Scout

Fixed

• This release takes the site's ID into account when updating and removing items from an index (#288)

• This release includes a bunch of performance improvements from PR's by @EpocSquadron (#171) and @aaronbushnell (#248)
• Better checks to prevent crashes when "searchable" isn't defined (#262)
• "lazy load" the index criteria (#269)

Added

• `scout/index/import now optionally takes a --queue=1` parameter to run the import(s) through the queue instead running them straigt away.

Added

• Added a config setting to keep using the orginal object in case `splittedObjects` only contains 1 item. (#193 & #219, thanks @gregkohn)

Fixed

• MakeSearchable job now uses the index criteria (#177)
• Disable relations on delete if so configured, should improve performance on larger installs (#227)

Added

• Added support for PHP 8.x
• Added support for algolia/algoliasearch-client-php 3.x

Added

• Added `indexRelations` config setting (true by default) (#175 & #205)
• Exposed options to update and dump index settings from the CP utility (#209)

Added

• Added ShouldBeSearchableEvent event, docs here (#205)

Fixed

• CP Utility now works for indexes using the wildcard selector for siteId (#163)
• CP Utility labels are now translatable
• Fixed compatibility with Craft Commerce (#178)

• Don't include indices in Settings::toArray() by default

• Remove pro requirement for Utility

• Allow tightenco/collect ^7.0

• Fix Fractal dependency to stay compatible with element-api

• Fix dependencies, now requires PHP >= 7.1

• Fix an issue where engine could be null - #121

• Fix queued indexing error

• Fix Index settings not being forwarded to replicas

• Don't use element as a prop, as it will get serialized with the whole class int the job queue.
• Defer related element propagation to queue

• Use a custom Serializer for Fractal (thanks @santi6291)

• Fixed a type hinting issue

• Fixed a bug when using ->site('*') in your criteria, thanks @timkelty

This is a major release, the way you configure Scout has been changed, please read the docs on how to upgrade.

• Rewrite of the full plugin
• Added tests for all functionality
• Added a comprehensive syntax to configure the Indices
• Added a settings page for all settings except Indices

• Resolve fractal conflict

• Fixed a bug when skipping elements when splitElementIndex is also set (#81)

• Now requires Craft 3.2
• Added a fix for Element drafts & revisions in Craft 3.2

• Revert untested change

• Fixed an issue with settings call

Added

• Scout can now skip elements if you return an empty array from the transformer
• Documented the --force option

Changed

• Scout now uses v2 of the Algolia API in the background. Nothing has changed in Scout usage.

Added

• search_api_key setting to model
• Twig variables for accessing Algolia settings in front end templates

Changed

• Normalised the mixture of quotes in the config example documentation

• Fixed a regression by the previous release when deleting elements would not deIndex them.

• Fixed an issue with deindexing elements, thanks @chrislam

• Fixes issue with console commands on Folder capitalization, thanks @philipzaengle

• Added an indexSettings option to mappings that allows for code-based index settings. Thanks to @timkelty
• Added ./craft scout/settings/update and ./craft scout/settings/dump commands.

• Fixed validator int --> integer
• Added siteId to splitElementIndex

Fixed

• cast site ids to int as craft/yii returns int as string (#39) @larsboldt
• add connect timeout option (#38) @larsboldt
• Replace call to deIndexElements with indexElements (#37)

Fixed

• Fixed an issue where objects weren't being deleted from the index
• Fixed an issue where splitting indices on site or siteId wouldn't work

Added

• Added a "refresh" command, thanks to @JorgeAnzola

Fixed

• Object IDs are now unique for multisites

Added

• Added a sync setting to enable/disable automatic indexing of elements.

Added

• Records can now be split (see: https://www.algolia.com/doc/guides/indexing/structuring-your-data/?language=php#indexing-long-documents) thanks to @larsboldt

Changed

• Changed how the queueing of indexing works to prevent errors and inconsistencies

Fixed

• Fixed an error when trying to serialize an Entry that cannot be serialized

Changed

• Updated the League/Fractal requirement to match that of Element API to avoid conflicts

Added

• Craft Scout now listens to Category events to determine if it needs to reindex elements

Added

• A new icon!

Added

• Console commands to flush & import indexes

  Changed

• Index filtering is now based on criteria instead of a function

Added

• Move indexing to jobs

Added

• Initial release