* Perform manual isolation of display names etc. instead of relying on BidiFormatter.
Fixes#1921
* Make follow request notification header formatting more like other notifications
* Delete scheduled post after editing
This ensures that editing a scheduled toot will not duplicate it.
Fixes#1924
* Fix scheduled toot editing
* Convert isEditingScheduledToot to property
* Replace MediatorLiveData with combineLiveData
* Remove dead code
* Fix observer emission
* Fix "Finish media upload" appearing for split second
* Open photos embedded in preview cards in the internal image viewer instead of opening the browser
* Enable toolbar for single image viewer
* Apply review feedback
* Change characters left color to red when negative
* Use val for the integers, optimize variable assignment
* Added tusky_red color
* remove extra line typo
* make characters left bold
* change red hue
* fix typo
* Improve notifications fetching
- Only fetch notifications up to the latest fetched one
- Use timeline markers to avoid showing already seen notifications
* Apply some of the suggestions
* convert EmojiPreference and EmojiCompatFont to Kotlin
* move preference related to to dedicated preference package
* update proguard-rules.pro
* reformat & add comment
* maintain disposable information in EmojiPreference instead of EmojiCompatFont
* Increase character limit for media descriptions to 1500
It was increased in Mastodon 3.0.0 which was released in October 2019.
* Improve image description view
Since media descriptions can be longer now, we need to adjust the UI.
It is a common problem that description takes up the whole screen, it's
hard for readers and also discourages people from adding descriptions.
This commit uses bottom sheet to hide most of the description. Since we
know how much screen space it will cover, we can use darker background
which makes reading text easier.
* Adjust description handle
* Fix unable to dismiss image caption
This commit does 3 things:
1. Replaces PhotoView (which is abandonware) with modern TouchImageView
2. Fixes an issue with panning images. Gesture was not intercepted
properly and pager was taking control instead of image being moved.
3. Adds feedback to dismissing of images with vertical gesture.
At some point settings DSL was refactored to first add a preference and
then run the builder. We shouldn't add a preference to the hierarchy
without setting a key for the preference first because preference gets
it's default value in `onAttachedToHierarchy()` and if the key is not
set then no default value will be set either.
This commit changes the order to execute builder (and set the key)
first and and preference to the point later.
* Refactor main preferences to use DSL
* Refactor account preferences to use DSL
* Use DSL in rest of the preference screens
* Preferences cleanup
* Fix preference dependencies
