RT index data, and also of attributes updates of plain disk indices that would otherwise only be stored in RAM until flush. When logging is enabled, every transaction COMMIT-ted into RT index gets written into a log file. Logs are then automatically replayed on startup after an unclean shutdown, recovering the logged changes.

binlog_path directive specifies the binary log files location. It should contain just the path; searchd will create and unlink multiple binlog.* files in that path as necessary (binlog data, metadata, and lock files, etc).

Empty value disables binary logging. That improves performance, but puts RT index data at risk.

WARNING! It is strongly recommended to always explicitly define 'binlog_path' option in your config. Otherwise, the default path, which in most cases is the same as working folder, may point to the folder with no write access (for example, /usr/local/var/data). In this case, the searchd will not start at all.


binlog_path = # disable logging
binlog_path = /var/data # /var/data/binlog.001 etc will be created

12.4.26. binlog_flush

Binary log transaction flush/sync mode. Optional, default is 2 (flush every transaction, sync every second). Introduced in version 1.10-beta.

This directive controls how frequently will binary log be flushed to OS and synced to disk. Three modes are supported:

  • 0, flush and sync every second. Best performance, but up to 1 second worth of committed transactions can be lost both on daemon crash, or OS/hardware crash.

  • 1, flush and sync every transaction. Worst performance, but every committed transaction data is guaranteed to be saved.

  • 2, flush every transaction, sync every second. Good performance, and every committed transaction is guaranteed to be saved in case of daemon crash. However, in case of OS/hardware crash up to 1 second worth of committed transactions can be lost.

For those familiar with MySQL and InnoDB, this directive is entirely similar to innodb_flush_log_at_trx_commit. In most cases, the default hybrid mode 2 provides a nice balance of speed and safety, with full RT index data protection against daemon crashes, and some protection against hardware ones.


binlog_flush = 1 # ultimate safety, low speed

12.4.27. binlog_max_log_size

Maximum binary log file size. Optional, default is 0 (do not reopen binlog file based on size). Introduced in version 1.10-beta.

A new binlog file will be forcibly opened once the current binlog file reaches this limit. This achieves a finer granularity of logs and can yield more efficient binlog disk usage under certain borderline workloads.


binlog_max_log_size = 16M

12.4.28. snippets_file_prefix

A prefix to prepend to the local file names when generating snippets. Optional, default is empty. Introduced in version 2.1.1-beta.

This prefix can be used in distributed snippets generation along with load_files or load_files_scattered options.

Note how this is a prefix, and not a path! Meaning that if a prefix is set to "server1" and the request refers to "file23", searchd will attempt to open "server1file23" (all of that without quotes). So if you need it to be a path, you have to mention the trailing slash.

Note also that this is a local option, it does not affect the agents anyhow. So you can safely set a prefix on a master server. The requests routed to the agents will not be affected by the master's setting. They will however be affected by the agent's own settings.

This might be useful, for instance, when the document storage locations (be those local storage or NAS mountpoints) are inconsistent across the servers.


snippets_file_prefix = /mnt/common/server1/

12.4.29. collation_server

Default server collation. Optional, default is libc_ci. Introduced in version 2.0.1-beta.

Specifies the default collation used for incoming requests. The collation can be overridden on a per-query basis. Refer to Section 5.12, “Collations” section for the list of available collations and other details.


collation_server = utf8_ci

12.4.30. collation_libc_locale

Server libc locale. Optional, default is C. Introduced in version 2.0.1-beta.

Specifies the libc locale, affecting the libc-based collations. Refer to Section 5.12, “Collations” section for the details.


collation_libc_locale = fr_FR

12.4.31. mysql_version_string

A server version string to return via MySQL protocol. Optional, default is empty (return Sphinx version). Introduced in version 2.0.1-beta.

Several picky MySQL client libraries depend on a particular version number format used by MySQL, and moreover, sometimes choose a different execution path based on the reported version number (rather than the indicated capabilities flags). For instance, Python MySQLdb 1.2.2 throws an exception when the version number is not in X.Y.ZZ format; MySQL .NET connector 6.3.x fails internally on version numbers 1.x along with a certain combination of flags, etc. To workaround that, you can use mysql_version_string directive and have searchd report a different version to clients connecting over MySQL protocol. (By default, it reports its own version.)


mysql_version_string = 5.0.37

12.4.32. rt_flush_period

RT indexes RAM chunk flush check period, in seconds. Optional, default is 10 hours. Introduced in version 2.0.1-beta.

Actively updated RT indexes that however fully fit in RAM chunks can result in ever-growing binlogs, impacting disk use and crash recovery time. With this directive the search daemon performs periodic flush checks, and eligible RAM chunks can get saved, enabling consequential binlog cleanup. See Section 4.4, “Binary logging” for more details.


rt_flush_period = 3600 # 1 hour

12.4.33. thread_stack

Per-thread stack size. Optional, default is 1M. Introduced in version 2.0.1-beta.

In the workers = threads mode, every request is processed with a separate thread that needs its own stack space. By default, 1M per thread are allocated for stack. However, extremely complex search requests might eventually exhaust the default stack and require more. For instance, a query that matches a thousands of keywords (either directly or through term expansion) can eventually run out of stack. Previously, that resulted in crashes. Starting with 2.0.1-beta, searchd attempts to estimate the expected stack use, and blocks the potentially dangerous queries. To process such queries, you can either the thread stack size by using the thread_stack directive (or switch to a different workers setting if that is possible).

A query with N levels of nesting is estimated to require approximately 30+0.16*N KB of stack, meaning that the default 64K is enough for queries with upto 250 levels, 150K for upto 700 levels, etc. If the stack size limit is not met, searchd fails the query and reports the required stack size in the error message.


thread_stack = 256K

12.4.34. expansion_limit

The maximum number of expanded keywords for a single wildcard. Optional, default is 0 (no limit). Introduced in version 2.0.1-beta.

When doing substring searches against indexes built with dict = keywords enabled, a single wildcard may potentially result in thousands and even millions of matched keywords (think of matching 'a*' against the entire Oxford dictionary). This directive lets you limit the impact of such expansions. Setting expansion_limit = N restricts expansions to no more than N of the most frequent matching keywords (per each wildcard in the query).


expansion_limit = 16

12.4.35. watchdog

Threaded server watchdog. Optional, default is 1 (watchdog enabled). Introduced in version 2.0.1-beta.

A crashed query in threads multi-processing mode (workers = threads) can take down the entire server. With watchdog feature enabled, searchd additionally keeps a separate lightweight process that monitors the main server process, and automatically restarts the latter in case of abnormal termination. Watchdog is enabled by default.


watchdog = 0 # disable watchdog

12.4.36. prefork_rotation_throttle

Delay between restarting preforked children on index rotation, in milliseconds. Optional, default is 0 (no delay). Introduced in version 2.0.2-beta.

When running in workers = prefork mode, every index rotation needs to restart all children to propagate the newly loaded index data changes. Restarting all of them at once might put excessive strain on CPU and/or network connections. (For instance, when the application keeps a bunch of open persistent connections to different children, and all those children restart.) Those bursts can be throttled down with prefork_rotation_throttle directive. Note that the children will be restarted sequentially, and thus "old" results might persist for a few more seconds. For instance, if prefork_rotation_throttle is set to 50 (milliseconds), and there are 30 children, then the last one would only be actually restarted 1.5 seconds (50*30=1500 milliseconds) after the "rotation finished" message in the searchd event log.


prefork_rotation_throttle = 50 # throttle children restarts by 50 msec each

12.4.37. sphinxql_state

Path to a file where current SphinxQL state will be serialized. Available since version 2.1.1-beta.

On daemon startup, this file gets replayed. On eligible state changes (eg. SET GLOBAL), this file gets rewritten automatically. This can prevent a hard-to-diagnose problem: If you load UDF functions, but Sphinx crashes, when it gets (automatically) restarted, your UDF and global variables will no longer be available; using persistent state helps a graceful recovery with no such surprises.


sphinxql_state = uservars.sql

12.4.38. ha_ping_interval

Interval between agent mirror pings, in milliseconds. Optional, default is 1000. Added in 2.1.1-beta.

For a distributed index with agent mirrors in it (see more in ???), master sends all mirrors a ping command during the idle periods. This is to track the current agent status (alive or dead, network roundtrip, etc). The interval between such pings is defined by this directive.

To disable pings, set ha_ping_interval to 0.


ha_ping_interval = 0

12.4.39. ha_period_karma

Agent mirror statistics window size, in seconds. Optional, default is 60. Added in 2.1.1-beta.

For a distributed index with agent mirrors in it (see more in ???), master tracks several different per-mirror counters. These counters are then used for failover and balancing. (Master picks the best mirror to use based on the counters.) Counters are accumulated in blocks of ha_period_karma seconds.

After beginning a new block, master may still use the accumulated values from the previous one, until the new one is half full. Thus, any previous history stops affecting the mirror choice after 1.5 times ha_period_karma seconds at most.

Despite that at most 2 blocks are used for mirror selection, upto 15 last blocks are actually stored, for instrumentation purposes. They can be inspected using SHOW AGENT STATUS statement.


ha_period_karma = 120

12.4.40. persistent_connections_limit

The maximum # of simultaneous persistent connections to remote persistent agents. Each time connecting agent defined under 'agent_persistent' we try to reuse existing connection (if any), or connect and save the connection for the future. However we can't hold unlimited # of such persistent connections, since each one holds a worker on agent size (and finally we'll receive the 'maxed out' error, when all of them are busy). This very directive limits the number. It affects the num of connections to each agent's host, across all distributed indexes.

It is reasonable to set the value equal or less than max_children option of the agents.


persistent_connections_limit = 29 # assume that each host of agents has max_children = 30 (or 29).

12.4.41. rt_merge_iops

A maximum number of I/O operations (per second) that the RT chunks merge thread is allowed to start. Optional, default is 0 (no limit). Added in 2.1.1-beta.

This directive lets you throttle down the I/O impact arising from the OPTIMIZE statements. It is guaranteed that all the RT optimization activity will not generate more disk iops (I/Os per second) than the configured limit. Modern SATA drives can perform up to around 100 I/O operations per second, and limiting rt_merge_iops can reduce search performance degradation caused by merging.


rt_merge_iops = 40

12.4.42. rt_merge_maxiosize

A maximum size of an I/O operation that the RT chunks merge thread is allowed to start. Optional, default is 0 (no limit). Added in 2.1.1-beta.

This directive lets you throttle down the I/O impact arising from the OPTIMIZE statements. I/Os bigger than this limit will be broken down into 2 or more I/Os, which will then be accounted as separate I/Os with regards to the rt_merge_iops limit. Thus, it is guaranteed that all the optimization activity will not generate more than (rt_merge_iops * rt_merge_maxiosize) bytes of disk I/O per second.


rt_merge_maxiosize = 1M

12.4.43. predicted_time_costs

Costs for the query time prediction model, in nanoseconds. Optional, default is "doc=64, hit=48, skip=2048, match=64" (without the quotes). Added in 2.1.1-beta.

Terminating queries before completion based on their execution time (via either SetMaxQueryTime() API call, or SELECT ... OPTION max_query_time SphinxQL statement) is a nice safety net, but it comes with an inborn drawback: indeterministic (unstable) results. That is, if you repeat the very same (complex) search query with a time limit several times, the time limit will get hit at different stages, and you will get different result sets.

Starting with 2.1.1-beta, there is a new option, SELECT ... OPTION max_predicted_time, that lets you limit the query time and get stable, repeatable results. Instead of regularly checking the actual current time while evaluating the query, which is indeterministic, it predicts the current running time using a simple linear model instead:

predicted_time =
    doc_cost * processed_documents +
    hit_cost * processed_hits +
    skip_cost * skiplist_jumps +
    match_cost * found_matches

The query is then terminated early when the predicted_time reaches a given limit.

Of course, this is not a hard limit on the actual time spent (it is, however, a hard limit on the amount of processing work done), and a simple linear model is in no way an ideally precise one. So the wall clock time may be either below or over the target limit. However, the error margins are quite acceptable: for instance, in our experiments with a 100 msec target limit the majority of the test queries fell into a 95 to 105 msec range, and all of the queries were in a 80 to 120 msec range. Also, as a nice side effect, using the modeled query time instead of measuring actual run time results in somewhat less gettimeofday() calls, too.

No two server makes and models are identical, so predicted_time_costs directive lets you configure the costs for the model above. For convenience, they are integers, counted in nanoseconds. (The limit in max_predicted_time is counted in milliseconds, and having to specify cost values as 0.000128 ms instead of 128 ns is somewhat more error prone.) It is not necessary to specify all 4 costs at once, as the missed one will take the default values. However, we strongly suggest to specify all of them, for readability.


predicted_time_costs = doc=128, hit=96, skip=4096, match=128

12.4.44. shutdown_timeout

searchd --stopwait wait time, in seconds. Optional, default is 3 seconds. Added in 2.2.1-beta.

When you run searchd --stopwait your daemon needs to perform some activities before stopping like finishing queries, flushing RT RAM chunk, flushing attributes and updating binlog. And it requires some time. searchd --stopwait will wait up to shutdown_time seconds for daemon to finish its jobs. Suitable time depends on your index size and load.


shutdown_timeout = 5 # wait for up to 5 seconds

12.4.45. ondisk_attrs_default

Instance-wide defaults for ondisk_attrs directive. Optional, default is 0 (all attributes are loaded in memory). This directive lets you specify the default value of ondisk_attrs for all indexes served by this copy of searchd. Per-index directives take precedence, and will overwrite this instance-wide default value, allowing for fine-grain control.

12.4.46. query_log_min_msec

Limit (in milliseconds) that prevents the query from being written to the query log. Optional, default is 0 (all queries are written to the query log). This directive specifies that only queries with execution times that exceed the specified limit will be logged.

12.4.47. agent_connect_timeout

Instance-wide defaults for agent_connect_timeout parameter. The last defined in distributed (network) indexes.

12.4.48. agent_query_timeout

Instance-wide defaults for agent_query_timeout parameter. The last defined in distributed (network) indexes, or also may be overrided per-query using OPTION clause.

12.4.49. agent_retry_count

Integer, specifies how many times sphinx will try to connect and query remote agents in distributed index before reporting fatal query error. Default is 0 (i.e. no retries). This value may be also specified on per-query basis using 'OPTION retry_count=XXX' clause. If per-query option exists, it will override the one specified in config.

12.4.50. agent_retry_delay

Integer, in milliseconds. Specifies the delay sphinx rest before retrying to query a remote agent in case it fails. The value has sense only if non-zero agent_retry_count or non-zero per-query OPTION retry_count specified. Default is 500. This value may be also specified on per-query basis using 'OPTION retry_delay=XXX' clause. If per-query option exists, it will override the one specified in config.

12.5. Common section configuration options

12.5.1. lemmatizer_base

Lemmatizer dictionaries base path. Optional, default is /usr/local/share (as in --datadir switch to ./configure script). Added in version 2.1.1-beta.

Our lemmatizer implementation (see Section 12.2.6, “morphology” for a discussion of what lemmatizers are) is dictionary driven. lemmatizer_base directive configures the base dictionary path. File names are hardcoded and specific to a given lemmatizer; the Russian lemmatizer uses ru.pak dictionary file. The dictionaries can be obtained from the Sphinx website.


lemmatizer_base = /usr/local/share/sphinx/dicts/

12.5.2. on_json_attr_error

What to do if JSON format errors are found. Optional, default value is ignore_attr (ignore errors). Applies only to sql_attr_json attributes. Added in 2.1.1-beta.

By default, JSON format errors are ignored (ignore_attr) and the indexer tool will just show a warning. Setting this option to fail_index will rather make indexing fail at the first JSON format error.


on_json_attr_error = ignore_attr

12.5.3. json_autoconv_numbers

Automatically detect and convert possible JSON strings that represent numbers, into numeric attributes. Optional, default value is 0 (do not convert strings into numbers). Added in 2.1.1-beta.

When this option is 1, values such as "1234" will be indexed as numbers instead of strings; if the option is 0, such values will be indexed as strings. This conversion applies to any data source, that is, JSON attributes originating from either SQL or XMLpipe2 sources will all be affected.


json_autoconv_numbers = 1

12.5.4. json_autoconv_keynames

Whether and how to auto-convert key names within JSON attributes. Known value is 'lowercase'. Optional, default value is unspecified (do not convert anything). Added in 2.1.1-beta.

When this directive is set to 'lowercase', key names within JSON attributes will be automatically brought to lower case when indexing. This conversion applies to any data source, that is, JSON attributes originating from either SQL or XMLpipe2 sources will all be affected.


json_autoconv_keynames = lowercase

12.5.5. rlp_root

Path to the RLP root folder. Mandatory if RLP is used. Added in 2.2.1-beta.


rlp_root = /home/myuser/RLP

12.5.6. rlp_environment

RLP environment configuration file. Mandatory if RLP is used. Added in 2.2.1-beta.


rlp_environment = /home/myuser/RLP/rlp-environment.xml

12.5.7. rlp_max_batch_size

Maximum total size of documents batched before processing them by the RLP. Optional, default is 51200. Do not set this value to more than 10Mb because sphinx splits large documents to 10Mb chunks before processing them by the RLP. This option has effect only if morphology = rlp_chinese_batched is specified. Added in 2.2.1-beta.


rlp_max_batch_size = 100k

12.5.8. rlp_max_batch_docs

Maximum number of documents batched before processing them by the RLP. Optional, default is 50. This option has effect only if morphology = rlp_chinese_batched is specified. Added in 2.2.1-beta.


rlp_max_batch_docs = 100

12.5.9. plugin_dir

Trusted location for the dynamic libraries (UDFs). Optional, default is empty (no location). Introduced in version 2.0.1-beta.

Specifies the trusted directory from which the UDF libraries can be loaded. Requires workers = thread to take effect.


plugin_dir = /usr/local/sphinx/lib

Appendix A. Sphinx revision history

Table of Contents

A.1. Version 2.2.10-release, 07 sep 2015
A.2. Version 2.2.9-release, 16 apr 2015
A.3. Version 2.2.8-release, 09 mar 2015
A.4. Version 2.2.7-release, 20 jan 2015
A.5. Version 2.2.6-release, 13 nov 2014
A.6. Version 2.2.5-release, 06 oct 2014
A.7. Version 2.2.4-release, 11 sep 2014
A.8. Version 2.2.3-beta, 13 may 2014
A.9. Version 2.2.2-beta, 11 feb 2014
A.10. Version 2.2.1-beta, 13 nov 2013
A.11. Version 2.1.9-release, 03 jul 2014
A.12. Version 2.1.8-release, 28 apr 2014
A.13. Version 2.1.7-release, 30 mar 2014
A.14. Version 2.1.6-release, 24 feb 2014
A.15. Version 2.1.5-release, 22 jan 2014
A.16. Version 2.1.4-release, 18 dec 2013
A.17. Version 2.1.3-release, 12 nov 2013
A.18. Version 2.1.2-release, 10 oct 2013
A.19. Version 2.1.1-beta, 20 feb 2013
A.20. Version 2.0.11-dev, xx xxx xxxx
A.21. Version 2.0.10-release, 22 jan 2014
A.22. Version 2.0.9-release, 26 aug 2013
A.23. Version 2.0.8-release, 26 apr 2013
A.24. Version 2.0.7-release, 26 mar 2013
A.25. Version 2.0.6-release, 22 oct 2012
A.26. Version 2.0.5-release, 28 jul 2012
A.27. Version 2.0.4-release, 02 mar 2012
A.28. Version 2.0.3-release, 23 dec 2011
A.29. Version 2.0.2-beta, 15 nov 2011
A.30. Version 2.0.1-beta, 22 apr 2011
A.31. Version 1.10-beta, 19 jul 2010
A.32. Version 0.9.9-release, 02 dec 2009
A.33. Version 0.9.9-rc2, 08 apr 2009
A.34. Version 0.9.9-rc1, 17 nov 2008
A.35. Version, 30 oct 2008
A.36. Version 0.9.8, 14 jul 2008
A.37. Version 0.9.7, 02 apr 2007
A.38. Version 0.9.7-rc2, 15 dec 2006
A.39. Version 0.9.7-rc1, 26 oct 2006
A.40. Version 0.9.6, 24 jul 2006
A.41. Version 0.9.6-rc1, 26 jun 2006

A.1. Version 2.2.10-release, 07 sep 2015

New features and changes

  • added #2310, --replay-flags=ignore-open-errors switch to replay binlogs even if some files are missing

  • added #2234, support for empty string values (stringattr='') in WHERE clause

  • added #2233, support for IN() filters with string values

  • added #2232, string collation support in SELECT expressions

  • added #2121, "where flt<>val" support, "where fltcol=intval" and "where fltcol!=intval" conditions

  • added #2119, new indexer exit code 2 on a --rotate failure

  • fixed #2207, unified min_prefix_len, min_infix_len behavior between RT and plain indexes

  • fixed #2020, unified (and greatly shortened) the list of SphinxQL reserved keywords between indexer checks, SphinxQL parser checks, and the documentation

Major bug fixes

  • fixed #2251, expressions dependent on aggregation results (eg. as in SELECT MAX(id) m1, m1+10 m2) were not computed properly in RT indexes

  • fixed #2248, LENGTH() was 2x off for 64-bit MVA attributes

  • fixed #2146, OPTIMIZE could occasionally break big RT indexes (by violating 4/16 GB string/MVA per chunk size limits)

  • fixed #2118, multi-wordforms with clashing prefixes were processed in a wrong order

  • fixed #1926, disabled and later re-enabled indexes were not picked up again by searchd on SIGHUP

Minor bug fixes

  • fixed #2312, using FACTORS() along with a subtree cache could crash (because on wrong qpos values from the cache passed to the ranker)

  • fixed #2310, comparing a non-existent JSON field with a string constant (as in jcol.some_typo='abc') could crash

  • fixed #2309, UDFs with BIGINT return were saved without a type into sphinxql_state file

  • fixed #2305, punctuation chars not mentioned in charset_table could still occasionally affect term position in the query

  • fixed #2303, a combination of hitless_words, lemmatizer_all, and a phrase operator could match a wrong result set

  • fixed #2301, searchd could sometimes crash on shutdown (at pid file unlink()) if the config was reloaded

  • fixed #2296, wordforms with multiple destination tokens broke snippet highlighting

  • fixed #2290, error in the middle of a multi-query batch did not abort SphinxQL packet, causing problems with some MySQL drivers like PHP mysqlnd

  • fixed #2286, multi-quries with different string filters were incorrectly considered identical

  • fixed #2280, HTML stripper incorrectly parsed hexadecimal NCRs (like eg. #xC0)

  • fixed #2273, a bit better error message when OPTIMIZE fails on a too big chunk

  • fixed #2258, some ranking FACTORS() were off when lemmatizer expansions yielded duplicate terms

  • fixed #2257, OR operator over conditional operators could crash

  • fixed #2242, added whitespaces support to SNIPPET() before_match/after_matches options, and fixed the handling of repeated %PASSAGE_ID% macros

  • fixed #2238, added a few safeguards to prevent crashes/freezes on loading damaged RT RAM chunks

  • fixed #2237, ATTACH-ing a part of a distributed index did not correctly invalidate it, could crash

  • fixed #2235, UPDATE ... OPTION strict=1 did not with plain indexes

  • fixed #2225, searchd crashed on startup if agent host string was empty

  • fixed #2127, indextool did not handle RT indexes with updated JSON attributes in them

  • fixed #2117, GEODIST() calls with hash {in=deg,out=mi} arguments on a distributed index did not parse correctly

  • fixed #2113, @@relaxed could occasonally crash ceratin complex queries

  • fixed #2106, using GROUP N BY with a custom ranker and FACTORS() caused crashes and memory leaks

  • fixed #2093, wildcard character at the end of the keyword could sometimes erroneously produce no matches

  • fixed #2088, NEAR operator with NOT argument could crash

  • fixed #1929, allowed `123abc` column names in SphinxQL SELECT (alas, they are still allowed in indexer)

  • fixed #1889, #1890, #1891, a few typo-style bugs in libsphinxclient sphinx_set_field_weights(), sphinx_set_index_weights(), sphinx_add_filter_entry()

  • fixed #1859, #2202, XML/TSV/CSV sources now works with control characters like EOF, and UTF BOM marks

  • fixed #1815, a number of SphinxSE issues (inet adress endpoint, too big numbers at MVA, and MVA inserts/replaces via SphinxQL)

  • fixed #1704, CONTAINS() now correctly handles polygons with duplicated points

  • fixed #1643, CRC32() is now properly evaluated as unsigned in BIGINT context

  • fixed #1567, #1747, #2245, column name quotation could fail in UDF expressions and distributed queries

  • fixed #1551, off-by-one blended keyword position errors in proximity queries (phrase, NEAR, etc)

  • fixed #1528, metaphone on too long (eg. Chinese or Japanese) strings crashed with a buffer overflow

  • fixed #1510, added an unknown field warning to SetFieldWeights() API call and SphinxQL OPTION field_weights

  • fixed #1367, remote agents (in distributed index) could not be accessed via UNIX sockets

  • fixed #1349, max_matches=0 was not handled correctly in libsphinxclient sphinx_set_limits()

  • fixed Github PR-1, SphinxSE TLS leak on table reopen

  • fixed searchd crash when trying to load a damaged index with an incorrect row count

  • fixed indextool MVA checks (an index error could sometimes be mistakenly reported)

A.2. Version 2.2.9-release, 16 apr 2015

Bug fixes

  • fixed #2228, removed searchd shutdown behavior on failed connection

  • fixed #2208, ZONESPANLIST() support for RT indexes

  • fixed #2203, legacy API SELECT list

  • fixed #2201, indextool false positive error on RT index

  • fixed #2201, crash with string comparison at expressions and expression ranker

  • fixed #2199, invalid packedfactors JSON output for index with stopwords

  • fixed #2197, TRUNCATE fails to remove disk chunk files after calling OPTIMIZE

  • fixed #2196, .NET connector issue (UTC_TIMESTAMP() support)

  • fixed #2190, incorrect GROUP BY outer JSON object

  • fixed #2176, agent used ha_strategy=random instead of specified in config

  • fixed #2144, query parser crash vs multiforms with heading numbers

  • fixed #2122, id64 daemon failed to load RT disk chunk with kill-list from id32 build

  • fixed #2120, aliased JSON elements support

  • fixed #1979, snippets generation and span length and lcs calculation in proximity queries

  • fixed truncated results (and a potential crash) vs long enough ZONESPANLIST() result

A.3. Version 2.2.8-release, 09 mar 2015

Minor features

  • added #2166, per agent HA strategy for distributed indexes

Bug fixes

  • fixed #2182, incorrect query results with multiple same destination wordforms

  • fixed #2181, improved error message on incorrect filters

  • fixed #2178, ZONESPAN operator for queries with more than two words

  • fixed #2172, incorrect results with field position fulltext operators

  • fixed #2171, some index options do not work for template indexes

  • fixed #2170, joined fields indexation with document id equals to 0

  • fixed #2110, crash on snippet generation

  • fixed WLCCS ranking factor computation

  • fixed memory leak on queries with ZONEs

A.4. Version 2.2.7-release, 20 jan 2015

Minor features

  • added #2112, string equal comparison support for IF() function (for JSON and string attributes)

  • added #2153, IN() support for mixed and top-level JSON arrays

Bug fixes

  • fixed #2158, crash at RT index after morphology changed to AOT after index was created

  • fixed #2155, stopwords got missed on disk chunk save at RT index

  • fixed #2151, agents statistics missed in case of huge amount of agents

  • fixed #2139, escape all special characters in JSON result set, according to RFC 4627

  • fixed #2123, no pid file created in x64 release built with vs2012

  • fixed #2115, indexer crash on wordforms with multiple destination keywords

  • fixed #2050, multi result set doesn't work without libmysqlclient

  • fixed #2003, lemmatize_XX_all handling of short and exact words

  • fixed #1912, reduce indextool memory usage during a check of a huge index

  • fixed off by one errors in filtering of BIGINT attributes

  • fixed seamless rotation in prefork mode

  • fixed snippets crash with blend chars at the beginning of a string

A.5. Version 2.2.6-release, 13 nov 2014

Bug fixes

  • fixed #2104, ALL()/ANY()/INDEXOF() support for distributed indexes

  • fixed #2102, show agent status misses warnings from agents

  • fixed #2100, crash of indexer while loading stopwords with tokenizer plugin

  • fixed #2098, arbitrary JSON subkeys and IS NULL for distributed indexes

  • fixed #2097, escaping of field-start modifier

  • fixed possibly memory leak in plugin creation function

  • indexation of duplicate documents

A.6. Version 2.2.5-release, 06 oct 2014

New minor features

  • added OPTION rand_seed which affects ORDER BY RAND()

Bug fixes

  • fixed #2042, indextool fails with field mask on 32+ fields

  • fixed #2031, wrong encoding with UnixODBC/Oracle source

  • fixed #2056, several bugs in RLP tokenizer

  • fixed #2054, SHOW THREADS hangs if queries in prefork mode

  • fixed #2057, WARNING at indexer on duplicated wordforms

  • fixed #2066, snippet generation with weight_order enabled

  • fixed exception parsing in queries

  • fixed crash in config parser

  • fixed MySQL protocol response when daemon maxed out

A.7. Version 2.2.4-release, 11 sep 2014

New major features

  • added ALTER RTINDEX rt1 RECONFIGURE which allows to change RT index settings on the fly

  • added SHOW INDEX idx1 SETTINGS statement

  • added ability to specify several destination forms for the same source wordform (as a result, N:M mapping is now available)

  • added blended chars support to exceptions

New minor features

Optimizations and removals

Bug fixes

  • fixed #2027, slow queries to multiple indexes with large kill-lists

  • fixed #2022, blend characters of matched word must not be outside of snippet passage

  • fixed #2021, output units in GEODIST() function

  • fixed #2018, different wildcard behaviour in RT and plain indexes

  • fixed #2005, aggregate functions improperly calculate aliased expressions

  • fixed #1972, daemon crashes trying to read a big (>8G) .spm file

  • fixed #1966, INTERVAL() function does not work with JSON fields

  • fixed #1963, GROUPBY() on JSON attributes sometimes yields NULL

  • fixed GROUPBY() on empty JSON arrays to return NULL instead of []

  • fixed buffer overrun when sizing packed factors (with way too many fields) in expression ranker

  • fixed cpu time logging for cases where work is done in child threads or agents

A.8. Version 2.2.3-beta, 13 may 2014

New features

Optimizations and removals

  • improved speed of concurrent insertion in RT indexes

  • removed max_matches config key

Bug fixes

  • fixed #1946, IN() function support for string attributes

  • fixed #1942, crash in SHOW THREADS command

  • fixed #1922, crash on snippet generation for queries with duplicated words

  • fixed #1919, TSV bitcount attributes indexation issue

  • fixed #1916, COUNT(*) with empty result set

  • fixed #1910, JSON parsing issue

  • fixed #1906, ZONE constraints for expanded terms

  • fixed #1904, race condition in RT indexes on saving disk chunk

  • fixed #1899, crash on CALL KEYWORDS

  • fixed #1893, searchd crashes on expressions like 'a<<(*!b)'

  • fixed #1884, crash with SNIPPET() function over distributed index

  • fixed #1883, crash at expanded keyword with hitless index

  • fixed #1870, crash on ORDER BY JSON attributes

  • fixed template index removing on rotation

A.9. Version 2.2.2-beta, 11 feb 2014

New features

  • added #1604, CALL KEYWORDS can show now multiple lemmas for a keyword


  • added ALTER for JSON/string/MVA attributes

  • added REMAP() function which surpasses SetOverride() API

  • added an argument to PACKEDFACTORS() to disable ATC calculation (syntax: PACKEDFACTORS({no_atc=1}))

  • added exact phrase query syntax

  • added flag '--enable-dl' to configure script which works with libmysqlclient, libpostgresql, libexpat, libunixobdc

  • added new plugin system: CREATE/DROP PLUGIN, SHOW PLUGINS, plugin_dir now in common, index/query_token_filter plugins

  • added ondisk_attrs support for RT indexes

  • added position shift operator to phrase operator

  • added possibility to add user-defined rankers (via plugins)

Optimizations, behavior changes, and removals

  • changed #1797, per-term statistics report (expanded terms fold to their respective substrings)

  • changed default thread_stack value to 1M

  • changed local directive in a distributed index which takes now a list (eg. local=shard1,shard2,shard3)

  • deprecated SetMatchMode() API call

  • deprecated SetOverride() API call

  • optimized infix searches for dict=keywords

  • optimized kill lists in plain and RT indexes

  • removed deprecated "address" and "port" config keys

  • removed deprecated CLI search and sql_query_info

  • removed deprecated charset_type and mssql_unicode

  • removed deprecated enable_star

  • removed deprecated ondisk_dict and ondisk_dict_default

  • removed deprecated str2ordinal attributes

  • removed deprecated str2wordcount attributes

  • removed support for client versions 0.9.6 and below

A.10. Version 2.2.1-beta, 13 nov 2013

Major new features

  • added ALTER TABLE that can add attributes to disk and RT indexes on the fly
  • added ATTACH support for non-empty RT target indexes
  • added Chinese segmentation with RLP (Rosette Linguistics platform) support
  • added English, German lemmatization support
  • added HAVING support to SELECT statement, filtering on aggregate values is now possible
  • added N-best GROUP BY extension to return more than 1 row per group
  • added RT index support for index_field_lengths=1, bitfield attributes, and multiforms
  • added CSV, TSV data sources
  • added full JSON attributes support, arbitrary JSON documents (with subobjects etc) can now be stored
  • added in-place JSON updates for scalar values
  • added index type=template directive (allows CALL KEYWORDS, CALL SNIPPETS)
  • added ondisk_attrs, ondisk_attrs_default directives that keep attributes on disk
  • added table functions mechanism, and REMOVE_REPEATS() table function
  • added support for arbitrary expressions in WHERE for DELETE queries

Ranking related features

  • added OPTION local_df=1, an option to aggregate IDFs over local indexes (shards)
  • added UDF XXX_reinit() method to reload UDFs with workers=prefork
  • added comma-separated syntax to OPTION idf, tfidf_unnormalized and tfidf_normalized flags
  • added lccs, wlccs, exact_order, min_gaps, and atc ranking factors
  • added sphinx_get_XXX_factors(), a faster interface to access PACKEDFACTORS() in UDFs
  • added support for exact_hit, exact_order field factors when using more than 32 fields (exact_hit, exact_order)

Instrumentation features

  • added DESCRIBE and --dumpheader support for tokencount attributes (generated by index_field_lengths=1 directive)
  • added RT index query profile, percentages, totals to SHOW PROFILE
  • added predicted_time, dist_predicted_time, fetched_docs, fetched_hits counters to SHOW META
  • added total_tokens and disk_bytes counters to SHOW INDEX STATUS

General features

  • added ALL(), ANY() and INDEXOF() functions for JSON subarrays
  • added MIN_TOP_WEIGHT(), MIN_TOP_SORTVAL() functions
  • added TOP() aggregate function to expression ranker
  • added a check for duplicated tail hit positions in indextool --check
  • added compact_in option to query_log_format=sphinxql
  • added distance units and calculation method options to GEODIST() function, optimized it a lot
  • added embedded stopwords/exceptions/wordforms to --dumpheader
  • added indexer --nohup and indextool --rotate switches to check index files before rotating them
  • added scientific notation support for JSON attributes (as per RFC 4627)
  • added several SphinxQL statements to fix MySQL Workbench connection issues (LIKE for session variables, etc.)
  • added shutdown_timeout directive to searchd config section
  • added signed values support for INTEGER() and UINT() function
  • added snippet generation options to SNIPPET() function
  • added string filter support in distributed queries, SphinxAPI, SphinxQL query log
  • added support for mixed distributed and local index queries (SELECT * FROM dist1,dist2,local3), and index_weights option for that case

Optimizations, behavior changes, and removals

  • optimized JSON attributes access (1.12x to 2.0x+ total query speedup depending on the JSON data)
  • optimized SELECT (1.02x to 3.5x speedup, depending on index schema size)
  • optimized UPDATE (up to 3x faster on big updates)
  • optimized away internal threads table mutex contention with workers=threads and 1000s of threads
  • changed [emptyword -foo] query behavior in cases when emptyword is a stopword or an overshort word, made such queries computable rather than erroneous
  • changed post-morphology wordforms behavior, now it works as 'if ( stem(token)==stem(abc) ) emit(def)'
  • changed the config defaults to id64, dict=keywords, charset_type=utf-8, enable_star=1, workers=threads, mem_limit=128M, rt_mem_limit=128M
  • changed the default SphinxAPI matching mode to SPH_MATCH_EXTENDED2
  • disallowed dashes in index names in API requests (just like in SphinxQL)
  • removed legacy xmlpipe data source v1, compat_sphinxql_magics directive, SetWeights() SphinxAPI call, and SPH_SORT_CUSTOM SphinxAPI mode

Bug fixes

  • fixed #1734, unquoted literal in json subscript could cause a crash, returns 'unknown column' now.
  • fixed #1683, under certain conditions stopwords were not taken into account in RT indexes
  • fixed #1648, #1644, when using AOT lemmas with snippet generation, not all the forms got highlighted
  • fixed #1549, OPTIONidf=tfidf_normalized was ignored for distributed queries
  • fixed that ORDER BY RAND() was not affected by index_weights
  • fixed that float updates with integer values in SphinxQL mistakenly set the float to 0
  • fixed that predicted_time was not accumulated with dist_threads
  • fixed GROUP_CONCAT result length limit (was implicitly limited by 1024 bytes)
  • fixed agent query distribution in HA mirroring
  • fixed duplicates check for quorum operator, it works ok now for expanded keywords
  • fixed off-by-1 query positions of words in indexes with wordforms and blended characters
  • fixed wrong lcs and min_best_span_pos ranking factor values when any expansion (expand_keywords or lemmatize) occurred
  • fixed a crash while creating indexes with sql_joined_field

A.11. Version 2.1.9-release, 03 jul 2014

Bug fixes

  • fixed #1994, parsing of empty JSON arrays

  • fixed #1987, handling of index_exact_words with AOT morphology and infixes on

  • fixed #1984, teaching HTML parser to handle hex numbers

  • fixed #1983, master and agents networking issue

  • fixed #1977, escaping of characters doens't work with exceptions

  • fixed #1968, parsing of WEIGHT() function (queries to distributed indexes affected)

A.12. Version 2.1.8-release, 28 apr 2014

Bug fixes

  • fixed #1937, crash at SENTENCE operator

  • fixed #1933, quorum operator works incorrectly if it's number is exception

  • fixed #1932, fixed daemon index recovery after failed rotation

  • fixed #1923, crash at indexer with dict=keywords

  • fixed #1918, fixed crash while hitless words are used within fulltext operators which require hits

  • fixed #1878, daemon doesn't reset regexp_filter after rotation with seamless_rotate=0

  • fixed #1769, crash after unsuccessful INSERT at RT index

  • fixed #1682, field end modifier doesn't work with words containing blended chars

A.13. Version 2.1.7-release, 30 mar 2014

Bug fixes

  • fixed #1917, field limit propagation outside of group

  • fixed #1915, exact form passes to index skipping stopwords filter

  • fixed #1905, multiple lemmas at the end of a field

  • fixed #1903, indextool check mode for hitless indexes and indexes with large amount of documents

  • fixed #1902, crash on JSON field in the IN() function

  • fixed #1884, crash at SNIPPET() with local indexes at distributed index

  • fixed #1802, loading large keywords dictionary

  • fixed #1786, indextool fails to handle indexes with AOT morphology

  • fixed crash of daemon on logging extra large message

  • fixed expression engine: division by zero, log and sqrt() functions of non-positive arguments

  • fixed LCS and min_best_span_pos computation

  • fixed unnecessary escaping in JSON result set

  • fixed Quick Tour documentation chapter

A.14. Version 2.1.6-release, 24 feb 2014

Bug fixes

  • fixed #1857, crash in arabic stemmer

  • fixed #1875, fixed crash on adding documents with long words in dict=keyword index with morphology and infixes enabled

  • fixed #1876, crash on words with large codepoints and infix searches

  • fixed #1880, crash on multiquery with one incorrect query

  • fixed #1882, race of periodic and forced FLUSHing on an RT index

  • fixed #1881, quorum syntax with '.' as blended char

  • fixed evaluating of LCS by an expression ranker

  • fixed #1864, indexer crash on badly formed JSON, e.g. '[,1,2,3,4,]'

  • fixed #1853, incomplete ORDER BY JSON attribute in distributed indexes

  • fixed #1847, broken infix searches in RT indexes

  • fixed #1844, clash of mix cased attribute and field names at CSV source

  • fixed #1840, filter by @uservar in distributes indexes

  • fixed #1832,#1833,#1834, some big endianess issues

  • fixed #1830, loss of ondisk_attrs after rotation

  • fixed #1762, memory leak in regexp_filter

  • fixed #1759, indextool false positives on persistent MVA checking

  • fixed GROUP BY id

  • fixed crash on sending empty snippet result

  • fixed index corruption in UPDATE queries with non-existent attributes

A.15. Version 2.1.5-release, 22 jan 2014

Bug fixes

  • fixed #1848, infixes and morphology clash

  • fixed #1823, indextool fails to handle indexes with lemmatizer morphology

  • fixed #1799, crash in queries to distributed indexes with GROUP BY on multiple values

  • fixed #1718, expand_keywords option lost in disk chunks of RT indexes

  • fixed documentation on rt_flush_period

  • fixed network protocol issue which results in timeouts of libmysqlclient for big Sphinx responses

A.16. Version 2.1.4-release, 18 dec 2013

Bug fixes

  • fixed #1778, indexes with more than 255 attributes

  • fixed #1777, ORDER BY WEIGHT()

  • fixed #1796, missing results in queries with quorum operator of indexes with some lemmatizer

  • fixed #1780, incorrect results while querying indexes with wordforms, some lemmatizer and enable_star=1

  • fixed, SHOW PROFILE for fullscan queries

  • fixed, --with-re2 check

A.17. Version 2.1.3-release, 12 nov 2013

Bug fixes

  • fixed #1753, path to re2 sources could not be set using --with-re2, options --with-re2-libs and --with-re2-includes added to configure

  • fixed #1739, erroneous conversion of RAM chunk into disk chunk when loading id32 index with id64 binary

  • fixed #1738, unlinking RAM chunk when converting it to disk chunk

  • fixed #1710, unable to filter by attributes created by index_field_lengths=1

  • fixed #1716, random crash with with multiple running threads

  • fixed crash while querying index with lemmatizer and wordforms

A.18. Version 2.1.2-release, 10 oct 2013

New features

  • added FLUSH RAMCHUNK statement

  • added SHOW PLAN statement

  • added support for GROUP BY on multiple attributes

  • added BM25F() function to SELECT expressions (now works with the expression based ranker)

  • added indextool --fold command and -q switch

  • added JSON debug check for RT index RAM chunk

  • added LENGTH() function for MVA

  • added missing rt_attr_bool directive

  • added support for selecting over 250 columns via SphinxQL

  • deprecated custom sort mode, and str2ordinal and str2wordcount attribute types

  • optimized SELECT, UPDATE for indexes with many attributes (up to 3.5x speedup in extreme cases)

  • JSON attributes (up to 5-20% faster SELECTs using JSON objects)

  • optimized xmlpipe2 indexing (up to 9 times faster on some schemas)

Bug fixes

  • fixed #1684, COUNT(DISTINCT smth) with implicit GROUP BY returns correct value now

  • fixed #1672, exact token AOT vs lemma (indexer skips exact form of token that passed AOT through tokenizer)

  • fixed #1659, fail while loading empty infix dictionary with dict=keywords

  • fixed #1638, force explicit JSON type conversion for aggregate functions

  • fixed #1628, GROUP_CONCAT() and GROUPBY() support for distributed agents

  • fixed #1619, INTEGER() conversion function doesn't support signed integers

  • fixed #1615, global IDF vs exact term (=term) fixed global IDF for missed terms fixed SphinxQL global_idf=0 option

  • fixed #1607, now ignoring binlog when running daemon with --console flag

  • fixed #1606, hard interruption of the daemon by Ctrl+C (SIGINT) signal

  • fixed #1592, duplicates vs expression ranker

  • fixed #1578, SORT BY string attribute via API attr_asc \ attr_desc

  • fixed #1575, crash of daemon on MVA receive from agents with dist_threads enabled

  • fixed #1574, agent got kill list of local indexes of distributed index

  • fixed #1573, ranker expression vs expanded terms

  • fixed #1572, BM25F vs negative terms

  • fixed #1550, float got cut at full-text part of a query

  • fixed #1541, BM25F expression in distributes indexes

  • fixed #1508, #1522, distributed index query lasts up to agent_connect_timeout with epoll path

  • fixed #1508, master failed to connect waiting agents up to agent_connect_timeout time

  • fixed #1489, filtering by integer field in JSON using floating point precision

  • fixed #1485, index_exact_words vs keyword dict with infix

  • fixed #1484, INSERT into RT vs no JSON attribute

  • fixed #1478, memory leaks at daemon PACKEDFACTORS() as UDF argument, index query tokenizer, expression ranker SUM()

  • fixed #1470, broken UDF unpack (since r3738 UDF version 2)

  • fixed #1468, multiple conditions in WHERE for JSON attributes

  • fixed #1466, index_field_lengths vs XML data source

  • fixed #1463, daemon shutdown vs RT index optimize (added forced terminate of long merging operation)

  • fixed #1460, aggregate functions AVG(), MAX(), MIN(), SUM() do not work for JSON attributes

  • fixed #1459, BM25F doesn't work with field_string fields

  • fixed #1458, factors to copy field_tf at UDF

  • fixed #1450, garbage in JSON fields when selecting them from a RT index

  • fixed #1449, broken build on Mac OS X

  • fixed #1446, WEIGHT() did not work in SELECT expressions

  • fixed #1445, field-start/field-end modifiers did not work for star-expanded keywords

  • fixed #1443, morphology=lemmatizer_ru_all now works with index_exact_words=1 (exact forms can be matches)

  • fixed #1442, incorrect COUNT(*) value in queries to distributed indexes with implicit GROUP BY

  • fixed #1439, filters on float values in JSON issue, string values quoting issue

  • fixed #1399, filter error message on string attribute

  • fixed #1384, added possibility to define any own DSN line with source=mssql (like as in source=odbc)

  • fixed ATTACH vs wordforms or stopwords; after daemon was restarted this setting was getting lost in RT indexes

  • fixed balancing of agents in HA

  • fixed co-working of index_exact_word + AOT lemmatizer

  • fixed epoll invoking and turned on by default

  • fixed incorrect handling of wildcards in tokenizer

  • fixed infix indexing with dict=keywords

  • fixed max_predicted_time integer overflows

  • fixed memory error in tokenizer

  • fixed several memory leaks

  • fixed PACKEDFACTORS() to work in different GROUP BY queries

  • fixed preprocessor definitions for RE2 in VS solution

  • fixed rotation of global IDF for workers=threads and seamless_rotate=1

  • fixed rotation of old indexes

  • fixed RT kill list survives TRUNCATE and works in newly ATTACHed index

  • fixed saving id32 RT index with id64 daemon

  • fixed stemmer vs RT index INSERT

  • fixed string case error with JSON attributes in select list of a query

  • fixed TOP_COUNT usage in misc/suggest and updated to PHP 5.3 and UTF-8

A.19. Version 2.1.1-beta, 20 feb 2013

Major new features

New features

New SphinxQL features

Major behavior changes and optimizations

  • changed that UDFs are now allowed in fork/prefork modes via sphinxql_state startup script

  • changed that compat_sphinxql_magics now defaults to 0

  • changed that small enough exceptions, wordforms, stopwords files are now embedded into the index header

  • changed that rt_mem_limit can now be over 2 GB (bug #1059)

  • optimized tokenizer (upto 1.25x indexing and snippets speedup)

  • optimized multi-keyword searching (added skiplists)

  • optimized filtering and scan in several frequent cases (single-value, 2-arg, 3-arg WHERE clauses)

A.20. Version 2.0.11-dev, xx xxx xxxx

Bug fixes

A.21. Version 2.0.10-release, 22 jan 2014

Bug fixes

  • fixed #1778, SENTENCE and PARAGRAPH operators and infix stars clash

  • fixed #1774, stack overflow on parsing large expressions

  • fixed #1744, daemon failed to write to log file bigger than 4G

  • fixed #1705, expression ranker handling of indexes with more than 32 fields

  • fixed #1700, crash and cutoff in fullscan reverse_scan=1 queries

  • fixed #1698, proper handling of stopword with blended chars

  • fixed #1682, field end modifier and index_exact_words clash

  • fixed #1678, memory leak in SUM() function of an expression ranker

  • fixed #1670, updating of MVA attributes in distributed indexes via API

  • fixed #1662, EscapeString() API escapes '<' too now

  • fixed #1520, SetLimits() API documentation

  • fixed #1491, documentation: space character is prohibited in charset_table

  • fixed memory leak in expressions with max_window_hits

  • fixed rt_flush_period - less stricter internal check and more often flushes overall

A.22. Version 2.0.9-release, 26 aug 2013

Bug fixes

  • fixed #1655, special characters like ()?* were not processed correctly by exceptions

  • fixed #1651, CREATE FUNCTION can now be used with BIGINT return type

  • fixed #1649, incorrect warning message (about statistics mismatch) was returned when mixing wildcards and regular keywords

  • fixed #1603, passing MVA64 arguments to non-MVA functions caused unpredicted behavior and crashes (now explicitly forbidden)

  • fixed #1601, negative numbers in IN() clause caused a syntax error

  • fixed #1581, dict=keywords and sql_joined_field occasionally caused indexer to build corrupted indexes

  • fixed #1546, file descriptor leaked on index rotation (that eventually prevented searchd to reload indexes)

  • fixed #1537, COUNT(*) and compat_sphinxql_magics=0 via SphinxAPI caused an incorrect error message

  • fixed #1531, #1589, several matching and highlighting issues when using both blend_chars and multi-wordforms

  • fixed #1521, indextool --check did not handle empty RT MVA and gave an incorrect warning

  • fixed #1392, SphinxSE builds with MySQL 5.6 now

  • fixed #1346, NEAR handles duplicated keywords properly now

  • fixed #757, wordforms shared between multiple indexes with different tokenizer settings failed to load (they now load with a warning)

  • fixed that batch queries did not batch in some cases (because of internal expression alias issues)

  • fixed that CALL KEYWORDS occasionally gave incorrect error messages

  • fixed searchd crashes on ATTACHing plain indexes with MVAs

  • fixed several deadlocks and other threading issues

  • fixed incorrect sorting order with utf8_general_ci

  • fixed that in some cases incorrect attribute values were returned when using expression aliases

  • optimized xmlpipe2 indexing

  • added a warning for missed stopwords, exception, wordforms files on index load and in indextool --check

A.23. Version 2.0.8-release, 26 apr 2013

Bug fixes

  • fixed #1515, log strings over 2KB were clipped when query_log_format=plain

  • fixed #1514, RT index disk chunk lose attribute update on daemon restart

  • fixed #1512, crash while formatting log messages

  • fixed #1511, crash on indexing PostgreSQL data source with MVA attributes

  • fixed #1509, blend_chars vs incomplete multi-form and overshort

  • fixed #1504, RT binlog replay vs descending tid on update

  • fixed #1499, sql_field_str2wordcount actually is int, not string

  • fixed #1498, now working with exceptions starting with number too

  • fixed #1496, multiple destination keywords in wordform

  • fixed #1494, lost 'mod', '%' operations in select list. Also corrected few typers in the doc.

  • fixed #1490, expand_keywords vs prefix

  • fixed #1487, `id` in expression fixed

  • fixed #1483, snippets limits fix

  • fixed #1481, shebang config changes check on rotation

  • fixed #1479, port handling in PHP Sphinx API

  • fixed #1474, daemon crash at SphinxQL packet overflows max_packet_size

  • fixed #1472, crash on loading index to indextool for check

  • fixed #1465, expansion_limit got lost in index rotation

  • fixed #1427, #1506, utf8 3 and 4-bytes codepoints

  • fixed #1405, between with mixed int float values

A.24. Version 2.0.7-release, 26 mar 2013

Bug fixes

  • fixed #1475, memory leak in the expression parser

  • fixed #1457, error messages over 2KB were clipped

  • fixed #1454, searchd did not display an error message when the binlog path did not exist

  • fixed #1441, SHOW META in a query batch was returning the last non-batch error

  • fixed #1435, typo in the documentation

  • fixed #1430, rt_flush_period now works even with a disabled binlog

  • fixed #1427, overlong 4-byte UTF-8 codes in source text could cause indexer crashes or index corruption

  • fixed #1418, warnings from local index searches were lost with dist_threads>0

  • fixed #1417, crash handler now works on searchd startup stage, too (eg. to report index load time crashes)

  • fixed #1410, bad numerics like '123abc' now result in a proper SphinxQL error message

  • fixed #1404, a tiny memory leak in shared mutex

  • fixed #1394, race in --iostats caused incorrect I/O statistics in threaded modes

  • fixed #1391, QUORUM operator vs docinfo=inline returned wrong attribute values

  • fixed #1389, edge case in the ORDER operator caused occasionally searchd crashes

  • fixed #1382, query parts with field limits but without real keywords (like '@name {') are now simply ignored and no longer cause a query syntax error

  • fixed #1370, Windows indexer builds failed to fetch rows from MSSQL 2012

  • fixed #1368, ORDER BY RAND() did not work in RT indexes

  • fixed #1364, queries with hitless words could occasionally crash searchd

  • fixed #1363, '*' in charset_table was causing query syntax errors with enable_star=1

  • fixed #1353, added filtering by 'id' syntax (in addition to '@id') to SphinxSE

  • fixed #1346, fixed NEAR operator behavior vs duplicated keywords

  • fixed #1345, invalid PROXIMITY operator threshold now causes a query syntax error rather than unexpected search behavior

  • fixed #1343, misconfigured indexes with 0 full text fields are now explicitly forbidden

  • fixed #1342, specific error messages (from the preload stage) went missing when failing to load the indexes

  • fixed #1339, no warning on inconsistent word statistics

  • fixed #1335, typo in searchd help screen

  • fixed #1334, typo in SELECT documentation

  • fixed #1316, PHRASE operator did not match in a rare self-repeating document/query case

  • fixed #1297, letting queries complete gracefully instead of killing them off in seamless_rotate=1, workers=prefork case

  • fixed #1295, mentioned index naming requirements (proper identifier) in the FROM clause docs

  • fixed #1221, incorrect results when using @groupby in select list via SphinxAPI with compat_sphinxql_magics=0

  • fixed #1180, special SPZ chars occasionally leaking into snippets

  • fixed #1171, preforked children did not reload logs on SIGUSR1

  • fixed #1150, added support for `id` syntax in DELETE and parents in WHERE

  • fixed #1135, crashes when using MVA/strings attributes in expression ranker

  • fixed #1124, corrupted attributes after merging with an empty index

  • fixed #1090, SphinxSE snippets UDF updated to support MySQL 5.5

  • fixed #1041, added initial support for MVA updates (and other mutex protected things) on FreeBSD

  • fixed #999, fullscan returned empty result sets in mixed batches of fullscan and fulltext queries

  • fixed #921, document count/bytes 32bit overflow in indexer progress output

  • fixed #539, added processing suffix rules with dots in .affix file to spelldump

  • fixed #481, rotation did not work on Windows with preopen=1

  • fixed #268, added warnings about duplicate elements in xmlpipe2

  • fixed CSphStaticMutex (double initialization issue)

  • fixed documentation typo in SQL data sources

  • fixed too-late initialization of mutex at daemon

  • fixed that an instance of searchd resurrected by watchdog could leak resources and/or crash

  • added a console message about crashes during index loading at startup

  • added more debug info about failed index loading

A.25. Version 2.0.6-release, 22 oct 2012

Bug fixes

  • fixed #1322, J connector seems to be broken in rel20 , but works in trunk

  • fixed #1321, 'set names utf8' passes, but 'set names utf-8' doesn't because of syntax error '-'

  • fixed #1318, unhandled float comparison operators at filter

  • fixed #1317, FD leaks on thread seamless rotation

  • fixed #1313, crash on stopping daemon with incorrect RT index config

  • fixed #1306, 'jolly roger ;)', and '(((((((((9 brackets)' crashes searchd

  • fixed #1304, OS X debug compilation

  • fixed #1302, daemon random crashes on OS X

  • fixed #1301, indexer fails to send rotate signal

  • fixed #1300, lost index settings on attach

  • fixed #1299, daemon failed to rotate ATTACHed plain index

  • fixed #1289, SENTENCE or PARAGRAPH searching leak memory

  • fixes #1285, crash on running searchd with syslog and watchdog

  • fixed #1279, linking against explicitly disabled iconv. Also added --with-libexpat to config options, which sometimes required on systems without XML support

  • fixed #1278, broken unixODBC detection in configure script.

  • fixed #1277, broken build on some toolchains (like uClibc) where not defined LLONG_MIN, added ULLONG_MAX

  • fixed #1274, large spa ( >4GB ) file hasn't loaded

  • fixed #1269, crash at RT index with MVA from disk chunk previously updated

  • fixed #1268, unuseful warning removed

  • fixed #1264, string and MVA attributes aliasing works again

  • fixed #1254, its now possible to add indexes using --rotate

  • fixed #1249, SphinxQL unusable with PHP >= 5.4.5

  • fixed #1246, attributes of 100 character length not being saved

  • fixed #1234, case sensitive GROUP BY attribute

  • fixed #1216, typos, mem_limit default size and RT documentation

  • fixed #1148, RT documentation updated

  • fixed #1140, mem_limit default value

  • fixed #1138, updated documentation on sql_attr_string

  • fixed #1129, snippets vs empty files and empty filenames

  • fixed #1123, configure compatibility fix

  • fixed #1122, 64bit sql_range_step

  • fixed #1082, crashes and deadlocks on OS X with workers=threads and field leak of read-write lock

  • fixed #1081, select only count distinct attr1 but group by attr2

  • fixed #1064, mistake while working with timestamp functions

  • fixed #1043, inaccurate distinct count in case many indexes or distributed index

  • fixed #1042, arithmetic expressions overflow

  • fixed #1007, Russian stemming on big endian systems

  • fixed #986, asserting in SetRankingMode (PHP API)

  • fixed #975, incorrect ranking in some rare cases

  • fixed #967, Python API type checking error

  • fixed #934, API vs fullscan vs non-empty query

  • fixed #899, error if using SetFilterRange as HAVING from SQL

  • fixed #867, indexer accepts index names starting with digit or _

  • fixed #699, signed vs unsigned 64-bit DocIDs in SphinxQL

  • fixed #668, now ignoring single @ character (incorrect field operator)

  • fixed #611, @! operator vs non-existent field, updated documentation

  • fixed #412, multiple --filter arguments work as they should in search utility

  • fixed #108, support for system libstemmer library. The sources of libstemmer placed into libstemmer_c is preferred, but the system lib will be tried if no sources found

  • fixed ORDER BY output at query log with SphinxQL mode

  • fixed documentation entry about sql_joined_field

  • fixed sample config file

  • fixed x64 configurations for libstemmer

A.26. Version 2.0.5-release, 28 jul 2012

Bug fixes

  • fixed #1258, xmlpipe2 refused to index indexes with docinfo=inline

  • fixed #1257, legacy groupby modes vs dist_threads could occasionally return wrong search results (race condition)

  • fixed #1253, missing single-word query performance optimization (simplified ranker) vs prefix-expanded keywords vs dict=keywords

  • fixed #1252, COUNT(*) vs dist_threads could occasionally crash (race condition)

  • fixed #1251, missing expression support in the IN() function

  • fixed #1245, FlushAttributes mistakenly disabled by attr_flush_period=0 setting

  • fixed #1244, per-API-command (search, update, etc) statistics were not updated by SphinxQL requests

  • fixed #1243, misc issues (broken statistics, weights, checks) with very long keywords having blended parts in RT indexes

  • fixed #1240, embedded xmlpipe2 schema with more attributes than the sphinx.conf one caused indexer to crash

  • fixed #1239, memory leak when optimizing ABS(const) and other 1-arg functions

  • fixed #1228, #761, #1183, #1190, #1198, misc issues occasonally caused by MVA updates (crash on SaveAttributes; index rotation vs index name and TID; looped MVA updates; persistent MVA removal on rotation)

  • fixed #1227, API queries with SetGeoAnchor() were logged incorrectly in SphinxQL-format query logs (query_log_format=sphinxql)

  • fixed #1214, phrase query parsing issues when blend_chars contained a quote (") symbol

  • fixed #1213, attribute aliases were not recognized by the subsequent SELECT items

  • fixed #1212, indextool failed to check hitless keywords

  • fixed #1210, crash when indexing an index with joined fields only (no regular fields)

  • fixed #1209, xmlpipe_fixup_utf8 off by a byte on certain (pretty rare) malformed sequences

  • fixed #1202, various issues with CALL KEYWORDS vs RT indexes (crashes vs dict=keywords, missing modifiers in output)

  • fixed #1201, snippets vs query_mode=1 vs complex OR-queries could occasionally crash

  • fixed #1197, indexer running out of disk space could either crash, or fail to display a proper error message

  • fixed #1185, keywords with wildcards were not handled when highlighting the entire document

  • fixed #1184, indexer crash when ngram_chars was set, but ngram_len=0

  • fixed #1182, indexer crash on certain combinations of docinfo=inline vs bitfields

  • fixed #1181, GROUP BY on a MVA64 was truncated at 32 bits

  • fixed #1179, passage_boundary in snippets could get ignored (when highlighting the entire document)

  • fixed #1178, indexer could crash when charset_table specified out-of-bounds codes

  • fixed #1177, SPZ queries in snippets erroneously required passage_boundary option to be explicitly set

  • fixed #1176, multi-queries with a GROUP/ORDER BY on a string attributed crashed

  • fixed #1175, connection id mismatch in SphinxQL-format query logs

  • fixed #1167, nested parentheses in a full-text query could mistakenly reset preceding field or zone limit operator

  • fixed #1158, float range filters were not supported in a multi-query batch optimizer

  • fixed #1157, broken gcc-4.7 build

  • fixed #1156, empty result set instead of an error message when querying distributed indexes with compat_sphinxql_magic=1 and hitting an error

  • fixed #1143, dash after a number incorrectly parsed as an operator NOT

  • fixed #1137, searchd --stopwait hanged when the running instance crashed during shutdown

  • fixed #1136, high idle CPU load on systems without pthread_timed_lock()

  • fixed #1134, issues with prefork workers on systems without pthread_timed_lock()

  • fixed #1133, BuildExcerpts() on a distributed index with load_files did not distribute the jobs

  • fixed #1126, inaccurate hits sorting progress report on joined field indexing

  • fixed #1121, occasional bad entries (wrong characters or invalid SQL) in SphinxQL-format query log

  • fixed #1118, libsphinxclient requests failed when using SPH_RANK_EXPR

  • fixed #1073, improved handling of wordforms/multiforms rules referring to stopwords

  • fixed #1062, bigint filter ranges truncated when searching via SphinxQL

  • fixed #1052, SphinxSE range arguments with leading zeroes mistakenly parsed as octal

  • fixed #1011, negative MVA64 values mistakenly converted to positive (on indexing and/or output)

  • fixed #974, crash when logging queries over 2048 bytes with performance counters enabled

  • fixed #909, field-end modifier was ignored when followed by a non-whitespace syntax character (eg quote or bracket)

  • fixed #907, issue with bigint filtering (large positive or negative values)

  • fixed #906, #1074, Mac OS X 10.7.3 builds (conflicting memory allocation routines in Sphinx and external libs)

  • fixed #901, #1066, sending bigger request packets was broken in Python API

  • fixed #879, filters on weight-dependent expressions did not work correctly

  • fixed #553, default/missing port value was not handled properly in SetServer() API call

  • fixed that blended vs multiforms vs min_word_len could hang the query parser

  • fixed missing command-line switches documentation

A.27. Version 2.0.4-release, 02 mar 2012

Bug fixes

  • fixed #605, pack vs mysql compress

  • fixed #783, #862, #917, #985, #990, #1032 documentation bugs

  • fixed #885, bitwise AND/OR were not available via API

  • fixed #984, crash on indexing data with MAGIC_CODE_ZONE symbol

  • fixed #1004, RT index loses words from dictionary on segments merging with id64 enabled

  • fixed #1035, daemon doesn't properly handle FDs in case of socket overflow FD_SETSIZE ( *nix, preopen_indexes=0, worker=threads )

  • fixed #1038, quoted string for API select

  • fixed #1046, head SPZ overflow, snippet generation at non fast with SPZ

  • fixed #1048, distributed index can't sort \ filter because of missed attributes

  • fixed #1050, expression ranker vs agents

  • fixed #1051, added MVA64 support to UDFs

  • fixed #1054, max_query_time not handled properly on searching at RT index

  • fixed #1055, expansion_limit on searching at RT disk chunks

  • fixed #1057, daemon crashes on generating snippet with 0 documents provided

  • fixed #1060, load_files_scattered don't work

  • fixed #1065, libsphinxclient vs distribute index (agents)

  • fixed #1067, modifiers were not escaped in legacy query emulation

  • fixed #1071, master - agent communication got slower for a large query

  • fixed #1076, #1077, (redundant copying, and a possible mutex leak with uservars)

  • fixed #1078, blended vs FIELD_END

  • fixed #1084 crash \ index corruption on loading persist MVA

  • fixed #1091, RT attach of plain index with string \ MVA attributes prior regular attributes

  • fixed #1092, update got binloged with wrong TID

  • fixed #1098, crash on creating large expression

  • fixed #1099, cleaning up temporary files on fail of indexing

  • fixed #1100, missing xmlpipe_attr_bigint config directive

  • fixed #1101, now ignoring dashes within keywords when dash is not in charset_table

  • fixed #1103, ZONE operator incorrectly works on more than one keywords in a simple zone

  • fixed #1106, optimized WHERE id=value, WHERE id IN (values_list) clauses used in SELECT, UPDATE statements

  • fixed #1112, Sphinx doesn't work out-of-the-box because the collision of binlog_path option

  • fixed #1116, crash on FLUSH RTINDEX unknown-index-name

  • fixed #1117, occasional RT headers corruption (leading to crashes and/or missing results)

  • fixed #1119, missing expression ranker support in SphinxSE

  • fixed #1120, negative total_found, docs and hits counter on huge indexes

A.28. Version 2.0.3-release, 23 dec 2011

Bug fixes

  • fixed #1031, SphinxQL parsing syntax for MVA at insert \ replace statements

  • fixed #1027, stalls on attribute update in high-concurrency load

  • fixed #1026, daemon crash on malformed API command

  • fixed #1021, max_children option has been ignored with worker=threads

  • fixed #1020, crash on large attribute files loading

  • fixed #1014, crash on rotation when index has been removed from config file (worker=threads, *nix box)

  • fixed #1001, broken MVA files in RT index while saving disk chunk

  • fixed #995, crash on empty MVA updates

  • fixed #994, crash on daemon shutdown with seamless_rotate=0 and workers=threads

  • fixed #993, #998, crash on replay DELETE statement vs RT index with dict=keywords, fixed sequential INSERT into dict=keywords index right after INSERT into dict=crc index

  • fixed #991, crash on indexing mssql source with mssql_unicode enabled

  • fixed #983, #950, crash on host name lookup (SphinxSE with MySQL 5.5)

  • fixed #981, snippet inconsistency with allow_empty=0

  • fixed #980, broken index produced by index merge in rare cases

  • fixed #971, absent error message at master on agent "maxed out"

  • fixed #695, #815, #835, #866, malformed warnings in SphinxQL

  • fixed build of SphinxSE with MySQL 5.1

  • fixed crash log for 'fork' and 'prefork' workers

A.29. Version 2.0.2-beta, 15 nov 2011

Major new features

New features

  • added support for upto 256 searchable fields (was upto 32 before)

  • added FIBONACCI() function to expressions

  • added load_files_scattered option to snippets

  • added implicit attribute type promotions in multi-index result sets (#939)

  • added index names to indexer progress message on merge (#928)

  • added --replay-flags switch to searchd

  • added string attribute support and a few previously missing snippets options to SphinxSE

  • added previously missing Status(), SetConnectTimeout() API calls to Python API

  • added ORDER BY RAND() support to SELECT statement

  • added Sphinx version to Windows crash log

  • added RT index support to indextool --check (checks disk chunks only) (#877)

  • added prefork_rotation_throttle directive (preforked children restart delay, in milliseconds) (#873)

  • added on_file_field_error directive (different sql_file_field handling modes)

  • added manpages for all the programs

  • added syslog logging support

  • added sentence, paragraph, and zone support in html_strip_mode=retain mode to snippets

  • optimized search performance with many ZONE operators

  • improved suggestion tool (added Levenshtein limit, removed extra DB fetch)

  • improved sentence extraction (handles salutations, starting initials better now)

  • changed max_filter_values sanity check to 10M values

New SphinxQL features

  • added FLUSH RTINDEX statement

  • added dist_threads directive (parallel processing), load_files, load_files_scattered, batch syntax (multiple documents) support to CALL SNIPPETS statement

  • added OPTION comment='...' support to SELECT statement (#944)

  • added SHOW VARIABLES statement

  • added dummy handlers for SET TRANSACTION, SET NAMES, SELECT @@sysvar statements, and for sql_auto_is_null, sql_mode, and @@-style variables (like @@tx_isolation) in SET statement (better MySQL frameworks/connectors support)

  • added complete SphinxQL error logging (all errors are logged now, not just SELECTs)

  • improved SELECT statement syntax, made expressions aliases optional

Bug fixes

  • fixed #982, empty binlogs prevented upgraded daemon from starting up

  • fixed #978, libsphinxclient build failed on sparc/sparc64 solaris

  • fixed #977, eliminated (most) compiler warnings

  • fixed #969, broken expression MVA/string argument type check prevented IF(IN(mva..)) and other valid expressions from working

  • fixed #966, NOT IN @global_var syntax was not supported

  • fixed #958, mem_limit over INT_MAX was not clamped

  • fixed #954, UTF-8 snippets could crash on malformed data

  • fixed #951, UTF-8 snippets could hang on malformed data

  • fixed #947, bad float column type was reported via SphinxQL, breaking some clients

  • fixed #940, group-by with a small enough max_matches limit could occasionaly crash and/or sort wrongly

  • fixed #932, sending huge queries to agents occasionally failed (mainly on Windows)

  • fixed #926, snippets did not highlight widlcard matches with morphology enabled

  • fixed #918, crash logger did not report a proper query in dist_threads case

  • fixed #916, watchdog caused (endless) respawns if there was a crash during shutdown

  • fixed #904, attribute names were not forcibly case-folded in some API calls (eg. SetGroupDistinct)

  • fixed #902, query parser did not support stopword_step=0

  • fixed #897, network sockets dangled (open but unattended) while replaying binlog

  • fixed #855, allow_empty option in snippets did not always work correctly

  • fixed #854, indexing with many bigint attributes and docinfo=inline crashed

  • fixed #838, RT MVA insertion did not sort MVA values, caused matching issues

  • fixed #833, duplicate MVA values were not eliminated on update

  • fixed #832, certain (overshort/incorrect) documents crashed indexing MS SQL Unicode columns

  • fixed #829, query parser did not properly handle numerics with blend_chars

  • fixed #814, group-by string attributes in RT indexes dit not always work correctly

  • fixed #812, utf8 stemming produced unexpected stems on words with single-byte chars

  • fixed #808, huge queries crashed logging with query_log_format=sphinxql

  • fixed #806, stray single-star keyword crashed on querying

  • fixed #798, snippets ignored index_exact_words in query_mode

  • fixed #797, RT klist loader had an occasional off-by-one crash

  • fixed #791, preopen_indexes erroneously defaulted to 0 on Windows

  • fixed #790, huge dictionaries (over 4 GB) did not work

  • fixed #786, inplace_enable could occasionally corrupt the indexes

  • fixed #775, doc had a typo (soundex vs metaphone)

  • fixed #772, snippets duplicated blended chars on a SPZ boundary

  • fixed #762, query parser truncated digit-only keywords over 15 digits

  • fixed #736, query parser dit not properly handle blended/special char sequence

  • fixed #726, rotation of an index with a changed attribute count crashed

  • fixed #687, querying multiple indexes with index weights and sort-by expression produced incorrect (unadjusted) weights

  • fixed #585, (unsupported) string ordinals were silently zeroed out with docinfo=inline (instead of failing)

  • fixed #583, certain keywords could occasionally crash multiforms

  • fixed that concurrent MVA updates could crash

  • fixed that query parser did not ignore a pure blended token with a leading modifier

  • fixed that query parser did not properly handle a modifier followed by a dash

  • fixed that substring indexing with dict=crc did not support index_exact_words and zones

  • fixed that in a rare edge case common subtree cache could crash

  • fixed that empty result set returned the full schema (rather than SELECT-ed columns)

  • fixed that SphinxQL did not have a sanity check for (currently unsupported) result set schemas over 250 attributes

  • fixed that updates on regular indexes were not binlogged

  • fixed that multi-query optimization check for expressions did not handle multi-index case

  • fixed that SphinxSE did not build vs MySQL 5.5 release

  • fixed that proximity_bm25 ranker could yield incorrect weight on duplicated keywords

  • fixed that prefix expansion with dict=keyword occasionally crashed

  • fixed that strip_path did not work on RT disk chunks

  • fixed that exclude filters were not properly logged in query_log_format=sphinxql mode

  • fixed that plain string attribute check in indextool --check was broken

  • fixed that Java API did not let specify a connection timeout

  • fixed that ordinal and wordcount attributes could not be fetched via SphinxQL

  • fixed that in a rare edge case OR/ORDER would not match properly

  • fixed that sending (huge) query response did not handle EINTR properly

  • fixed that SPH04 ranker could yield incorrectly high weight in some cases

  • fixed that C API did not let zero out cutoff, max_matches settings

  • fixed that on a persistent connection there were occasionally issues handling signals while doing network reads/waitss

  • fixed that in a rare edge case (field start modifier in a certain complex query) querying crashed

  • fixed that snippets did not support dist_threads with load_files=0

  • fixed that in some extremely rare edge cases tiny parts of an index could end up corrupted with dict=keywords

  • fixed that field/zone conditions were not propagated to expanded keywords with dict=keywords

A.30. Version 2.0.1-beta, 22 apr 2011

New general features

New SphinxQL features

New command-line switches

  • added --print-queries switch to indexer that dumps SQL queries it runs

  • added --sighup-each switch to indexer that rotates indexes one by one

  • added --strip-path switch to searchd that skips file paths embedded in the index(-es)

  • added --dumpconfig switch to indextool that dumps an index header in sphinx.conf format

Major changes and optimizations

  • changed default preopen_indexes value to 1

  • optimized English stemmer (results in 1.3x faster snippets and indexing with morphology=stem_en)

  • optimized snippets, 1.6x general speedup

  • optimized const-list parsing in SphinxQL

  • optimized full-document highlighting CPU/RAM use

  • optimized binlog replay (improved performance on K-list update)

Bug fixes

  • fixed #767, joined fields vs ODBC sources

  • fixed #757, wordforms shared by indexes with different settings

  • fixed #733, loading of indexes in formats prior to v.14

  • fixed #763, occasional snippets failures

  • fixed #648, occasionally missed rotations on multiple SIGHUPs

  • fixed #750, an RT segment merge leading to false positives and/or crashes in some cases

  • fixed #755, zones in snippets output

  • fixed #754, stopwords counting at snippet passage generation

  • fixed #723, fork/prefork index rotation in children processes

  • fixed #696, freeze on zero threshold in quorum operator

  • fixed #732, query escaping in SphinxSE

  • fixed #739, occasional crashes in MT mode on result set send

  • fixed #746, crash with a named list in SphinxQL option

  • fixed #674, AVG vs group order

  • fixed #734, occasional crashes attempting to report NULL errors

  • fixed #829, tail hits within field position modifier

  • fixed #712, missing query_mode, force_all_words snippet option defaults in Java API

  • fixed #721, added dupe removal on RT batch INSERT/REPLACE

  • fixed #720, potential extraneous highlighting after a blended keyword

  • fixed #702, exceptions vs star search

  • fixed #666, ext2 query grouping vs exceptions

  • fixed #688, WITHIN GROUP ORDER BY related crash

  • fixed #660, multi-queue batches vs dist_threads

  • fixed #678, crash on dict=keywords vs xmlpipe vs min_prefix_len

  • fixed #596, ECHILD vs scripted configs

  • fixed #653, dependency in expression, sorting, grouping

  • fixed #661, concurrent distributed searches vs workers=threads

  • fixed #646, crash on status query via UNIX socket

  • fixed #589, libexpat.dll missing from some Win32 build types

  • fixed #574, quorum match order

  • fixed multiple documentation issues (#372, #483, #495, #601, #623, #632, #654)

  • fixed that ondisk_dict did not affect RT indexes

  • fixed that string attributes check in indextool --check was erroneously sensitive to string data order

  • fixed a rare crash when using BEFORE operator

  • fixed an issue with multiforms vs BuildKeywords()

  • fixed an edge case in OR operator (emitted wrong hits order sometimes)

  • fixed aliasing in docinfo accessors that lead to very rare crashes and/or missing results

  • fixed a syntax error on a short token at the end of a query

  • fixed id64 filtering and performance degradation with range filters

  • fixed missing rankers in libsphinxclient

  • fixed missing SPH04 ranker in SphinxSE

  • fixed column names in sql_attr_multi sample (works with example.sql now)

  • fixed an issue with distributed local+remote setup vs aggregate functions

  • fixed case sensitive columns names in RT indexes

  • fixed a crash vs strings from multiple indexes in result set

  • fixed blended keywords vs snippets

  • fixed secure_connection vs MySQL protocol vs MySQL.NET connector

  • fixed that Python API did not works with Python 2.3

  • fixed overshort_step vs snippets

  • fixed keyword staistics vs dist_threads searching

  • fixed multiforms vs query parsing (vs quorum)

  • fixed missed quorum words vs RT segments

  • fixed blended keywords occasionally skipping extra character when querying (eg "abc[]")

  • fixed Python API to handle int32 values

  • fixed prefix and infix indexing of joined fields

  • fixed MVA ranged query

  • fixed missing blended state reset on document boundary

  • fixed a crash on missing index while replaying binlog

  • fixed an error message on filter values overrun

  • fixed passage duplication in snippets in weight_order mode

  • fixed select clauses over 1K vs remote agents

  • fixed overshort accounting vs soft-whitespace tokens

  • fixed rotation vs workers=threads

  • fixed schema issues vs distributed indexes

  • fixed blended-escaped sequence parsing issue

  • fixed MySQL IN clause (values order etc)

  • fixed that post_index did not execute when 0 documents were succesfully indexed

  • fixed field position limit vs many hits

  • fixed that joined fields missed an end marker at field end

  • fixed that xxx_step settings were missing from .sph index header

  • fixed libsphinxclient missing request cleanup in sphinx_query() (eg after network errors)

  • fixed that index_weights were ignored when grouping

  • fixed multi wordforms vs blend_chars

  • fixed broken MVA output in SphinxQL

  • fixed a few RT leaks

  • fixed an issue with RT string storage going missing

  • fixed an issue with repeated queries vs dist_threads

  • fixed an issue with string attributes vs buffer overrun in SphinxQL

  • fixed unexpected character data warnings within ignored xmlpipe tags

  • fixed a crash in snippets with NEAR syntax query

  • fixed passage duplication in snippets

  • fixed libsphinxclient SIGPIPE handling

  • fixed libsphinxclient vs VS2003 compiler bug

A.31. Version 1.10-beta, 19 jul 2010

  • added RT indexes support (Chapter 4, Real-time indexes)

  • added prefork and threads support (workers directives)

  • added multi-threaded local searches in distributed indexes (dist_threads directive)

  • added common subquery cache (subtree_docs_cache, subtree_hits_cache directives)

  • added string attributes support (sql_attr_string, sql_field_string, xml_attr_string, xml_field_string directives)

  • added indexing-time word counter (sql_attr_str2wordcount, sql_field_str2wordcount directives)

  • added CALL SNIPPETS(), CALL KEYWORDS() SphinxQL statements

  • added field_weights, index_weights options to SphinxQL SELECT statement

  • added insert-only SphinxQL-talking tables to SphinxSE (connection='sphinxql://host[:port]/index')

  • added select option to SphinxSE queries

  • added backtrace on crash to searchd

  • added SQL+FS indexing, aka loading files by names fetched from SQL (sql_file_field directive)

  • added a watchdog in threads mode to searchd

  • added automatic row phantoms elimination to index merge

  • added hitless indexing support (hitless_words directive)

  • added --check, --strip-path, --htmlstrip, --dumphitlist ... --wordid switches to indextool

  • added --stopwait, --logdebug switches to searchd

  • added --dump-rows, --verbose switches to indexer

  • added "blended" characters indexing support (blend_chars directive)

  • added joined/payload field indexing (sql_joined_field directive)

  • added FlushAttributes() API call

  • added query_mode, force_all_words, limit_passages, limit_words, start_passage_id, load_files, html_strip_mode, allow_empty options, and %PASSAGE_ID% macro in before_match, after_match options to BuildExcerpts() API call

  • added @groupby/@count/@distinct columns support to SELECT (but not to expressions)

  • added query-time keyword expansion support (expand_keywords directive, SPH_RANK_SPH04 ranker)

  • added query batch size limit option (max_batch_queries directive; was hardcoded)

  • added SINT() function to expressions

  • improved SphinxQL syntax error reporting

  • improved expression optimizer (better constant handling)

  • improved dash handling within keywords (no longer treated as an operator)

  • improved snippets (better passage selection/trimming, around option now a hard limit)

  • optimized index format that yields ~20-30% smaller indexes

  • optimized sorting code (indexing time 1-5% faster on average; 100x faster in worst case)

  • optimized searchd startup time (moved .spa preindexing to indexer), added a progress bar

  • optimized queries against indexes with many attributes (eliminated redundant copying)

  • optimized 1-keyword queries (performace regression introduced in 0.9.9)

  • optimized SphinxQL protocol overheads, and performance on bigger result sets

  • optimized unbuffered attributes writes on index merge

  • changed attribute handling, duplicate names are strictly forbidden now

  • fixed that SphinxQL sessions could stall shutdown

  • fixed consts with leading minus in SphinxQL

  • fixed AND/OR precedence in expressions

  • fixed #334, AVG() on integers was not computed in floats

  • fixed #371, attribute flush vs 2+ GB files

  • fixed #373, segfault on distributed queries vs certain libc versions

  • fixed #398, stopwords not stopped in prefix/infix indexes

  • fixed #404, erroneous MVA failures in indextool --check

  • fixed #408, segfault on certain query batches (regular scan, plus a scan with MVA groupby)

  • fixed #431, occasional shutdown hangs in preforked workers

  • fixed #436, trunk checkout builds vs Solaris sh

  • fixed #440, escaping vs parentheses declared as valid in charset_table

  • fixed #442, occasional non-aligned free in MVA indexing

  • fixed #447, occasional crashes in MVA indexing

  • fixed #449, pconn busyloop on aborted clients on certain arches

  • fixed #465, build issue on Alpha

  • fixed #468, build issue in libsphinxclient

  • fixed #472, multiple stopword files failing to load

  • fixed #489, buffer overflow in query logging

  • fixed #493, Python API assertion after error returned from Query()

  • fixed #500, malformed MySQL packet when sending MVAs

  • fixed #504, SIGPIPE in libsphinxclient

  • fixed #506, better MySQL protocol commands support in SphinxQL (PING etc)

  • fixed #509, indexing ranged results from stored procedures

A.32. Version 0.9.9-release, 02 dec 2009

  • added Open, Close, Status calls to libsphinxclient (C API)

  • added automatic persistent connection reopening to PHP, Python APIs

  • added 64-bit value/range filters, fullscan mode support to SphinxSE

  • MAJOR CHANGE, our IANA assigned ports are 9312 and 9306 respectively (goodbye, trusty 3312)

  • MAJOR CHANGE, erroneous filters now fail with an error (were silently ignored before)

  • optimized unbuffered .spa writes on merge

  • optimized 1-keyword queries ranking in extended2 mode

  • fixed #441 (IO race in case of highly conccurent load on a preopened)

  • fixed #434 (distrubuted indexes were not searchable via MySQL protocol)

  • fixed #317 (indexer MVA progress counter)

  • fixed #398 (stopwords not removed from search query)

  • fixed #328 (broken cutoff)

  • fixed #250 (now quoting paths w/spaces when installing Windows service)

  • fixed #348 (K-list was not updated on merge)

  • fixed #357 (destination index were not K-list-filtered on merge)

  • fixed #369 (precaching .spi files over 2 GBs)

  • fixed #438 (missing boundary proximity matches)

  • fixed #371 (.spa flush in case of files over 2 GBs)

  • fixed #373 (crashes on distributed queries via mysql proto)

  • fixed critical bugs in hit merging code

  • fixed #424 (ordinals could be misplaced during indexing in case of bitfields etc)

  • fixed #426 (failing SE build on Solaris; thanks to Ben Beecher)

  • fixed #423 (typo in SE caused crash on SHOW STATUS)

  • fixed #363 (handling of read_timeout over 2147 seconds)

  • fixed #376 (minor error message mismatch)

  • fixed #413 (minus in SphinxQL)

  • fixed #417 (floats w/o leading digit in SphinxQL)

  • fixed #403 (typo in SetFieldWeights name in Java API)

  • fixed index rotation vs persistent connections

  • fixed backslash handling in SphinxQL parser

  • fixed uint unpacking vs. PHP 5.2.9 (possibly other versions)

  • fixed #325 (filter settings send from SphinxSE)

  • fixed #352 (removed mysql wrapper around close() in SphinxSE)

  • fixed #389 (display error messages through SphinxSE status variable)

  • fixed linking with port-installed iconv on OS X

  • fixed negative 64-bit unpacking in PHP API

  • fixed #349 (escaping backslash in query emulation mode)

  • fixed #320 (disabled multi-query route when select items differ)

  • fixed #353 (better quorum counts check)

  • fixed #341 (merging of trailing hits; maybe other ranking issues too)

  • fixed #368 (partially; @field "" caused crashes; now resets field limit)

  • fixed #365 (field mask was leaking on field-limited terms)

  • fixed #339 (updated debug query dumper)

  • fixed #361 (added SetConnectTimeout() to Java API)

  • fixed #338 (added missing fullscan to mode check in Java API)

  • fixed #323 (added floats support to SphinxQL)

  • fixed #340 (support listen=port:proto syntax too)

  • fixed #332 (\r is legal SphinxQL space now)

  • fixed xmlpipe2 K-lists

  • fixed #322 (safety gaps in mysql protocol row buffer)

  • fixed #313 (return keyword stats for empty indexes too)

  • fixed #344 (invalid checkpoints after merge)

  • fixed #326 (missing CLOCK_xxx on FreeBSD)

A.33. Version 0.9.9-rc2, 08 apr 2009

  • added IsConnectError(), Open(), Close() calls to Java API (bug #240)

  • added read_buffer, read_unhinted directives

  • added checks for build options returned by mysql_config (builds on Solaris now)

  • added fixed-RAM index merge (bug #169)

  • added logging chained queries count in case of (optimized) multi-queries

  • added GEODIST() function

  • added --status switch to searchd

  • added MySpell (OpenOffice) affix file support (bug #281)

  • added ODBC support (both Windows and UnixODBC)

  • added support for @id in IN() (bug #292)

  • added support for aggregate functions in GROUP BY (namely AVG, MAX, MIN, SUM)

  • added MySQL UDF that builds snippets using searchd

  • added write_buffer directive (defaults to 1M)

  • added xmlpipe_fixup_utf8 directive

  • added suggestions sample

  • added microsecond precision int64 timer (bug #282)

  • added listen_backlog directive

  • added max_xmlpipe2_field directive

  • added initial SphinxQL support to mysql41 handler, SELECT .../SHOW WARNINGS/STATUS/META are handled

  • added support for different network protocols, and mysql41 protocol

  • added fieldmask ranker, updated SphinxSE list of rankers

  • added mysql_ssl_xxx directives

  • added --cpustats (requires clock_gettime()) and --status switches to searchd

  • added performance counters, Status() API call

  • added overshort_step and stopword_step directives

  • added strict order operator (aka operator before, eg. "one << two << three")

  • added indextool utility, moved --dumpheader there, added --debugdocids, --dumphitlist options

  • added own RNG, reseeded on @random sort query (bug #183)

  • added field-start and field-end modifiers support (syntax is "^hello world$"; field-end requires reindex)

  • added MVA attribute support to IN() function

  • added AND, OR, and NOT support to expressions

  • improved logging of (optimized) multi-queries (now logging chained query count)

  • improved handshake error handling, fixed protocol version byte order (omg)

  • updated SphinxSE to protocol 1.22

  • allowed phrase_boundary_step=-1 (trick to emulate keyword expansion)

  • removed SPH_MAX_QUERY_WORDS limit

  • fixed CLI search vs documents missing from DB (bug #257)

  • fixed libsphinxclient results leak on subsequent sphinx_run_queries call (bug #256)

  • fixed libsphinxclient handling of zero max_matches and cutoff (bug #208)

  • fixed Java API over-64K string reads (eg. big snippets) in Java API (bug #181)

  • fixed Java API 2nd Query() after network error in 1st Query() call (bug #308)

  • fixed typo-class bugs in SetFilterFloatRange (bug #259), SetSortMode (bug #248)

  • fixed missing @@relaxed support (bug #276), fixed missing error on @nosuchfield queries, documented @@relaxed

  • fixed UNIX socket permissions to 0777 (bug #288)

  • fixed xmlpipe2 crash on schemas with no fields, added better document structure checks

  • fixed (and optimized) expr parser vs IN() with huge (10K+) args count

  • fixed double EarlyCalc() in fullscan mode (minor performance impact)

  • fixed phrase boundary handling in some cases (on buffer end, on trailing whitespace)

  • fixes in snippets (aka excerpts) generation

  • fixed inline attrs vs id64 index corruption

  • fixed head searchd crash on config re-parse failure

  • fixed handling of numeric keywords with leading zeroes such as "007" (bug #251)

  • fixed junk in SphinxSE status variables (bug #304)

  • fixed wordlist checkpoints serialization (bug #236)

  • fixed unaligned docinfo id access (bug #230)

  • fixed GetRawBytes() vs oversized blocks (headers with over 32K charset_table should now work, bug #300)

  • fixed buffer overflow caused by too long dest wordform, updated tests

  • fixed IF() return type (was always int, is deduced now)

  • fixed legacy queries vs. special chars vs. multiple indexes

  • fixed write-write-read socket access pattern vs Nagle vs delays vs FreeBSD (oh wow)

  • fixed exceptions vs query-parser issue

  • fixed late calc vs @weight in expressions (bug #285)

  • fixed early lookup/calc vs filters (bug #284)

  • fixed emulated MATCH_ANY queries (empty proximity and phrase queries are allowed now)

  • fixed MATCH_ANY ranker vs fields with no matches

  • fixed index file size vs inplace_enable (bug #245)

  • fixed that old logs were not closed on USR1 (bug #221)

  • fixed handling of '!' alias to NOT operator (bug #237)

  • fixed error handling vs query steps (step failure was not reported)

  • fixed querying vs inline attributes

  • fixed stupid bug in escaping code, fixed EscapeString() and made it static

  • fixed parser vs @field -keyword, foo|@field bar, "" queries (bug #310)

A.34. Version 0.9.9-rc1, 17 nov 2008

  • added min_stemming_len directive

  • added IsConnectError() API call (helps distingusih API vs remote errors)

  • added duplicate log messages filter to searchd

  • added --nodetach debugging switch to searchd

  • added blackhole agents support for debugging/testing (agent_blackhole directive)

  • added max_filters, max_filter_values directives (were hardcoded before)

  • added int64 expression evaluation path, automatic inference, and BIGINT() enforcer function

  • added crash handler for debugging (crash_log_path directive)

  • added MS SQL (aka SQL Server) source support (Windows only, mssql_winauth and mssql_unicode directives)

  • added indexer-side column unpacking feature (unpack_zlib, unpack_mysqlcompress directives)

  • added nested brackers and NOTs support to query language, rewritten query parser

  • added persistent connections support (Open() and Close() API calls)

  • added index_exact_words feature, and exact form operator to query language ("hello =world")

  • added status variables support to SphinxSE (SHOW STATUS LIKE 'sphinx_%')

  • added max_packet_size directive (was hardcoded at 8M before)

  • added UNIX socket support, and multi-interface support (listen directive)

  • added star-syntax support to BuildExcerpts() API call

  • added inplace inversion of .spa and .spp (inplace_enable directive, 1.5-2x less disk space for indexing)

  • added builtin Czech stemmer (morphology=stem_cz)

  • added IDIV(), NOW(), INTERVAL(), IN() functions to expressions

  • added index-level early-reject based on filters

  • added MVA updates feature (mva_updates_pool directive)

  • added select-list feature with computed expressions support (see SetSelect() API call, test.php --select switch), protocol 1.22

  • added integer expressions support (2x faster than float)

  • added multiforms support (multiple source words in wordforms file)

  • added legacy rankers (MATCH_ALL/MATCH_ANY/etc), removed legacy matching code (everything runs on V2 engine now)

  • added field position limit modifier to field operator (syntax: @title[50] hello world)

  • added killlist support (sql_query_killlist directive, --merge-killlists switch)

  • added on-disk SPI support (ondisk_dict directive)

  • added indexer IO stats

  • added periodic .spa flush (attr_flush_period directive)

  • added config reload on SIGHUP

  • added per-query attribute overrides feature (see SetOverride() API call); protocol 1.21

  • added signed 64bit attrs support (sql_attr_bigint directive)

  • improved HTML stripper to also skip PIs (<? ... ?>, such as <?php ... ?>)

  • improved excerpts speed (upto 50x faster on big documents)

  • fixed a short window of searchd inaccessibility on startup (started listen()ing too early before)

  • fixed .spa loading on systems where read() is 2GB capped

  • fixed infixes vs morphology issues

  • fixed backslash escaping, added backslash to EscapeString()

  • fixed handling of over-2GB dictionary files (.spi)

A.35. Version, 30 oct 2008

  • added configure script to libsphinxclient

  • changed proximity/quorum operator syntax to require whitespace after length

  • fixed potential head process crash on SIGPIPE during "maxed out" message

  • fixed handling of incomplete remote replies (caused over-degraded distributed results, in rare cases)

  • fixed sending of big remote requests (caused distributed requests to fail, in rare cases)

  • fixed FD_SET() overflow (caused searchd to crash on startup, in rare cases)

  • fixed MVA vs distributed indexes (caused loss of 1st MVA value in result set)

  • fixed tokenizing of exceptions terminated by specials (eg. "GPS AT&T" in extended mode)

  • fixed buffer overrun in stemmer on overlong tokens occasionally emitted by proximity/quorum operator parser (caused crashes on certain proximity/quorum queries)

  • fixed wordcount ranker (could be dropping hits)

  • fixed --merge feature (numerous different fixes, caused broken indexes)

  • fixed --merge-dst-range performance

  • fixed prefix/infix generation for stopwords

  • fixed ignore_chars vs specials

  • fixed misplaced F_SETLKW check (caused certain build types, eg. RPM build on FC8, to fail)

  • fixed dictionary-defined charsets support in spelldump, added \x-style wordchars support

  • fixed Java API to properly send long strings (over 64K; eg. long document bodies for excerpts)

  • fixed Python API to accept offset/limit of 'long' type

  • fixed default ID range (that filtered out all 64-bit values) in Java and Python APIs

A.36. Version 0.9.8, 14 jul 2008


  • added support for 64-bit document and keyword IDs, --enable-id64 switch to configure

  • added support for floating point attributes

  • added support for bitfields in attributes, sql_attr_bool directive and bit-widths part in sql_attr_uint directive

  • added support for multi-valued attributes (MVA)

  • added metaphone preprocessor

  • added libstemmer library support, provides stemmers for a number of additional languages

  • added xmlpipe2 source type, that supports arbitrary fields and attributes

  • added word form dictionaries, wordforms directive (and spelldump utility)

  • added tokenizing exceptions, exceptions directive

  • added an option to fully remove element contents to HTML stripper, html_remove_elements directive

  • added HTML entities decoder (with full XHTML1 set support) to HTML stripper

  • added per-index HTML stripping settings, html_strip, html_index_attrs, and html_remove_elements directives

  • added IO load throttling, max_iops and max_iosize directives

  • added SQL load throttling, sql_ranged_throttle directive

  • added an option to index prefixes/infixes for given fields only, prefix_fields and infix_fields directives

  • added an option to ignore certain characters (instead of just treating them as whitespace), ignore_chars directive

  • added an option to increment word position on phrase boundary characters, phrase_boundary and phrase_boundary_step directives

  • added --merge-dst-range switch (and filters) to index merging feature (--merge switch)

  • added mysql_connect_flags directive (eg. to reduce indexing time MySQL network traffic and/or time)

  • improved ordinals sorting; now runs in fixed RAM

  • improved handling of documents with zero/NULL ids, now skipping them instead of aborting

Search daemon

  • added an option to unlink old index on succesful rotation, unlink_old directive

  • added an option to keep index files open at all times (fixes subtle races on rotation), preopen and preopen_indexes directives

  • added an option to profile searchd disk I/O, --iostats command-line option

  • added an option to rotate index seamlessly (fully avoids query stalls), seamless_rotate directive

  • added HTML stripping support to excerpts (uses per-index settings)

  • added 'exact_phrase', 'single_passage', 'use_boundaries', 'weight_order 'options to BuildExcerpts() API call

  • added distributed attribute updates propagation

  • added distributed retries on master node side

  • added log reopen on SIGUSR1

  • added --stop switch (sends SIGTERM to running instance)

  • added Windows service mode, and --servicename switch

  • added Windows --rotate support

  • improved log timestamping, now with millisecond precision


  • added extended engine V2 (faster, cleaner, better; SPH_MATCH_EXTENDED2 mode)

  • added ranking modes support (V2 engine only; SetRankingMode() API call)

  • added quorum searching support to query language (V2 engine only; example: "any three of all these words"/3)

  • added query escaping support to query language, and EscapeString() API call

  • added multi-field syntax support to query language (example: "@(field1,field2) something"), and @@relaxed field checks option

  • added optional star-syntax ('word*') support in keywords, enable_star directive (for prefix/infix indexes only)

  • added full-scan support (query must be fully empty; can perform block-reject optimization)

  • added COUNT(DISTINCT(attr)) calculation support, SetGroupDistinct() API call

  • added group-by on MVA support, SetArrayResult() PHP API call

  • added per-index weights feature, SetIndexWeights() API call

  • added geodistance support, SetGeoAnchor() API call

  • added result set sorting by arbitrary expressions in run time (eg. "@weight+log(price)*2.5"), SPH_SORT_EXPR mode

  • added result set sorting by @custom compile-time sorting function (see src/sphinxcustomsort.inl)

  • added result set sorting by @random value

  • added result set merging for indexes with different schemas

  • added query comments support (3rd arg to Query()/AddQuery() API calls, copied verbatim to query log)

  • added keyword extraction support, BuildKeywords() API call

  • added binding field weights by name, SetFieldWeights() API call

  • added optional limit on query time, SetMaxQueryTime() API call

  • added optional limit on found matches count (4rd arg to SetLimits() API call, so-called 'cutoff')

APIs and SphinxSE

  • added pure C API (libsphinxclient)

  • added Ruby API (thanks to Dmytro Shteflyuk)

  • added Java API

  • added SphinxSE support for MVAs (use varchar), floats (use float), 64bit docids (use bigint)

  • added SphinxSE options "floatrange", "geoanchor", "fieldweights", "indexweights", "maxquerytime", "comment", "host" and "port"; and support for "expr:CLAUSE"

  • improved SphinxSE max query size (using MySQL condition pushdown), upto 256K now


  • added scripting (shebang syntax) support to config files (example: #!/usr/bin/php in the first line)

  • added unified config handling and validation to all programs

  • added unified documentation

  • added .spec file for RPM builds

  • added automated testing suite

  • improved index locking, now fcntl()-based instead of buggy file-existence-based

  • fixed unaligned RAM accesses, now works on SPARC and ARM

Changes and fixes since 0.9.8-rc2

  • added pure C API (libsphinxclient)

  • added Ruby API

  • added SetConnectTimeout() PHP API call

  • added allowed type check to UpdateAttributes() handler (bug #174)

  • added defensive MVA checks on index preload (protection against broken indexes, bug #168)

  • added sphinx-min.conf sample file

  • added --without-iconv switch to configure

  • removed redundant -lz dependency in searchd

  • removed erroneous "xmlpipe2 deprecated" warning

  • fixed EINTR handling in piped read (bug #166)

  • fixup query time before logging and sending to client (bug #153)

  • fixed attribute updates vs full-scan early-reject index (bug #149)

  • fixed gcc warnings (bug #160)

  • fixed mysql connection attempt vs pgsql source type (bug #165)

  • fixed 32-bit wraparound when preloading over 2 GB files

  • fixed "out of memory" message vs over 2 GB allocs (bug #116)

  • fixed unaligned RAM access detection on ARM (where unaligned reads do not crash but produce wrong results)

  • fixed missing full scan results in some cases

  • fixed several bugs in --merge, --merge-dst-range

  • fixed @geodist vs MultiQuery and filters, @expr vs MultiQuery

  • fixed GetTokenEnd() vs 1-grams (was causing crash in excerpts)

  • fixed sql_query_range to handle empty strings in addition to NULL strings (Postgres specific)

  • fixed morphology=none vs infixes

  • fixed case sensitive attributes names in UpdateAttributes()

  • fixed ext2 ranking vs. stopwords (now using atompos from query parser)

  • fixed EscapeString() call

  • fixed escaped specials (now handled as whitespace if not in charset)

  • fixed schema minimizer (now handles type/size mismatches)

  • fixed word stats in extended2; stemmed form is now returned

  • fixed spelldump case folding vs dictionary-defined character sets

  • fixed Postgres BOOLEAN handling

  • fixed enforced "inline" docinfo on empty indexes (normally ok, but index merge was really confused)

  • fixed rare count(distinct) out-of-bounds issue (it occasionaly caused too high @distinct values)

  • fixed hangups on documents with id=DOCID_MAX in some cases

  • fixed rare crash in tokenizer (prefixed synonym vs. input stream eof)

  • fixed query parser vs "aaa (bbb ccc)|ddd" queries

  • fixed BuildExcerpts() request in Java API

  • fixed Postgres specific memory leak

  • fixed handling of overshort keywords (less than min_word_len)

  • fixed HTML stripper (now emits space after indexed attributes)

  • fixed 32-field case in query parser

  • fixed rare count(distinct) vs. querying multiple local indexes vs. reusable sorter issue

  • fixed sorting of negative floats in SPH_SORT_EXTENDED mode

A.37. Version 0.9.7, 02 apr 2007

  • added support for sql_str2ordinal_column

  • added support for upto 5 sort-by attrs (in extended sorting mode)

  • added support for separate groups sorting clause (in group-by mode)

  • added support for on-the-fly attribute updates (PRE-ALPHA; will change heavily; use for preliminary testing ONLY)

  • added support for zero/NULL attributes

  • added support for 0.9.7 features to SphinxSE

  • added support for n-grams (alpha, 1-grams only for now)

  • added support for warnings reported to client

  • added support for exclude-filters

  • added support for prefix and infix indexing (see max_prefix_len, max_infix_len)

  • added @* syntax to reset current field to query language

  • added removal of duplicate entries in query index order

  • added PHP API workarounds for PHP signed/unsigned braindamage

  • added locks to avoid two concurrent indexers working on same index

  • added check for existing attributes vs. docinfo=none case

  • improved groupby code a lot (better precision, and upto 25x times faster in extreme cases)

  • improved error handling and reporting

  • improved handling of broken indexes (reports error instead of hanging/crashing)

  • improved mmap() limits for attributes and wordlists (now able to map over 4 GB on x64 and over 2 GB on x32 where possible)

  • improved malloc() pressure in head daemon (search time should not degrade with time any more)

  • improved test.php command line options

  • improved error reporting (distributed query, broken index etc issues now reported to client)

  • changed default network packet size to be 8M, added extra checks

  • fixed division by zero in BM25 on 1-document collections (in extended matching mode)

  • fixed .spl files getting unlinked

  • fixed crash in schema compatibility test

  • fixed UTF-8 Russian stemmer

  • fixed requested matches count when querying distributed agents

  • fixed signed vs. unsigned issues everywhere (ranged queries, CLI search output, and obtaining docid)

  • fixed potential crashes vs. negative query offsets

  • fixed 0-match docs vs. extended mode vs. stats

  • fixed group/timestamp filters being ignored if querying from older clients

  • fixed docs to mention pgsql source type

  • fixed issues with explicit '&' in extended matching mode

  • fixed wrong assertion in SBCS encoder

  • fixed crashes with no-attribute indexes after rotate

A.38. Version 0.9.7-rc2, 15 dec 2006

  • added support for extended matching mode (query language)

  • added support for extended sorting mode (sorting clauses)

  • added support for SBCS excerpts

  • added mmap()ing for attributes and wordlist (improves search time, speeds up fork() greatly)

  • fixed attribute name handling to be case insensitive

  • fixed default compiler options to simplify post-mortem debugging (added -g, removed -fomit-frame-pointer)

  • fixed rare memory leak

  • fixed "hello hello" queries in "match phrase" mode

  • fixed issue with excerpts, texts and overlong queries

  • fixed logging multiple index name (no longer tokenized)

  • fixed trailing stopword not flushed from tokenizer

  • fixed boolean evaluation

  • fixed pidfile being wrongly unlink()ed on bind() failure

  • fixed --with-mysql-includes/libs (they conflicted with well-known paths)

  • fixes for 64-bit platforms

A.39. Version 0.9.7-rc1, 26 oct 2006

  • added alpha index merging code

  • added an option to decrease max_matches per-query

  • added an option to specify IP address for searchd to listen on

  • added support for unlimited amount of configured sources and indexes

  • added support for group-by queries

  • added support for /2 range modifier in charset_table

  • added support for arbitrary amount of document attributes

  • added logging filter count and index name

  • added --with-debug option to configure to compile in debug mode

  • added -DNDEBUG when compiling in default mode

  • improved search time (added doclist size hints, in-memory wordlist cache, and used VLB coding everywhere)

  • improved (refactored) SQL driver code (adding new drivers should be very easy now)

  • improved exceprts generation

  • fixed issue with empty sources and ranged queries

  • fixed querying purely remote distributed indexes

  • fixed suffix length check in English stemmer in some cases

  • fixed UTF-8 decoder for codes over U+20000 (for CJK)

  • fixed UTF-8 encoder for 3-byte sequences (for CJK)

  • fixed overshort (less than min_word_len) words prepended to next field

  • fixed source connection order (indexer does not connect to all sources at once now)

  • fixed line numbering in config parser

  • fixed some issues with index rotation

A.40. Version 0.9.6, 24 jul 2006

  • added support for empty indexes

  • added support for multiple sql_query_pre/post/post_index

  • fixed timestamp ranges filter in "match any" mode

  • fixed configure issues with --without-mysql and --with-pgsql options

  • fixed building on Solaris 9

A.41. Version 0.9.6-rc1, 26 jun 2006

  • added boolean queries support (experimental, beta version)

  • added simple file-based query cache (experimental, beta version)

  • added storage engine for MySQL 5.0 and 5.1 (experimental, beta version)

  • added GNU style configure script

  • added new searchd protocol (all binary, and should be backwards compatible)

  • added distributed searching support to searchd

  • added PostgreSQL driver

  • added excerpts generation

  • added min_word_len option to index

  • added max_matches option to searchd, removed hardcoded MAX_MATCHES limit

  • added initial documentation, and a working example.sql

  • added support for multiple sources per index

  • added soundex support

  • added group ID ranges support

  • added --stdin command-line option to search utility

  • added --noprogress option to indexer

  • added --index option to search

  • fixed UTF-8 decoder (3-byte codepoints did not work)

  • fixed PHP API to handle big result sets faster

  • fixed config parser to handle empty values properly

  • fixed redundant time(NULL) calls in time-segments mode