Skip to content

Releases: ChristopherRabotin/bungiesearch

Version 1.3.0

04 Dec 12:55
Compare
Choose a tag to compare

Release notes

New features

None.

Improvements

  • Much better handling of elasticsearch clusters: now waits for green status when building the index, run tests on a cluster (instead of a single instance).

Bug fixes

  • Fixed potential 503 issue when rebuilding too soon if running on an elasticsearch cluster.

Release contributors

Many thanks to @NullSoldier who wrote all the pull requests for this release.

Pull request summary

  • Default test settings to localhost #126
  • Run all travis tests on a real elasticsearch cluster #125
  • Wait for cluster to reach green status after creation #124
  • Removed pyc that was accidently checked in #123
  • Promoted requirements to dev requirments and not just for testing #122
  • Wait for create API call to finish #117

Version 1.2.2

05 Nov 12:19
Compare
Choose a tag to compare

Release notes

Improvements

None. =(

Bug fixes

  • Fixed #109 via #110 : Options update and index no longer mutually exclusive when updating the index.
  • Fixed #114 via #113 : bound supported versions of elasticsearch.

Release contributors

Thanks to @einvalentin for #113 and @NullSoldier for #116 .

Installation

Please use PyPI to install bungiesearch. I forgot to commit 42a273d before creating this tag, but the version on PyPI has the correct setup.py file.

Version 1.2.1

10 Aug 10:50
Compare
Choose a tag to compare

Release notes

Improvements

  • Function update_index of utils now supports any model_items which are iterable, i.e. have the __iter__ attribute.

Bug fixes

  • Fixed #105 : version 1.2.0 could not be installed due to README.md file not found error.

Release contributors

Thanks to @afrancis13 for finding the bug with the README in distributed package (I now check locally that the dist file works before uploading it), and for proposing to allow for different iterable model_items instead of just list and tuple.

Version 1.2.0

07 Aug 15:32
Compare
Choose a tag to compare

Release notes

New features

  • Supports prepare_{} functions in ModelIndex definitions, similar to haystack.
  • Supports templates for models, similar to haystack.
  • Additional management commands: clear_index and rebuild_index.

Improvements

  • README improvements by @meninoebom. More needed for the new features (cf. #101).
  • Django manager now using contribute_to_class (as it should have since the start).
  • Coverage test reports added, and improved.

Bug fixes

  • indexing_query is now correctly called when defined in the ModelIndex.

Release contributors

Many thanks to @afrancis13 and @meninoebom for their contributions.

Pull request summary

  • Remove sleep and add ability to bulk delete from index #88 by @afrancis13
  • Update README.md with installation instructions #91 by @meninoebom
  • Add the ability to use custom analyzers in bungiesearch. #94 by @afrancis13
  • Add data template handling #95 by @afrancis13
  • Adds coverall coverage test and badge to README file #97 by me
  • Fix indexing_query in update command #98 by @afrancis13
  • Performance boost for bulk delete, and additional abstraction in utils #99 by @afrancis13
  • Add prepare_{} field evaluation #100 by @afrancis13
  • Use contribute_to_class in Bungiesearch Manager #103 by @afrancis13
  • Additional management commands for bungiesearch #102 by @afrancis13
  • Added guilty-as-charged flag in clear_index command. #104 by me

Version 1.1.0

26 Jun 12:45
Compare
Choose a tag to compare

Release notes

New features

  • Supports elasticsearch-dsl-py==0.0.4.
  • Supports custom primary key field names.
  • Supports custom signals classes (to add functionality to the post save and pre delete signals for example).

Bug fixes

  • Requesting only the _id field via fields() does not prevent automatic database mapping.
  • StringFields whose value is None is now indexed as null instead of as an empty string, allowing for proper filtering.

Release contributors

Many thanks to @afrancis13 for #76, #77 and #79.

Pull request summary

  • Version 1.1.0 changes #86
  • Fixes fetching only specific fields, including only _id. #84
  • Formatted code, organized imports and separated tests #83
  • StringField returns None if the serialized value is None (instead of an empty string). #82
  • Add ability to create custom signal processing classes #79
  • Fix id_fields issue #77
  • Make bungiesearch compatible with elasticsearch-dsl 0.0.4 #76

Version 1.0.2

05 Jun 17:08
Compare
Choose a tag to compare

Release notes

New features

  • Now on PyPI (first version 1.0.2º).
  • Travis CI

º: I created and uploaded to PyPI an incorrect distribution for version 1.0.0, and PyPI does not support overwriting of releases (cf. this issue).

Bug fixes

  • Update mapping no longer crashes if a given model is defined on multiple indices.
  • Redundant slicing of results prevented correct iteration if start position was set and non-zero.

Release contributors

Many thanks to @JoshStegmaier for coding up #70 and to @afrancis13 for coding up #74. Thanks also to @Folcon for reporting #64.

Pull request summary

  • Fixes update mapping when model is defined on multiple indices #69
  • Removed redundant slicing that caused incorrect results #70
  • Add continuous integration #74
  • Adds PyPI support with LICENSE. Bungiesearch now in version 1.0.2. #75

Version 0.1.0

09 Mar 15:33
Compare
Choose a tag to compare

Supports elasticsearch-dsl Search's features, as well as Django model integration with Managers, settings and the whole lot. More information in the README.md file.