that the scanner was run, the ttlcache was also created each time.
This caused (under testing with 166 genres in the database) the
memory consumed by navidrome to 101.18MB over approx 3 days; 96%
of which is in instances of this cache. Swapping to a singleton
has reduced this to down to ~ 2.6MB
Co-authored-by: Rob Emery <git@mintsoft.net>
* Fix album image_files being null.
* Fix small nitpick.
* Use ExecContext instead of Exec.
* Change more columns to not null and set default values.
* Remove columns that don't need to be changed from migration.
* Fix typo.
* Remove unnecessary select statements.
* Remove duplicate code.
* Do not apply changes to radio table.
* Do not apply changes full_text columns and respective indexes.
* Fix musicbrainz columns.
* Rename migration.
* Make ExternalInfoUpdatedAt nullable
* Make Share's timestamps nullable
---------
Co-authored-by: Deluan Quintão <deluan@navidrome.org>
* draft commit
* time to fight pipeline
* round 2 changes
* remove unnecessary line
* fight taglib. again
* make taglib work again???
* add id3 tags
* taglib 1.12 vs 1.13
* use int instead for windows
* store as json now
* add migration, more tests
* support repeated line, multiline
* fix ms and support .m, .mm, .mmm
* address some concerns, make cpp a bit safer
* separate responses from model
* remove [:]
* Add trace log
* Try to unblock pipeline
* Fix merge errors
* Fix SIGSEGV error (proper handling of empty frames)
* Add fallback artist/title to structured lyrics
* Rename conflicting named vars
* Fix tests
* Do we still need ffmpeg in the pipeline?
* Revert "Do we still need ffmpeg in the pipeline?"
Yes we do.
This reverts commit 87df7f6df7.
* Does this passes now, with a newer ffmpeg version?
* Revert "Does this passes now, with a newer ffmpeg version?"
No, it does not :(
This reverts commit 372eb4b0ae.
* My OCD made me do it :P
---------
Co-authored-by: Deluan Quintão <deluan@navidrome.org>
* Start migration to dbx package
* Fix annotations and bookmarks bindings
* Fix tests
* Fix more tests
* Remove remaining references to beego/orm
* Add PostScanner/PostMapper interfaces
* Fix importing SmartPlaylists
* Renaming
* More renaming
* Fix artist DB mapping
* Fix playlist updates
* Remove bookmarks at the end of the test
* Remove remaining `orm` struct tags
* Fix user timestamps DB access
* Fix smart playlist evaluated_at DB access
* Fix search3
* Update mapping.go
fallback in the case there's no Date tagged but Original Date or Release Date are tagged
* Add tests
---------
Co-authored-by: Deluan <deluan@navidrome.org>
* Update mapping.go
CreatedAt = BirthTime
* Update metadata.go
Add BirthTime() function
* Update spread_fs.go
Replacing djherbis/atime package with djherbis/times, as times includes the functionality of atime
* Update go.mod
remove djherbis/atime, add djherbis/times
* Update mapping.go
time package not used anymore
* Update go.sum
removed djherbis/atime, added djherbis/times
* Update spread_fs.go
revert to previous, cannot get rid of /atime after all since it's a dependency of /fscache
* Update go.mod
djherbis/times 1.6.0 now released
* Update go.sum
new sums
* Update metadata.go
Inverted if statement, more readable
* Update go.mod
format fix
* Update go.sum
format fix
* Update go.sum
format fix
* Update go.sum
format fix
* Update metadata.go
variable name times -> fileProperties
check for errors
* Update metadata.go
reverse order of error check
* Update metadata.go
typo
* Update metadata.go
https://github.com/navidrome/navidrome/pull/2553#issuecomment-1787967615
* wip: API endpoint for creating playlists from m3u files
* wip: get user id from context
* temporarily disable failing test
* custom logic for playlist route to accomodate m3u content type
* incorporate playlist parsing into existing logic in core
* re-enable test
* fix locally failing test
* Address requested changes.
* Improve ImportFile tests.
* Remove ownerID as a parameter of ImportM3U.
* Write tests for ImportM3U.
* Separate ImportM3U test into two.
* Test OwnerID and playlist Name.
---------
Co-authored-by: Sam Watson <SwatsonCodes@users.noreply.github.com>
Co-authored-by: caiocotts <caio@cotts.com.br>
* Add support for lyrics tag `unsynced_lyrics`
* Update metadata.go
* Update metadata.go
resolve lint issue
* format the code with `goimports`
format the code with `goimports`
* Fix MB recording ID parameter name for ListenBrainz submission
This follows the ListenBrainz API documentation.
Fixes#1657
* Rename MediaFile.MbzTrackID to MbzRecordingID
This better reflects the actual data. That the MusicBrainz
recording ID is stored in file metadata as musicbrainz_trackid
is a historical artifact.
* Rename database column mbz_track_id to mbz_recording_id
Rob Emery