audiomuze / tagminder Goto Github PK

Current table names do not follow a convention making it harder to know which to refer to. Decide on naming convention and implement. e.g. use utility_ or _info as a prefix for all utility tables.

Leave LYRICIST & WRITER tgs intact, but incorporate into COMPOSER because most music servers do not distinguish them from COMPOSER.

e.g.

• adding DATE tag to albums /tracks without dates
• GENRE tag to albums without genres etc
• no track artist
• no track title

Incorporate MusicBrainz and Discogs DB's as metadata enrichment sources for contributor metadata.

As above,

Add MusicBrainz identifiers to all LABEL, WORK & PART tags leveraging a download of tables from the MusicBrainz database. Codebase already enriches from existing metadata in alib or leverages musicbrainz table if present - extend to include above mentioned metadata.

Create optional function to strip enhanced metadata from album names ... will be useful when a music server properly leverages VERSION, bit depth and sample rate information in order to differentiate releases from one another.

Organise code into logical modules to enhance maintainability.

As above.

Add capability to rename folders holding albums, leveraging:

• albumartist
• album
• /OST -
• /VA -
• compilation = 1
• VERSION tag
• file bit depth and sampling rate signifying > redbook or DSD
• Mixed resolution (increasingly common with re-releases and bonus material)

e.g.:

albumartist - album [version] [bit-depth/sampling rate kHz]
albumartist - album [version] [Mixes Res]

Undo MusicBrainz' unholy mess of handling classical composers by shoving them into the album name. Starting point would be to build a table of composer names where GENRE = Classical, then working from there.

For original releases set both dates the same

For reissues and remasters investigate how music servers handle this (principally Logitehmediaserver or whatever it gets renamed to) and implement accordingly.

Add switch to toggle in-flight chatter status

As above.

Code already developed - generate and incorporate albums_missing_tracks table

As above.

Code base to ID exceptions already exists - just needs implementation.

Remove zero-stuffing from track numbers where not needed i.e. all values 0-9 in disc and track numbers do not require zero-stuffing

Leverage

• inferences based on existing track related metadata in table e.g. composer name
• MusicBrainz
• Discogs
• Allmusic
• RYM

Would need adding album identifiers from each metadata source - refer wikidata.

https://rateyourmusic.com/wiki/RYM:Capitalization

Most efficient method is likely to involve using a transformation function in Pandas to enable scalar updates and quick honing in on deltas requiring updates to alib.

Currently disambiguated metadata lives independently of lists produced by string-grouper.

Enhance code to remove previously assigned false-positives from result set, avoiding rework and improving metadata cleanup outcomes.

Improve documentation to enhance usability

Implement argv to enable passing of switches determining which metadata enhancement routines are run and to enable pointing to source/target root folder rather than using only current dir and hard-coded defaults.