v5.0.0

• Added a complete caching system to improve the speed of searching by content and copying references to the clipboard
  • Added all necessary unit tests to ensure correct behavior
• Added language support for Swahili (sw)
• Rewrote preferences filter to be more robust
  • Added and revised tests for filtering preferences
• Added unit tests ensuring that all source files have the proper module/encoding headers
• Fixed a major bug in which running unit tests would manipulate the host machine's user preferences (rather than manipulating a separate, temporary copy of user preferences)

v4.6.0

• Fixed a bug where non-ASCII languages (such as Español or العربية) could not be filtered via the yvset filter
• Renamed the searchEngine preference to search_engine for greater consistency with everything else
  • Added code to maintain backwards compatibility; the next time you run YouVersion Suggest, your preferred search engine will be kept
  • The now-deprecated searchEngine key will be removed and declared unsupported in the soon-to-be-released YouVersion Suggest v5. To ensure that your preferred search engine is kept, upgrade to v4.6 before upgrading to v5. If you haven't changed your preferred search engine to something other than Google (the default), then you can upgrade directly to v5 without any trouble.

v4.5.0

• Added language support for Persian (fa)
• Added language support for Bulgarian (bg)
• Corrected several misspelled book names in the Spanish (es) language file
• Fixed several self-cancelling bugs in the 'search reference' unit tests
• Added schema for the search engines data file
• Revised other data file schemas to improve precision, reduce redundancy, and improve overall consistency
• Upgraded coverage and flake8 packages
• Added additional code comments to improve code quality
• Restructured conditional logic in HTML parsers to reduce redundancy and improve readability

v4.4.3

• Upgraded project dependencies
• Added several unit tests to bring coverage back up to 100%
  • Tests handle cases where some originally-assumed givens are nonexistent
• Made minor tweaks to the verse-copying script to reduce complexity
• Revised several code comments to improve clarity and consistency
• Rewrote several getter functions to be simpler and clearer

v4.4.2

• Fixed major bug introduced in v4.0.0 (but only applicable to v4.4.0+) where existing user preferences would not be updated with newly-added preferences (e.g. the recently-introduced new search engine preference)

v4.4.1

• Updated README
  • Revised and reformatted markup
  • Added documentation for setting your preferred search engine
• Fixed a bug where choosing a preference (when using yvset) would not autocomplete the preference name
  • Added unit test to ensure this bug does not recur
• Improve leniency of query parser which handles setting preferences such that

v4.4.0

• Added alternative search engines including Bing and DuckDuckGo
  • You can change your preferred search engine using the yvset keyword
  • As described in the README, you can search the selected reference using your preferred search engine using the ctrl modifier key
• Eliminated inefficiency which caused default preferences to be loaded in addition to user preferences even if user preferences already existed
• Added support for Polish (pl)
• Non-numbered queries no longer match numbered books when filtering references
  • e.g. The query c3 now matches Colossians 3 alone and no longer includes 1 Corinthians 3

v4.3.0

• Added support for Korean (ko)

v4.2.0

• Added support for the following languages:
  • Spanish - Spain (es-ES)
  • Italian (it)
  • Portuguese - Brazil (pt)
  • Portuguese - Portgual (pt-PT)
  • French (fr)
  • German (de)
  • Japanese (ja)
  • Hindi (hi)
  • Russian (ru)
  • Arabic (ar)
• Fixed bug in which punctuation in a book name would be ignored when filtering by reference
  • Added unit test to cover this edge case

v4.1.0

• Added support for Finnish (fi)
• Upgraded project dependencies
• Covered case where workflow attempts to create data directory when it already exists
  • Added unit test to verify this
  • This fixes test coverage, which was listed at 100% but was actually at 99.55% (and thus rounded to 100%). Coverage is now actually 100%
• Fixed style of test docstrings to conform to Python convention
• Fixed a bug in which the workflow utility would always write to the installed owrkflow's info file even if nothing has changed.