mirrors/navidrome
1
0
Fork 0
mirror of https://github.com/navidrome/navidrome.git synced 2025-04-03 20:47:35 +03:00
Code Issues Projects Releases Packages Wiki Activity

Fix formatting

Browse source
This commit is contained in:
Deluan 2022-09-26 21:28:10 -04:00
parent 4cf43ed735
commit a7a0e23956
2 changed files with 12 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

14
scanner/tag_scanner.go
View file

@ -56,17 +56,17 @@ const (
// Load all directories from the DB
// Traverse the music folder, collecting each subfolder's ModTime (self or any non-dir children, whichever is newer)
// For each changed folder: get all files from DB whose path starts with the changed folder (non-recursively), check each file:
// if file in folder is newer, update the one in DB
// if file in folder does not exists in DB, add it
// for each file in the DB that is not found in the folder, delete it from DB
// - if file in folder is newer, update the one in DB
// - if file in folder does not exists in DB, add it
// - for each file in the DB that is not found in the folder, delete it from DB
// Compare directories in the fs with the ones in the DB to find deleted folders
// For each deleted folder: delete all files from DB whose path starts with the delete folder path (non-recursively)
// Create new albums/artists, update counters:
// collect all albumIDs and artistIDs from previous steps
// refresh the collected albums and artists with the metadata from the mediafiles
// - collect all albumIDs and artistIDs from previous steps
// - refresh the collected albums and artists with the metadata from the mediafiles
// For each changed folder, process playlists:
// If the playlist is not in the DB, import it, setting sync = true
// If the playlist is in the DB and sync == true, import it, or else skip it
// - If the playlist is not in the DB, import it, setting sync = true
// - If the playlist is in the DB and sync == true, import it, or else skip it
// Delete all empty albums, delete all empty artists, clean-up playlists
func (s *TagScanner) Scan(ctx context.Context, lastModifiedSince time.Time, progress chan uint32) (int64, error) {
ctx = s.withAdminUser(ctx)

8
utils/cache/spread_fs.go vendored
View file

@ -22,9 +22,11 @@ type spreadFS struct {
// NewSpreadFS returns a FileSystem rooted at directory dir. This FS hashes the key and
// distributes all files in a layout like XX/XX/XXXXXXXXXX. Ex:
// Key is abc123.300x300.jpg
// Hash would be: c574aeb3caafcf93ee337f0cf34e31a428ba3f13
// File in cache would be: c5 / 74 / c574aeb3caafcf93ee337f0cf34e31a428ba3f13
//
// Key is abc123.300x300.jpg
// Hash would be: c574aeb3caafcf93ee337f0cf34e31a428ba3f13
// File in cache would be: c5 / 74 / c574aeb3caafcf93ee337f0cf34e31a428ba3f13
//
// The idea is to avoid having too many files in one dir, which could potentially cause performance issues
// and may hit limitations depending on the OS.
// See discussion here: https://github.com/djherbis/fscache/issues/8#issuecomment-614319323