* Rework startup args
* Preserve previous startup args on re-launch
* Ignore subprocess when checking for duplicate processes
* Cleanup code
* Remove extra process list grabbing
* Use `IsUpgradePrefix` when using `RestartApplication`
* Change `ProxyServerPrefix` to `ProxyUrlPrefix`
This permits .NET to download the Windows targeting packs, which enables VRCX development on Linux.
It's not recommended to build the Windows apps with Linux though, so actual release builds should likely stick with Windows CI/CD, as per Microsoft:
> This property is currently recommended to allow development on non-Windows platforms. But when the application is ready to be released, it should be built on Windows. When building on a non-Windows platform, the output may not be the same as when building on Windows. In particular, the executable is not marked as a Windows application (which means that it will always launch a console window) and won't have an icon embedded.
https://learn.microsoft.com/en-us/dotnet/core/project-sdk/msbuild-props#enablewindowstargeting
* Optimized search screenshots by metadata
* feat: Screenshot metadata search bar
* fix: Reset search when selecting a file manually
* refactor: Re-do the whole search thing. Add number of results to dialog when searching
* fix: Add check & error for null metadata
* fix: Add sourceFile to error obj on return
* fix: Fix screenshot file dialog not sending path back to JS
* fix: Stop lfs parsing from dying if a value doesn't exist
* fix: Fix and optimize FileStream reading of metadata for searches
* fix: Reset search data and revert to normal when user clears out search box
* refactor: Remove/optimize some old screenshot helper stuff
- Use FileStream in ReadPNGResolution
- Limit the FindChunkIndex search range used when writing metadata
- Remove old ReadPNGDescription, just use filestream version now
* fix: Reset metadata search state if a file is added manually
* feat: Move viewer popover dialog to the fullscreen image viewer
* refactor: Change how parsing errors are handled... again
* refactor: Let the search carousel loop around
* fix: Re-do legacy parsing /wo JObject. Fix legacy instance ids/pos.
Also adds further docs to the legacy parsing for the various formats
* feat: Add persistent metadata cache for search
* Clean up
* fix: Fix viewer dying
sourceFile wasn't being included for vrcx pics
* refactor: Cache the state of files with no metadata
This is so we're not constantly re-processing these files with no metadata on every first search after a restart; These files won't magically gain metadata and this could cause a lot of hitching for someone that had potentially thousands of screenshots before using VRCX.
* Screenshot viewer loading
---------
Co-authored-by: Nekromateion <43814053+Nekromateion@users.noreply.github.com>
Co-authored-by: Natsumi <cmcooper123@hotmail.com>
* fix: Fix world-db not getting current world properly when traveling
* fix: Stop redundant and exception-prone funcResult definition
* Fix: fetching current location
* refactor: Move constructing responseData to its own functions
* Fix: ignore own string requests
* feat: Add NLog dependency, and add some logging to .NET, mostly worlddb
* fix: I missed a semicolon
* refactor: Add more debug logging, change log format, archive less
* feat: Add /getall endpoint
---------
Co-authored-by: Natsumi <cmcooper123@hotmail.com>
* chore: Change vscode workspace settings to work with omnisharp
* refactor(.NET): Use connection string builder to init sqlite database
* docs(.NET): Add method documentation to most things that matter
* docs(.NET): Add more docs I forgot to commit apparently
* feat: Add PoC world database structure ^& http listener
* fix: Send a response if VRCX isn't initialized rather than hanging
* feat: Initialize world db schema on startup
* feat: Allow worlds to store data in db through logfile
* use existing current location for worldDB
* Add tooltips
* chore: Make it so vscode can format C# files without prettier
* refactor: Add sqlite-net to (eventually) replace sqlite impl
* refactor: Make use of sqlite-net for world database
* docs: Add todo for fixing some random exception
* refactor: Remove now-unused SQLiteWorld
* refactor: Fix DB init query and change table structure again
* refactor: Add WorldDataRequest, add attributes for camelcase json keys
* Support current user location from API in addition to gameLog
* Change current location check for worldDB
* feat: Take store requests in JSON, identify worlds by GUID on store.
* refactor: Remove unused worldId param from connection key generator
* docs: Add more documentation to the methods for the world database
* fix: Hey wait that's not a primary key
* feat: Add a 10MB data cap for worlds shared across all of their rows.
* fix: Don't calculate size of world date twice when inserting
* refactor: Discard the guid variable since we only check for validity
* docs: Add docs/comments for new data cap functionality
* feat: Implement /getbulk API endpoint
* fix: Correct WorldDB init query typo
* fix: Update data entries properly instead of using 'OR REPLACE'
* refactor: Move endpoint processing to separate methods
* refactor: Add another check for error 503, remove old code
* feat: Add debug capability to /vrcx/getbulk
* fix: Correct the usage of getUser in actuallyGetCurrentLocation
* feat: Add store errors, implement external reading, stop 404ing
* docs: Add docs for new world db funcs
* refactor: Change world db server listen port to 22500
* fix: Use getUser correctly, dumb dumb
* fix: This error set shouldn't be here
* feat: Future-proof api endpoints. Add /status endpoint
---------
Co-authored-by: Natsumi <cmcooper123@hotmail.com>
