Fix album splitting in media servers for multi-source downloads
- Files downloaded from different soulseek sources carried conflicting MusicBrainz Release IDs and disc number tags that were never overwritten during metadata enhancement, causing Navidrome/Plex to split a single album into multiple entries even when folder structure and album name were identical - disc_number is now written to file metadata (sourced from Spotify/iTunes API), overwriting stale tags from soulseek sources - MusicBrainz IDs are now stripped during metadata enhancement so media servers group by album name + artist instead of mismatched release IDs - The verification worker now applies _resolve_album_group() for album name consistency with the stream processor path - Fixed source reuse retry logic: used_sources is no longer reset between retries, so a source that errors during transfer is skipped on retry and the system falls through to a normal search instead of retrying the same failing source 3 times
This commit is contained in:
Broque Thomas
parent
af7279a3ea
commit
863f04c5ae
1 changed files with 101 additions and 12 deletions
113
web_server.py
113
web_server.py
|
|
@ -7705,6 +7705,61 @@ from mutagen.mp4 import MP4, MP4Cover
|
|||
from mutagen.oggvorbis import OggVorbis
|
||||
import urllib.request
|
||||
|
||||
def _strip_musicbrainz_ids(audio_file) -> None:
|
||||
"""
|
||||
Remove MusicBrainz release/album/track IDs from audio file tags.
|
||||
Files from different soulseek sources carry different MusicBrainz IDs,
|
||||
which causes media servers (Navidrome, Plex) to split identically-named
|
||||
albums into separate entries even when album name and artist match.
|
||||
Operates on a non-easy-mode MutagenFile object (caller must save).
|
||||
"""
|
||||
try:
|
||||
removed = []
|
||||
|
||||
# ID3 (MP3): MusicBrainz IDs stored as TXXX custom frames
|
||||
if hasattr(audio_file, 'tags') and audio_file.tags and hasattr(audio_file.tags, 'getall'):
|
||||
mb_descs = [
|
||||
'MusicBrainz Album Id',
|
||||
'MusicBrainz Release Group Id',
|
||||
'MusicBrainz Release Track Id',
|
||||
'MusicBrainz Artist Id',
|
||||
'MusicBrainz Album Artist Id',
|
||||
'MusicBrainz Album Type',
|
||||
'MusicBrainz Album Status',
|
||||
'MusicBrainz Album Release Country',
|
||||
'MUSICBRAINZ_ALBUMID',
|
||||
'MUSICBRAINZ_RELEASEGROUPID',
|
||||
]
|
||||
for txxx in audio_file.tags.getall('TXXX'):
|
||||
if txxx.desc in mb_descs:
|
||||
audio_file.tags.delall(f'TXXX:{txxx.desc}')
|
||||
removed.append(f'TXXX:{txxx.desc}')
|
||||
# Also remove UFID (unique file identifier) MusicBrainz frames
|
||||
for ufid_key in [k for k in audio_file.tags if k.startswith('UFID:')]:
|
||||
if 'musicbrainz' in ufid_key.lower():
|
||||
del audio_file.tags[ufid_key]
|
||||
removed.append(ufid_key)
|
||||
|
||||
# FLAC / OGG Vorbis: MusicBrainz IDs stored as vorbis comments
|
||||
if hasattr(audio_file, 'tags') and audio_file.tags and hasattr(audio_file.tags, 'keys'):
|
||||
mb_vorbis_keys = [k for k in audio_file.tags.keys() if 'musicbrainz' in k.lower()]
|
||||
for key in mb_vorbis_keys:
|
||||
del audio_file.tags[key]
|
||||
removed.append(key)
|
||||
|
||||
# MP4 (M4A/AAC): MusicBrainz IDs stored as freeform atoms
|
||||
if hasattr(audio_file, 'tags') and audio_file.tags:
|
||||
mb_mp4_keys = [k for k in audio_file.tags.keys()
|
||||
if isinstance(k, str) and 'musicbrainz' in k.lower()]
|
||||
for key in mb_mp4_keys:
|
||||
del audio_file.tags[key]
|
||||
removed.append(key)
|
||||
|
||||
if removed:
|
||||
print(f"🧹 Stripped MusicBrainz IDs: {', '.join(removed)}")
|
||||
except Exception as e:
|
||||
print(f"⚠️ Error stripping MusicBrainz IDs (non-fatal): {e}")
|
||||
|
||||
def _enhance_file_metadata(file_path: str, context: dict, artist: dict, album_info: dict) -> bool:
|
||||
"""
|
||||
Core function to enhance audio file metadata using Spotify data.
|
||||
|
|
@ -7749,12 +7804,18 @@ def _enhance_file_metadata(file_path: str, context: dict, artist: dict, album_in
|
|||
|
||||
audio_file.save()
|
||||
|
||||
# Embed album art if enabled
|
||||
if config_manager.get('metadata_enhancement.embed_album_art', True):
|
||||
# Re-open in non-easy mode for embedding art
|
||||
audio_file_art = MutagenFile(file_path)
|
||||
_embed_album_art_metadata(audio_file_art, metadata)
|
||||
audio_file_art.save()
|
||||
# Re-open in non-easy mode for MusicBrainz cleanup and album art embedding.
|
||||
# Files from different soulseek sources carry different MusicBrainz Release IDs,
|
||||
# which causes media servers (Navidrome, Plex) to split identically-named albums
|
||||
# into separate entries. Stripping these ensures grouping by album name + artist.
|
||||
audio_file_raw = MutagenFile(file_path)
|
||||
if audio_file_raw is not None:
|
||||
_strip_musicbrainz_ids(audio_file_raw)
|
||||
|
||||
if config_manager.get('metadata_enhancement.embed_album_art', True):
|
||||
_embed_album_art_metadata(audio_file_raw, metadata)
|
||||
|
||||
audio_file_raw.save()
|
||||
|
||||
print("✅ Metadata enhanced successfully.")
|
||||
return True
|
||||
|
|
@ -7877,6 +7938,17 @@ def _extract_spotify_metadata(context: dict, artist: dict, album_info: dict) ->
|
|||
metadata['track_number'] = 1
|
||||
metadata['total_tracks'] = 1
|
||||
|
||||
# Always write disc_number to overwrite any stale tags from the soulseek source.
|
||||
# Without this, original disc tags persist and can cause media servers (Plex) to
|
||||
# split a single album into standard/deluxe based on differing disc numbers.
|
||||
# Priority: original_search context (from API) > album_info > default to 1
|
||||
disc_num = original_search.get('disc_number')
|
||||
if disc_num is None and album_info:
|
||||
disc_num = album_info.get('disc_number')
|
||||
if disc_num is None:
|
||||
disc_num = 1
|
||||
metadata['disc_number'] = disc_num
|
||||
|
||||
if spotify_album and spotify_album.get('release_date'):
|
||||
metadata['date'] = spotify_album['release_date'][:4]
|
||||
|
||||
|
|
@ -12414,9 +12486,20 @@ def _run_post_processing_worker(task_id, batch_id):
|
|||
'confidence': 0.9,
|
||||
'source': 'verification_worker_corrected'
|
||||
}
|
||||
|
||||
print(f"🎯 [Verification] Created proper album_info - track_number: {track_number}, album: {spotify_album.get('name')}")
|
||||
|
||||
|
||||
# Apply album grouping for consistency with stream processor path.
|
||||
# Without this, the verification worker could write a different album
|
||||
# name than the stream processor (e.g. raw API name vs resolved name),
|
||||
# causing media servers to split tracks into separate albums.
|
||||
try:
|
||||
original_album_ctx = original_search.get('album') if isinstance(original_search.get('album'), str) else None
|
||||
consistent_album_name = _resolve_album_group(spotify_artist, album_info, original_album_ctx)
|
||||
album_info['album_name'] = consistent_album_name
|
||||
except Exception as group_err:
|
||||
print(f"⚠️ [Verification] Album grouping failed, using raw name: {group_err}")
|
||||
|
||||
print(f"🎯 [Verification] Created proper album_info - track_number: {track_number}, album: {album_info['album_name']}")
|
||||
|
||||
print(f"🎵 [Post-Processing] Attempting metadata enhancement for: {found_file}")
|
||||
enhancement_success = _enhance_file_metadata(found_file, context, spotify_artist, album_info)
|
||||
|
||||
|
|
@ -12879,21 +12962,25 @@ def _attempt_download_with_candidates(task_id, candidates, track, batch_id=None)
|
|||
enhanced_payload['artists'] = [{'name': artist} for artist in track.artists] if track.artists else []
|
||||
print(f"✨ [Context] Using clean Spotify metadata - Album: '{track.album}', Title: '{track.name}'")
|
||||
|
||||
# CRITICAL FIX: Get track_number from Spotify API like GUI does
|
||||
# CRITICAL FIX: Get track_number and disc_number from Spotify API like GUI does
|
||||
if hasattr(track, 'id') and track.id:
|
||||
try:
|
||||
detailed_track = spotify_client.get_track_details(track.id)
|
||||
if detailed_track and 'track_number' in detailed_track:
|
||||
enhanced_payload['track_number'] = detailed_track['track_number']
|
||||
print(f"🔢 [Context] Added Spotify track_number: {detailed_track['track_number']}")
|
||||
enhanced_payload['disc_number'] = detailed_track.get('disc_number', 1)
|
||||
print(f"🔢 [Context] Added Spotify track_number: {detailed_track['track_number']}, disc_number: {enhanced_payload['disc_number']}")
|
||||
else:
|
||||
enhanced_payload['track_number'] = track_info.get('track_number', 1)
|
||||
enhanced_payload['disc_number'] = track_info.get('disc_number', 1)
|
||||
print(f"⚠️ [Context] No track_number in detailed_track, using track_info fallback: {enhanced_payload['track_number']}")
|
||||
except Exception as e:
|
||||
enhanced_payload['track_number'] = track_info.get('track_number', 1)
|
||||
enhanced_payload['disc_number'] = track_info.get('disc_number', 1)
|
||||
print(f"❌ [Context] Error getting track_number, using track_info fallback: {enhanced_payload['track_number']} ({e})")
|
||||
else:
|
||||
enhanced_payload['track_number'] = track_info.get('track_number', 1)
|
||||
enhanced_payload['disc_number'] = track_info.get('disc_number', 1)
|
||||
print(f"⚠️ [Context] No track.id available, using track_info fallback track_number: {enhanced_payload['track_number']}")
|
||||
|
||||
# Determine if this should be treated as album download
|
||||
|
|
@ -13081,6 +13168,8 @@ def _try_source_reuse(task_id, batch_id, track):
|
|||
_sr.info(f"Artist verification: artist='{artist_name}', verified={len(verified)}, using={len(final_candidates)} candidates")
|
||||
|
||||
# Initialize task state for download attempt
|
||||
# IMPORTANT: Preserve used_sources from previous attempts (e.g. monitor error retries)
|
||||
# so that _attempt_download_with_candidates skips sources that already failed.
|
||||
with tasks_lock:
|
||||
if task_id in download_tasks:
|
||||
download_tasks[task_id]['status'] = 'searching'
|
||||
|
|
@ -13088,7 +13177,7 @@ def _try_source_reuse(task_id, batch_id, track):
|
|||
download_tasks[task_id]['current_candidate_index'] = 0
|
||||
download_tasks[task_id]['retry_count'] = 0
|
||||
download_tasks[task_id]['candidates'] = []
|
||||
download_tasks[task_id]['used_sources'] = set()
|
||||
# Don't reset used_sources — the download monitor marks failed sources here
|
||||
|
||||
# Attempt download
|
||||
success = _attempt_download_with_candidates(task_id, final_candidates, track, batch_id)
|
||||
|
|
|
|||
Loading…
Reference in a new issue