Indexer¶

The indexer is responsible for indexing the music library on the filesystem and populating the database with the available metadata and information.

The indexer runs in two stages:

Scan music directories - Scan the filesystem for new music files and store them in the database.
Extract album art - Extract and store the album art from newly indexed releases. Generate thumbnails for the frontend.

The following function executes both stages:

src.indexer.run_indexer()¶

Return type: None

Code documentation for the indexer implementation is as follows:

Scan music directories¶

Functions:

`catalog_file`(filepath, conn)	Given a file, enter its information into the database.
`handle_track_batch`(track_batch)	Every time we have a batch of tracks, run some logic.
`scan_directories`()	Read the music directories to be indexed from the configuration and scan them for new files.
`scan_directory`(directory)	Scan a given directory for music files and catalog the discovered files.

src.indexer.scanner.catalog_file(filepath, conn)¶

Given a file, enter its information into the database. If associated database objects, e.g. artists and albums, don’t exist, they are created with information from the track.

If a track with this file’s sha256 already exists in the database, the filepath of the existing database row will be updated to the new filepath. No metadata updating will happen.

Parameters

filepath (str) – The filepath of the music file.
conn (sqlite3.Connection) – A connection to the database.

Return type

src.library.track.T

src.indexer.scanner.handle_track_batch(track_batch)¶

Every time we have a batch of tracks, run some logic. This is logic that has a cost–we don’t want to run it once every track, but not to the point where we want to only run it once.

Return type: None

src.indexer.scanner.scan_directories()¶

Read the music directories to be indexed from the configuration and scan them for new files.

Return type: None

src.indexer.scanner.scan_directory(directory)¶

Scan a given directory for music files and catalog the discovered files.

Parameters: directory (pathlib.Path) – The directory to scan.
Raises: NotADirectoryError – If the directory does not exist.
Return type: None

Extract album art¶

Functions:

`save_image`(tf, conn)	If the track has attached cover art, save it to the cover_arts dir with the sha256 of the cover art as the filename.
`save_pending_covers`()	For the releases with pending covers-to-save, look at their first track for a cover art to save.

src.indexer.covers.save_image(tf, conn)¶

If the track has attached cover art, save it to the cover_arts dir with the sha256 of the cover art as the filename.

Otherwise, look in the directory of the file and the directory above it for a cover or folder file (case insensitive) if embedded art does not exist.

If neither exist, return None.

Parameters: tf (tagfiles.TagFile) – The tagfile whose image we want to save.
Return type: typing.Optional[src.library.image.T]
Returns: The filepath of the saved image, if one was saved.

src.indexer.covers.save_pending_covers()¶

For the releases with pending covers-to-save, look at their first track for a cover art to save. If a cover art is found, save it to the COVER_ART_DIR, generate a thumbnail, and update the database with the filename.

Return type: None