* OK key now attempts to send DPAD_CENTER or ENTER (if needed), in attempt to make it work consistently accross all devices
* OK is no longer ignored in numeric mode
improved auto space adjustment around '-' and '/'
improved auto space adjustment when the next character is a punctuation mark
code cleanup: moved auto text case code, auto space code and suggestion loading code to their own classes
* Language IDs are now automatically generated. No more manual housekeeping.
* The database words are properly migrated to avoid dictionary re-import, because of the new IDs
* No unnecessary notifications when 'change language' key is pressed, but there is only one language
* Database code cleanup
* Moved the database exceptions and room classes to separate packages
* Migrations done the right way
* Updated the contribution guide
* more dictionary validations during build time (empty line and punctuation presence)
* slightly improved dictionary loading speed
* removed a blank line from Ukrainian
* OK now performs the default application action correctly, when an action is allowed (e.g. send a message, go to a URL or the next form field)
* updated manual
All categories are now on separate screens
Added a 'Clear Unselected' dictionary option
The 'auto space' setting is now on by default
Reorganized the emoji and added 4 new ones
There was previously no way to access the underscore - it is a necessary character for entering URLs, amongst other things
Added right behind the dash, though where it exists doesnt matter so much as THAT it exists as a choice
* validateDictionaries gradle task now makes use of caching for much faster builds
* lowered the severity of missing translations linting rule
* fixed some more errors in the translations
* added linting task to the GitHub CI validation workflow
* enabled GitHub CI validation on push to master
* all suggestions are now ordered by length, then by frequency
* word frequency is normalized to 255, instead of to 5; normalization now makes sense
* only maxed out languages are normalized, not all
* all words are normalized at once, instead of only the one that has reached the limit
* normalization now happens on start up, instead of using a trigger
* fixed word frequency not updating when a punctuation mark is appended at the end, for example: 'try,'
* switched the positions of ; and :
* updated documentation
* added a github workflow for closing stale issues
* added github PR validation workflows
* fixed the dictionary validator missing digits in words
* fixed the build failing completely when there no git tags
* updated androidx.room 2.4.3 -> 2.5.0
* the dictionary loader now supports word frequencies
* word frequency validation upon building
* added default word frequencies to all dictionaries
* updated documentation
