• Added the dev settings UI
• Added export and import functionality (import is unfinished)
• Updated the export logs to allow for either sharing or saving direct to files
• Fixed an issue where the input view wouldn't reappear when deleting messages for everyone in a one-to-one conversation
• Fixed an issue where the input view would be visible while the loading indicator was visible when deleting from a community
• Fixed an issue where notifications weren't being removed after processing an UnsendRequest
• Fixed an issue where legacy groups had a "Clear for everyone" option but it didn't do anything
• Updated the copy for legacy groups to say "Delete for everyone" instead of "Clear for everyone"
• Updated the SessionCallManager to be an updated singleton type (cleaned up more in Groups Rebuild)
• Updated the PushRegistrationManager to be an updated singleton type (cleaned up more in Groups Rebuild)
• Injected dependencies correctly in a bunch of places
• Added a few functions to retrieve conversation settings from libSession
• Updated the Note to Self swipe action to be "Hide" (hides the conversation but does not delete the messages)
• Updated the one-to-one deletion behaviour (now syncs both hiding the conversation and deleting it's messages)
• Updated the logic to retrieve the relevant disappearing messages setting from libSession when creating a thread if it doesn't exist (allows us to delete threads without worrying about losing settings)
• Updated a bunch of dependency management & injection code so the unit tests would pass
• Refactored the 'RecipientState' type to just be columns on the 'Interaction' table
• Added some syntactic sugar to clean up the 'CallbackWrapper' boilerplate within the LibSession+Networking code
• Fixed an issue where sending a message into a group where you are the only member would result in the message permanently being in the sending state
• Fixed an issue when initially setting up the interaction database structure where the community whisper mechanism needed another column
• Fixed a bug where the MessageSendJob wouldn't time out during path building
• Fixed a bug where the MessageSendJob wouldn't retry 8 times before detecting a timeout (meaning it would retry up to 80 times in total before permanently failing)
• Added new 'LintControl' mechanisms to allow for ignoring sections of code
• Added new 'MatchType' mechanisms for excluding unlocalized cases
• Updated the code to explicitly handle localized template strings (both single and multiline)
• Updated the code to process files across multiple threads to improve performance
• Updated the code to use Swift 5.7 regex and store in static variables to prevent reconstruction every time they are used
• Removed the list of individual files which are ignored (now just output a count)
• Fixed an issue where having a localized string on a subsequent line could result in an unlocalized (or incorrectly localized) string not being detected
• Fixed an issue where having multiple strings on a single line could result in an unlocalized string not being detected
• Fixed an issue where zero-width characters would result in the variable count comparison between translations failing
• Fixed a number of localization warnings
• Fixed a bug where the libSession logs wouldn't all come through correctly
• Added logic to handle response headers from libSession
• Minor optimisation to closed groups poller setup
• Minor logging tweaks
• Reworked the NotificationServiceExtension to just always reset and reload it's state to avoid weird bugs
• Updated the legacy group messages to fallback to using a locally generated serverHash if one isn't provided (always happens for legacy PNs)
• Include error info when failing to process extension logs
• Made a bunch of the Storage functions instance functions instead of static functions
Doing a config push was working via `Atomic.wrappedValue` instead of `Atomic.mutate` - we need to use `Atomic.mutate` in this case because we want to block other threads from modifying the config while we are doing a push (otherwise it can crash due to data changing while generating the push payload)
There were a few logs related to bad memory access within libSession and the `toLibSession` convention which we'd originally setup made me nervous as C-friendly objects would be allocated in Swift and then assigned to C struct properties but I'm not sure how the memory would actually behave in this case.
This updated approach unfortunately involves a bunch of duplicate code within 'TypeConversion+Utilities' (and some horrible type aliases for tuples) but as a result we now have compile-time safe 'get' and 'set' functions for any C struct which conforms to `CAccessible` and `CMutable`. The other nice benefit about this change is that the new `set` functions copy memory across into the C structs so we can have more confidence that the memory ownership has shifted to the libSession side of things.
Morgan Pretty
Ryan ZHAO
Morgan Pretty
RyanZhao