All notable changes to this project (at least, from v3.0.0 onwards) are documented in this file.
- Support for Manticore 6.0 (#1242)
sphinx-prefixed search methods, in case the standardsearchis overridden from something unrelated. (#1265)none/search_nonescopes that can be chained to searches and will return no results.- Added
ThinkingSphinx::Processor#syncto synchronise updates/deletions based on a real-time index's scope, by @akostadinov in @1258.
- Improved Rails 7.1 support, by @jdelstrother in #1252.
- Handle both SQL and RT indices correctly for inheritance column checks, by @akostadinov in #1249.
- Ensure tests and CI work with recent Manticore versions, by @jdelstrother in #1263.
- Use
rm -rfto delete test and temporary directories (instead ofrm -r).
- Fixed total count of results in pagination information for Manticore 5.0+, by disabling the cutoff limit. (#1239).
- ThinkingSphinx::Processor, a public interface to perform index-related operations on model instances or model name/id combinations. In collaboration with @akostadinov (#1215).
- Confirmed support by testing against Ruby 3.1 and 3.2 by @jdelStrother (#1237).
- Fix YAML loading, by @aepyornis (#1217).
- Further fixes for File.exist? instead of the deprecated File.exists?, by @funsim (#1221) and @graaf (1233).
- Treat unknown column errors as QueryErrors, so retrying the query occurs automatically.
- Fix MariaDB error handling.
- Rails 7 support, including contributions from @anthonyshull in #1205.
- Confirmed support by testing against Manticore 4.0 and Sphinx 3.4.
- Include instance_exec in ThinkingSphinx::Search::CORE_METHODS by @jdelStrother in #1210.
- Use File.exist? instead of the deprecated File.exists? (#1211).
- StaleIdsExceptions now include a URL in their error message with recommendations on how to resolve the problem.
- Fire real-time callbacks on
after_commit(including deletions) to ensure data is fully persisted to the database before updating Sphinx. More details in #1204.
- Ensure Thinking Sphinx's ActiveRecord components are loaded by either Rails' after_initialise hook or ActiveSupport's on_load notification, because the order of these two events are not consistent.
- Remove
app/indicesfrom eager_load_paths in Rails 4.2 and 5, to match the behaviour in 6.
- Ensure ActiveRecord components are loaded for rake tasks, but only after the Rails application has initialised. More details in #1199.
- Confirmed support for Ruby 3.0.
- Orphaned records in real-time indices can now be cleaned up without running
rails ts:rebuild. Disabled by default, can be enabled by settingreal_time_tidyto true per environment inconfig/thinking_sphinx.yml(and will needts:rebuildto restructure indices upon initial deploy). More details in #1192.
- Avoid loading ActiveRecord during Rails initialisation so app configuration can still have an impact (@jdelStrother in #1194).
- Remove
app/indices(in both the Rails app and engines) from Rails' eager load paths, which was otherwise leading to indices being loaded more than once. (See #1191 and #1195).
- Support for Sphinx v3.3 and Manticore v3.5.
- Support for Rails 6.1 (via joiner v0.6.0).
enable_staris no longer available as a configuration option, as it's been enabled by default in Sphinx since v2.2.2, and is no longer allowed in Sphinx v3.3.1.- All timestamp attributes are now considered plain integer values from Sphinx's perspective. Sphinx was already expecting integers, but since Sphinx v3.3.1 it doesn't recognise timestamps as a data type. There is no functional difference with this change - Thinking Sphinx was always converting times to their UNIX epoch integer values.
- Allow configuration of the maximum statement length (@kalsan in #1179).
- Respect
:pathvalues to navigate associations for Thinking Sphinx callbacks on SQL-backed indices. Discussed in #1182.
- Don't attempt to update delta flags on frozen model instances.
- New interface for adding callbacks to indexed models (which is no longer done automatically). Discussed in #1173 and committed via #1175. This is a breaking change - you will need to add these callbacks. See the full release notes for examples.
- Fields and attributes can be overriden - whichever's defined last with a given name is the definition that's used. This is an edge case, but useful if you want to override any of the default fields/indices. (Requested by @kalsan in #1172.)
- Custom index_set_class implementations can now expect the
:instancesoption to be set alongside:classes, which is useful in cases to limit the indices returned if you're splitting index data for given classes/models into shards. (Introduced in PR #1171 after discussions with @lunaru in #1166.)
- Sphinx 2.2.11 or newer is required, or Manticore 2.8.2 or newer.
- Ruby 2.4 or newer is required.
- Rails 4.2 or newer is required.
- Remove internal uses of
send, replaced withpublic_sendas that's available in all supported Ruby versions. - Deletion statements are simplified by avoiding the need to calculate document keys/offsets (@njakobsen via #1134).
- Real-time data is deleted before replacing it, to avoid duplicate data when offsets change (@njakobsen via #1134).
- Use
reference_nameas per customindex_set_classdefinitions. Previously, the class method was called onThinkingSphinx::IndexSeteven if a custom subclass was configured. (As per discussions with @kalsan in #1172.)
- Automatically remove
app/indicesfrom Zeitwerk's autoload paths in Rails 6.0 onwards (if using Zeitwerk as the autoloader).
- Confirmed Rails 6.0 support.
- Added ability to have custom real-time index processors (which handles all indices) and populators (which handles a particular index). These are available to get/set via
ThinkingSphinx::RealTime.processorandThinkingSphinx::RealTime.populator(and discussed in more detail in the release notes).
- Improve failure message when tables don't exist for models associated with Sphinx indices (Kiril Mitov in #1139).
- Injected has-many/habtm collection search calls as default extensions to associations in Rails 5+, as it's a more reliable approach in Rails 6.0.0.
- Reverted loading change behaviour from v4.3.1 for Rails v5 (Eduardo J. in #1138).
- Fixed loading of index files to work with Rails 6 and Zeitwerk (#1137).
- Allow overriding of Sphinx's running state, which is useful when Sphinx commands are interacting with a remote Sphinx daemon. As per discussions in #1131.
- Allow skipping of directory creation, as per discussions in #1131.
- Use ActiveSupport's lock monitor where possible (Rails 5.1.5 onwards) to avoid database deadlocks. Essential investigation by Jonathan del Strother (#1132).
- Allow facet searching on distributed indices (#1135).
- Allow changing the default encoding for MySQL database connections from utf8 to something else via the
mysql_encodingsetting inconfig/thinking_sphinx.yml. In the next significant release, the default will change to utf8mb4 (which is supported in MySQL 5.5.3 and newer). - Added Rails 6.0 and Manticore 2.8 to the test matrix.
- Use Arel's SQL literals for generated order clauses, to avoid warnings from Rails 6.
- Fix usage of alternative primary keys in update and deletion callbacks and attribute access.
- Ensure
respond_to?takes Sphinx scopes into account (Jonathan del Strother in #1124). - Add
:excerptsas a known option for search requests. - Fix depolymorphed association join construction with Rails 6.0.0.beta2.
- Reset ThinkingSphinx::Configuration's cached values when Rails reloads, to avoid holding onto stale references to ActiveRecord models (#1125).
- Don't join against associations in
sql_queryif they're only used by query-sourced properties (Hans de Graaff in #1127).
- The
:sqlsearch option can now accept per-model settings with model names as keys. e.g.ThinkingSphinx.search "foo", :sql => {'Article' => {:include => :user}}(Sergey Malykh in #1120).
- Drop MRI 2.2 from the test matrix, and thus no longer officially supported (though the code will likely continue to work with 2.2 for a while).
- Added MRI 2.6, Sphinx 3.1 and Manticore 2.7 to the test matrix.
- Real-time indices now work with non-default integer primary keys (alongside UUIDs or other non-integer primary keys).
- Support Sphinx 3.0.
- Allow disabling of docinfo setting via
skip_docinfo: trueinconfig/thinking_sphinx.yml. - Support merging of delta indices into their core counterparts using ts:merge.
- Support UNIX sockets as an alternative for TCP connections to the daemon (MRI-only).
- Translate relative paths to absolute when generating configuration when
absolute_paths: trueis set per environment inconfig/thinking_sphinx.yml.
- Drop Sphinx 2.0 support.
- Drop auto-typing of filter values.
- INDEX_FILTER environment variable is applied when running ts:index on SQL-backed indices.
- Drop MRI 2.0/2.1 support.
- Display a useful error message if processing real-time indices but the daemon isn't running.
- Refactor interface code into separate command classes, and allow for a custom rake interface.
- Add frozen_string_literal pragma comments.
- Log exceptions when processing real-time indices, but don't stop.
- Update polymorphic properties to support Rails 5.2.
- Allow configuration of the index guard approach.
- Output a warning if guard files exist when calling ts:index.
- Delete index guard files as part of ts:rebuild and ts:clear.
- Handle situations where no exit code is provided for Sphinx binary calls.
- Don't attempt to interpret indices for models that don't have a database table.
- Allow use of deletion callbacks for rollback events.
- Remove extra deletion code in the Populator - it's also being done by the real-time rake interface.
- Real-time callback syntax for namespaced models accepts a string (as documented).
- Fix up logged warnings.
- Add missing search options to known values to avoid incorrect warnings.
- Treat "Lost connection to MySQL server" as a connection error (Manuel Schnitzer).
- Index normalisation will now work even when index model tables don't exist.
- Rake tasks are now unified, so the original tasks will operate on real-time indices as well.
- Output warnings when unknown options are used in search calls.
- Allow generation of a single real-time index (Tim Brown).
- Automatically use UTF8 in Sphinx for encodings that are extensions of UTF8.
- Basic type checking for attribute filters.
- Delta callback logic now prioritises checking for high level settings rather than model changes.
- Allow for unsaved records when calculating document ids (and return nil).
- Display SphinxQL deletion statements in the log.
- Add support for Ruby's frozen string literals feature.
- Use saved_changes if it's available (in Rails 5.1+).
- Set a default connection timeout of 5 seconds.
- Don't search multi-table inheritance ancestors.
- Handle non-computable queries as parse errors.
- Index normalisation now occurs consistently, and removes unneccesary sphinx_internal_class_name fields from real-time indices.
- Fix Sphinx connections in JRuby.
- Fix long SphinxQL query handling in JRuby.
- Always close the SphinxQL connection if Innertube's asking (@cmaion).
- Get bigint primary keys working in Rails 5.1.
- Fix handling of attached starts of Sphinx (via Henne Vogelsang).
- Fix multi-field conditions.
- Use the base class of STI models for polymorphic join generation (via Andrés Cirugeda).
- Ensure ts:index now respects rake silent/quiet flags.
- Real-time callbacks can now be used with after_commit hooks if that's preferred over after_save.
- Allow for custom batch sizes when populating real-time indices.
- Only toggle the delta value if the record has changed or is new (rather than on every single save call).
- Delta indexing is now quiet by default (rather than verbose).
- Use Riddle's reworked command interface for interacting with Sphinx's command-line tools.
- Respect Rake's quiet and silent flags for the Thinking Sphinx rake tasks.
- ts:start and ts:stop tasks default to verbose.
- Sort engine paths for loading indices to ensure they're consistent.
- Custom exception class for invalid database adapters.
- Memoize the default primary keys per context.
- Explicit source method in the SQLQuery Builder instead of relying on method missing, thus avoiding any global methods named 'source' (Asaf Bartov).
- Load indices before deleting index files, to ensure the files are actually found and deleted.
- Avoid loading ActiveRecord earlier than necessary. This avoids loading Rails out of order, which caused problems with Rails 5.
- Handle queries that are too long for Sphinx.
- Improve Rails 5 / JRuby support.
- Fixed handling of multiple field tokens in wildcarding logic.
- Ensure custom primary key columns are handled consistently (Julio Monteiro).
- Add JSON attribute support for real-time indices.
- Add ability to disable all Sphinx-related callbacks via ThinkingSphinx::Callbacks.suspend! and ThinkingSphinx::Callbacks.resume!. Particularly useful for unit tests.
- Add native OutOfBoundsError for search queries outside the pagination bounds.
- Support MySQL SSL options on a per-index level (@arrtchiu).
- Allow for different indexing strategies (e.g. all at once, or one by one).
- Allow rand_seed as a select option (Mattia Gheda).
- Add primary_key option for index definitions (Nathaneal Gray).
- Add ability to start searchd in the foreground (Andrey Novikov).
- Improved error messages for duplicate property names and missing columns.
- Don't populate search results when requesting just the count values (Andrew Roth).
- Reset delta column before core indexing begins (reverting behaviour introduced in 3.1.0). See issue #958 for further discussion.
- Use Sphinx's bulk insert ability (Chance Downs).
- Reduce memory/object usage for model references (Jonathan del Strother).
- Disable deletion callbacks when real-time indices are in place and all other real-time callbacks are disabled.
- Only use ERB to parse the YAML file if ERB is loaded.
- Ensure SQL table aliases are reliable for SQL-backed index queries.
- Fixed mysql2 compatibility for memory references (Roman Usherenko).
- Fixed JRuby compatibility with camelCase method names (Brandon Dewitt).
- Fix stale id handling for multiple search contexts (Jonathan del Strother).
- Handle quoting of namespaced tables (Roman Usherenko).
- Make preload_indices thread-safe.
- Improved handling of marshalled/demarshalled search results.
- Add JSON as a Sphinx type for attributes (Daniel Vandersluis).
- minimal_group_by? can now be set in config/thinking_sphinx.yml to automatically apply to all index definitions.
- Add a contributor code of conduct.
- Remove polymorphic association and HABTM query support (when related to Thinking Sphinx) when ActiveRecord 3.2 is involved.
- Remove default charset_type - no longer required for Sphinx 2.2.
- Removing sql_query_info setting, as it's no longer used by Sphinx (nor is it actually used by Thinking Sphinx).
- Kaminari expects prev_page to be available.
- Don't try to delete guard files if they don't exist (@exAspArk).
- Handle database settings reliably, now that ActiveRecord 4.2 uses strings all the time.
- More consistent with escaping table names.
- Bug fix for association creation (with polymophic fields/attributes).
- Allow for custom offset references with the :offset_as option - thus one model across many schemas with Apartment can be treated differently.
- Allow for custom IndexSet classes.
- Log excerpt SphinxQL queries just like the search queries.
- Load Railtie if Rails::Railtie is defined, instead of just Rails (Andrew Cone).
- Convert raw Sphinx results to an array when querying (Bryan Ricker).
- Add bigint support for real-time indices, and use bigints for the sphinx_internal_id attribute (mapped to model primary keys) (Chance Downs).
- Generate de-polymorphised associations properly for Rails 4.2
- Use reflect_on_association instead of reflections, to stick to the public ActiveRecord::Base API.
- Don't load ActiveRecord early - fixes a warning in Rails 4.2.
- Don't double-up on STI filtering, already handled by Rails.
- Allow for custom paths for index files using :path option in the ThinkingSphinx::Index.define call.
- Allow the binlog path to be an empty string (Bobby Uhlenbrock).
- Add status task to report on whether Sphinx is running.
- Real-time index callbacks can take a block for dynamic scoping.
- Allow casting of document ids pre-offset as bigints (via big_documents_id option).
- regenerate task now only deletes index files for real-time indices.
- Raise an exception when a populated search query is modified (as it can't be requeried).
- Log indices that aren't processed due to guard files existing.
- Paginate records by 1000 results at a time when flagging as deleted.
- Default the Capistrano TS Rails environment to use rails_env, and then fall back to stage.
- rebuild task uses clear between stopping the daemon and indexing.
- Ensure indexing guard files are removed when an exception is raised (Bobby Uhlenbrock).
- Don't update real-time indices for objects that are not persisted (Chance Downs).
- Use STI base class for polymorphic association replacements.
- Convert database setting keys to symbols for consistency with Rails (@dimko).
- Field weights and other search options are now respected from set_property.
- Models with more than one index have correct facet counts (using Sphinx 2.1.x or newer).
- Some association fixes for Rails 4.1.
- Clear connections when raising connection errors.
- Allow for common section in generated Sphinx configuration files for Sphinx 2.2.x (disabled by default, though) (Trevor Smith).
- Basic support for HABTM associations and MVAs with query/ranged-query sources.
- Real-time indices callbacks can be disabled (useful for unit tests).
- ThinkingSphinx::Test has a clear method and no-index option for starting for real-time setups.
- Allow disabling of distributed indices.
- Include full statements when query execution errors are raised (uglier, but more useful when debugging).
- Connection error messages now mention Sphinx, instead of just MySQL.
- Raise an exception when a referenced column does not exist.
- Capistrano tasks use thinking_sphinx_rails_env (defaults to standard environment) (Robert Coleman).
- Alias group and count columns for easier referencing in other clauses.
- Log real-time index updates (Demian Ferreiro).
- All indices now respond to a public attributes method.
- Don't apply attribute-only updates to real-time indices.
- Don't instantiate blank strings (via inheritance type columns) as constants.
- Don't presume all indices for a model have delta pairs, even if one does.
- Always use connection options for connection information.
- respond_to? works reliably with masks (Konstantin Burnaev).
- Avoid null values in MVA query/ranged-query sources.
- Don't send unicode null characters to real-time Sphinx indices.
- :populate option is now respected for single-model searches.
- :thinking_sphinx_roles is now used consistently in Capistrano v3 tasks.
- Only expand log directory if it exists.
- Handle JDBC connection errors appropriately (Adam Hutchison).
- Fixing wildcarding of Unicode strings.
- Improved handling of association searches with real-time indices, including via has_many :though associations (Rob Anderton).
- Support for Capistrano v3 (Alexander Tipugin).
- JRuby support (with Sphinx 2.1 or newer).
- Support for Sphinx 2.2.x's HAVING and GROUP N BY SphinxQL options.
- Adding max_predicted_time search option (Sphinx 2.2.x).
- Wildcard/starring can be applied directly to strings using ThinkingSphinx::Query.wildcard('pancakes'), and escaping via ThinkingSphinx::Query.escape('pancakes').
- Capistrano recipe now includes tasks for realtime indices.
- :group option within :sql options in a search call is passed through to the underlying ActiveRecord relation (Siarhei Hanchuk).
- Persistent connections can be disabled if you wish.
- Track what's being indexed, and don't double-up while indexing is running. Single indices (e.g. deltas) can be processed while a full index is happening, though.
- Pass through :delta_options to delta processors (Timo Virkalla).
- All delta records can have their core pairs marked as deleted after a suspended delta (use ThinkingSphinx::Deltas.suspend_and_update instead of ThinkingSphinx::Deltas.suspend).
- Set custom database settings within the index definition, using the set_database method. A more sane approach with multiple databases.
- Updating Riddle requirement to >= 1.5.10.
- Extracting join generation into its own gem: Joiner.
- Geodist calculation is now prepended to the SELECT statement, so it can be referred to by other dynamic attributes.
- Auto-wildcard/starring (via :star => true) now treats escaped characters as word separators.
- Capistrano recipe no longer automatically adds thinking_sphinx:index and thinking_sphinx:start to be run after deploy:cold.
- UTF-8 forced encoding is now disabled by default (in line with Sphinx 2.1.x).
- Sphinx functions are now the default, instead of the legacy special variables (in line with Sphinx 2.1.x).
- Rails 3.1 is no longer supported.
- MRI 1.9.2 is no longer supported.
- Insist on at least * for SphinxQL SELECT statements.
- Reset the delta column to true after core indexing is completed, instead of before, and don't filter out delta records from the core source.
- Provide a distributed index per model that covers both core and delta indices.
- Indices will be detected in Rails engines upon configuration.
- Destroy callbacks are ignored for non-persisted objects.
- Blank STI values are converted to the parent class in Sphinx index data (Jonathan Greenberg).
- Track indices on parent STI models when marking documents as deleted.
- Separate per_page/max_matches values are respected in facet searches (Timo Virkkala).
- Don't split function calls when casting timestamps (Timo Virkalla).
- Raise an error if no indices match the search criteria (Bryan Ricker).
- skip_time_zone setting is now available per environment via config/thinking_sphinx.yml to avoid the sql_query_pre time zone command.
- Added new search options in Sphinx 2.1.x.
- Added ability to disable UTF-8 forced encoding, now that Sphinx 2.1.2 returns UTF-8 strings by default. This will be disabled by default in Thinking Sphinx 3.1.0.
- Added ability to switch between Sphinx special variables and the equivalent functions. Sphinx 2.1.x requires the latter, and that behaviour will become the default in Sphinx 3.1.0.
- Adding search_for_ids on scoped search calls.
- MySQL users can enable a minimal GROUP BY statement, to speed up queries: set_property :minimal_group_by? => true.
- Updating Riddle dependency to be >= 1.5.9.
- Separated directory preparation from data generation for real-time index (re)generation tasks.
- Have tests index UTF-8 characters where appropriate (Pedro Cunha).
- Always use DISTINCT in group concatenation.
- Sphinx connection failures now have their own class, ThinkingSphinx::ConnectionError, instead of the standard Mysql2::Error.
- Don't clobber custom :select options for facet searches (Timo Virkkala).
- Automatically load Riddle's Sphinx 2.0.5 compatability changes.
- Realtime fields and attributes now accept symbols as well as column objects, and fields can be sortable (with a _sort prefix for the matching attribute).
- Insist on the log directory existing, to ensure correct behaviour for symlinked paths. (Michael Pearson).
- Rake's silent mode is respected for indexing (@endoscient).
- Cast every column to a timestamp for timestamp attributes with multiple columns.
- Don't use Sphinx ordering if SQL order option is supplied to a search.
- Custom middleware and mask options now function correctly with model-scoped searches.
- Suspended deltas now no longer update core indices as well.
- Use alphabetical ordering for index paths consistently (@grin).
- Convert very small floats to fixed format for geo-searches.
- Allow scoping of real-time index models.
- Updating Riddle dependency to be >= 1.5.8.
- Real-time index population presentation and logic are now separated.
- Using the connection pool for update callbacks, excerpts, deletions.
- Don't add the sphinx_internal_class_name unless STI models are indexed.
- Use Mysql2's reconnect option and have it turned on by default.
- Improved auto-starring with escaped characters.
- Respect existing sql_query_range/sql_query_info settings.
- Don't add select clauses or joins to sql_query if they're for query/ranged-query properties.
- Set database timezones as part of the indexing process.
- Chaining scopes with just options works again.
- ts:regenerate rake task for rebuilding Sphinx when realtime indices are involved.
- ts:clear task removes all Sphinx index and binlog files.
- Facet search calls now respect the limit option (which otherwise defaults to max_matches) (Demian Ferreiro).
- Excerpts words can be overwritten with the words option (@groe).
- The :facets option can be used in facet searches to limit which facets are queried.
- A separate role can be set for Sphinx actions with Capistrano (Andrey Chernih).
- Facet searches can now be called from Sphinx scopes.
- Updating Riddle dependency to be >= 1.5.7.
- Glaze now responds to respond_to? (@groe).
- Deleted ActiveRecord objects are deleted in realtime indices as well.
- Realtime callbacks are no longer automatically added, but they're now more flexible (for association situations).
- Cleaning and refactoring so Code Climate ranks this as A-level code (Philip Arndt, Shevaun Coker, Garrett Heinlen).
- Exceptions raised when communicating with Sphinx are now mentioned in the logs when queries are retried (instead of STDOUT).
- Excerpts now use just the query and standard conditions, instead of parsing Sphinx's keyword metadata (which had model names in it).
- Get database connection details from ActiveRecord::Base, not each model, as this is where changes are reflected.
- Default Sphinx scopes are applied to new facet searches.
- Empty queries with the star option set to true are handled gracefully.
- Excerpts are now wildcard-friendly.
- Facet searches now use max_matches value (with a default of 1000) to ensure as many results as possible are returned.
- The settings cache is now cleared when the configuration singleton is reset (Pedro Cunha).
- Escaped @'s in queries are considered part of each word, instead of word separators.
- Internal class name conditions are ignored with auto-starred queries.
- RDoc doesn't like constant hierarchies split over multiple lines.
- INDEX_ONLY environment flag is passed through when invoked through Capistrano (Demian Ferreiro).
- use_64_bit option returns as cast_to_timestamp instead (Denis Abushaev).
- Collection of hooks (lambdas) that get called before indexing. Useful for delta libraries.
- Updating Riddle dependency to be >= 1.5.6
- Delta jobs get common classes to allow third-party delta behaviours to leverage Thinking Sphinx.
- Raise ThinkingSphinx::MixedScopesError if a search is called through an ActiveRecord scope.
- GroupEnumeratorsMask is now a default mask, as masks need to be in place before search results are populated/the middleware is called (and previously it was being added within a middleware call).
- The current_page method is now a part of ThinkingSphinx::Search, as it is used when populating results.
- Update to association handling for Rails/ActiveRecord 4.0.0.rc1.
- Cast and concatenate multi-column attributes correctly.
- Don't load fields or attributes when building a real-time index - otherwise the index is translated before it has a chance to be built.
- Default search panes are cloned for each search.
- Index-level settings (via set_property) are now applied consistently after global settings (in thinking_sphinx.yml).
- All string values returned from Sphinx are now properly converted to UTF8.
- The default search masks are now cloned for each search, instead of referring to the constant (and potentially modifying it often).
- Ruby 2.0 support.
- Rails 4.0.0 beta1 support.
- Indexes defined in app/indices in engines are now loaded (Antonio Tapiador del Dujo).
- Query errors are classified as such, instead of getting the base SphinxError.
- per_page now accepts an optional paging limit, to match WillPaginate's behaviour. If none is supplied, it just returns the page size.
- Strings and regular expressions in ThinkingSphinx::Search::Query are now treated as UTF-8.
- Setting a custom framework will rebuild the core configuration around its provided settings (path and environment).
- Search masks don't rely on respond_to?, and so Object/Kernel methods are passed through to the underlying array instead.
- Empty search conditions are now ignored, instead of being appended with no value (Nicholas Klick).
- Custom conditions are no longer added to the sql_query_range value, as they may involve associations.
- :utf8? option within index definitions is now supported, and defaults to true if the database configuration's encoding is set to 'utf8'.
- indices_location and configuration_file values in thinking_sphinx.yml will be applied to the configuration.
- Primary keys that are not 'id' now work correctly.
- Search options specified in index definitions and thinking_sphinx.yml are now used in search requests (eg: max_matches, field_weights).
- Custom association conditions are no longer presumed to be an array.
- Capistrano tasks use the correct ts rake task prefix (David Celis).
- Provide Capistrano deployment tasks (David Celis).
- Allow specifying of Sphinx version. Is only useful for Flying Sphinx purposes at this point - has no impact on Riddle or Sphinx.
- Support new JDBC configuration style (when JDBC can be used) (Kyle Stevens).
- Mysql2::Errors are wrapped as ThinkingSphinx::SphinxErrors, with subclasses of SyntaxError and ParseError used appropriately. Syntax and parse errors do not prompt a retry on a new connection.
- Polymorphic associations can be used within index definitions when the appropriate classes are set out.
- Allow custom strings for SQL joins in index definitions.
- indexer and searchd settings are added to the appropriate objects from config/thinking_sphinx.yml (@ygelfand).
- Use connection pool for search queries. If a query fails, it will be retried on a new connection before raising if necessary.
- Glaze always passes methods through to the underlying ActiveRecord::Base object if they don't exist on any of the panes.
- Referring to associations via polymorphic associations in an index definition now works.
- Don't override foreign keys for polymorphic association replacements.
- Quote namespaced model names in class field condition.
- New lines are maintained and escaped in custom source queries.
- Subclasses of indexed models fire delta callbacks properly.
- Thinking Sphinx can be loaded via thinking/sphinx, to satisfy Bundler.
- New lines are maintained and escaped in sql_query values.
- Initial realtime index support, including the ts:generate task for building index datasets. Sphinx 2.0.6 is required.
- SphinxQL connection pooling via the Innertube gem.
- Updating Riddle dependency to 1.5.4.
- UTF-8 is now the default charset again (as it was in earlier Thinking Sphinx versions).
- Removing ts:version rake task.
- Respect source options as well as underlying settings via the set_property method in index definitions.
- Load real-time index definitions when listing fields, attributes, and/or conditions.
- Source type support (query and ranged query) for both attributes and fields. Custom SQL strings can be supplied as well.
- Wordcount attributes and fields now supported.
- Support for Sinatra and other non-Rails frameworks.
- A sphinx scope can be defined as the default.
- An index can have multiple sources, by using define_source within the index definition.
- sanitize_sql is available within an index definition.
- Providing :prefixes => true or :infixes => true as an option when declaring a field means just the noted fields have infixes/prefixes applied.
- ThinkingSphinx::Search#query_time returns the time Sphinx took to make the query.
- Namespaced model support.
- Default settings for index definition arguments can be set in config/thinking_sphinx.yml.
- A custom Riddle/Sphinx controller can be supplied. Useful for Flying Sphinx to have an API layer over Sphinx commands, without needing custom gems for different Thinking Sphinx/Flying Sphinx combinations.
- Correctly escape nulls in inheritance column (Darcy Laycock).
- Use ThinkingSphinx::Configuration#render_to_file instead of ThinkingSphinx::Configuration#build in test helpers (Darcy Laycock).
- Suppressing delta output in test helpers now works (Darcy Laycock).
First pre-release of v3. Not quite feature complete, but the important stuff is certainly covered. See the README for more the finer details.