aherman
/
SoulSync
mirror of https://github.com/Nezreka/SoulSync.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix metadata cross-contamination when downloading concurrent albums

Browse Source
pull/153/head
Broque Thomas 3 months ago
parent 3644422ab8
commit d4d2568f32
1 changed files with 88 additions and 35 deletions
Show Stats Download Patch File Download Diff File

123
web_server.py
Unescape View File

@ -150,6 +150,16 @@ def extract_filename(full_path):
else:
return full_path
def _make_context_key(username, filename):
"""Build a unique context key from username and full Soulseek path.
Uses the full remote path (not just filename) to prevent collisions
when different tracks from the same user share a filename
(e.g., two albums both containing '01 - Intro.flac').
"""
normalized = filename.replace('\\', '/').lstrip('/') if filename else ''
return f"{username}::{normalized}"
# --- Initialize Core Application Components ---
print("🚀 Initializing SoulSync services for Web UI...")
try:
@ -352,7 +362,7 @@ def get_cached_transfer_data():
file_info['username'] = username
all_transfers.append(file_info)
for transfer in all_transfers:
key = f"{transfer.get('username')}::{extract_filename(transfer.get('filename', ''))}"
key = _make_context_key(transfer.get('username'), transfer.get('filename', ''))
live_transfers_lookup[key] = transfer
# Also add YouTube downloads (through orchestrator)
@ -361,7 +371,7 @@ def get_cached_transfer_data():
for download in all_downloads:
# Only add YouTube downloads (Soulseek ones are already in the lookup)
if download.username == 'youtube':
key = f"{download.username}::{extract_filename(download.filename)}"
key = _make_context_key(download.username, download.filename)
# Convert DownloadStatus to transfer dict format
live_transfers_lookup[key] = {
'id': download.id,
@ -577,7 +587,7 @@ class WebUIDownloadMonitor:
task_username = task.get('username') or task.get('track_info', {}).get('username')
if task_filename and task_username:
lookup_key = f"{task_username}::{extract_filename(task_filename)}"
lookup_key = _make_context_key(task_username, task_filename)
live_info = live_transfers_lookup.get(lookup_key)
if live_info:
@ -664,7 +674,7 @@ class WebUIDownloadMonitor:
for directory in user_data['directories']:
if 'files' in directory:
for file_info in directory['files']:
key = f"{username}::{extract_filename(file_info.get('filename', ''))}"
key = _make_context_key(username, file_info.get('filename', ''))
live_transfers[key] = file_info
# Also get YouTube downloads (through orchestrator)
@ -673,7 +683,7 @@ class WebUIDownloadMonitor:
for download in all_downloads:
# Only add YouTube downloads (Soulseek ones are already in the lookup)
if download.username == 'youtube':
key = f"{download.username}::{extract_filename(download.filename)}"
key = _make_context_key(download.username, download.filename)
# Convert DownloadStatus to transfer dict format for monitor compatibility
live_transfers[key] = {
'id': download.id,
@ -718,7 +728,7 @@ class WebUIDownloadMonitor:
if not task_filename or not task_username:
return False
lookup_key = f"{task_username}::{extract_filename(task_filename)}"
lookup_key = _make_context_key(task_username, task_filename)
live_info = live_transfers_lookup.get(lookup_key)
if not live_info:
@ -760,7 +770,7 @@ class WebUIDownloadMonitor:
# Defer orphan cleanup to outside the lock (needs matched_context_lock)
if username and filename:
old_context_key = f"{username}::{extract_filename(filename)}"
old_context_key = _make_context_key(username, filename)
_orphaned_download_keys.add(old_context_key)
deferred_ops.append(('cleanup_orphan', old_context_key))
@ -845,7 +855,7 @@ class WebUIDownloadMonitor:
# Defer orphan cleanup to outside the lock (needs matched_context_lock)
if username and filename:
old_context_key = f"{username}::{extract_filename(filename)}"
old_context_key = _make_context_key(username, filename)
_orphaned_download_keys.add(old_context_key)
deferred_ops.append(('cleanup_orphan', old_context_key))
@ -932,7 +942,7 @@ class WebUIDownloadMonitor:
# Defer orphan cleanup to outside the lock (needs matched_context_lock)
if username and filename:
old_context_key = f"{username}::{extract_filename(filename)}"
old_context_key = _make_context_key(username, filename)
_orphaned_download_keys.add(old_context_key)
deferred_ops.append(('cleanup_orphan', old_context_key))
@ -4065,7 +4075,7 @@ def start_download():
))
if download_id:
# Register download for post-processing (simple transfer to /Transfer)
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
matched_downloads_context[context_key] = {
'search_result': {
@ -4112,7 +4122,7 @@ def start_download():
if download_id:
# Register download for post-processing (simple transfer to /Transfer)
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
is_youtube = username == 'youtube'
with matched_context_lock:
matched_downloads_context[context_key] = {
@ -4186,21 +4196,35 @@ def _find_completed_file_robust(download_dir, api_filename, transfer_dir=None):
# Consolidate multiple spaces
return ' '.join(text.split()).strip()
def _path_matches_api_dirs(file_path):
"""Check if ALL api directory components appear in the file's path."""
path_parts = set(p.lower() for p in file_path.replace('\\', '/').split('/'))
return all(d in path_parts for d in api_dir_parts)
def search_in_directory(search_dir, location_name):
"""Search for the file in a specific directory."""
best_match_path = None
highest_similarity = 0.0
best_fuzzy_path = None
highest_fuzzy_similarity = 0.0
exact_matches = []
# Walk through the entire directory
for root, dirs, files in os.walk(search_dir):
# Skip quarantine folder — contains known-wrong files from AcoustID verification
dirs[:] = [d for d in dirs if d != 'ss_quarantine']
for file in files:
# Direct match is the best case
# Direct basename match
if os.path.basename(file) == target_basename:
file_path = os.path.join(root, file)
print(f"✅ Found exact match in {location_name}: {file_path}")
return file_path, 1.0
# Fast path: if path aligns with expected directory structure, return now
if api_dir_parts and _path_matches_api_dirs(file_path):
print(f"✅ Found path-confirmed match in {location_name}: {file_path}")
return file_path, 1.0
if not api_dir_parts:
# No directory info to disambiguate — return first match (original behavior)
print(f"✅ Found exact match in {location_name}: {file_path}")
return file_path, 1.0
exact_matches.append(file_path)
continue
# Check for slskd dedup suffix (e.g. "Song_639067852665564677.flac")
# slskd appends _<timestamp> when a file with the same name already exists
@ -4208,23 +4232,53 @@ def _find_completed_file_robust(download_dir, api_filename, transfer_dir=None):
stripped_stem = re.sub(r'_\d{10,}$', '', file_stem)
if stripped_stem != file_stem and stripped_stem + file_ext_part == target_basename:
file_path = os.path.join(root, file)
print(f"✅ Found dedup-suffix match in {location_name}: {file_path}")
return file_path, 1.0
if api_dir_parts and _path_matches_api_dirs(file_path):
print(f"✅ Found path-confirmed dedup match in {location_name}: {file_path}")
return file_path, 1.0
if not api_dir_parts:
print(f"✅ Found dedup-suffix match in {location_name}: {file_path}")
return file_path, 1.0
exact_matches.append(file_path)
continue
# Fuzzy matching for variations
normalized_file = normalize_for_finding(file)
similarity = SequenceMatcher(None, normalized_target, normalized_file).ratio()
if similarity > highest_similarity:
highest_similarity = similarity
best_match_path = os.path.join(root, file)
return best_match_path, highest_similarity
if similarity > highest_fuzzy_similarity:
highest_fuzzy_similarity = similarity
best_fuzzy_path = os.path.join(root, file)
# Return best exact match (disambiguated by path), or fall back to fuzzy
if exact_matches:
if len(exact_matches) == 1:
print(f"✅ Found exact match in {location_name}: {exact_matches[0]}")
return exact_matches[0], 1.0
# Multiple files share the basename — pick the one whose path best
# matches the expected directory structure from the Soulseek remote path
best = exact_matches[0]
best_score = -1
for m in exact_matches:
m_parts = set(p.lower() for p in m.replace('\\', '/').split('/'))
score = sum(1 for d in api_dir_parts if d in m_parts)
if score > best_score:
best_score = score
best = m
print(f"⚠️ Found {len(exact_matches)} files named '{target_basename}' in {location_name}, picked best path match: {best}")
return best, 1.0
return best_fuzzy_path, highest_fuzzy_similarity
# Extract filename using the helper function
target_basename = extract_filename(api_filename)
normalized_target = normalize_for_finding(target_basename)
# Extract directory components from the API path for disambiguation.
# When multiple downloads produce the same basename (e.g., "01 - Silent Night.flac"
# from different albums/users), these let us pick the correct file on disk.
api_path_normalized = api_filename.replace('\\', '/') if api_filename else ''
api_dir_parts = [p.lower() for p in api_path_normalized.split('/')[:-1] if p]
# First search in downloads directory
best_downloads_path, downloads_similarity = search_in_directory(download_dir, 'downloads')
@ -4290,8 +4344,7 @@ def get_download_status():
if not filename_from_api: continue
# Check if this completed download has a matched context
# CRITICAL FIX: Use extract_filename() to match storage key format
context_key = f"{username}::{extract_filename(filename_from_api)}"
context_key = _make_context_key(username, filename_from_api)
# Check if this is an orphaned download that completed after retry
if context_key in _orphaned_download_keys:
@ -4440,7 +4493,7 @@ def get_download_status():
# Check if YouTube download is completed and needs post-processing
if download.state and ('succeeded' in download.state.lower() or 'completed' in download.state.lower()):
context_key = f"{download.username}::{extract_filename(download.filename)}"
context_key = _make_context_key(download.username, download.filename)
with matched_context_lock:
context = matched_downloads_context.get(context_key)
@ -6735,7 +6788,7 @@ def _start_enhanced_album_download(enhanced_tracks, unmatched_tracks, spotify_ar
download_id = run_async(soulseek_client.download(username, filename, size))
if download_id:
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# Create context with FULL Spotify track metadata (like Download Missing Tracks modal)
matched_downloads_context[context_key] = {
@ -6779,7 +6832,7 @@ def _start_enhanced_album_download(enhanced_tracks, unmatched_tracks, spotify_ar
download_id = run_async(soulseek_client.download(username, filename, size))
if download_id:
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# Basic context for unmatched tracks (simple cleanup)
matched_downloads_context[context_key] = {
@ -6864,7 +6917,7 @@ def _start_album_download_tasks(album_result, spotify_artist, spotify_album):
download_id = run_async(soulseek_client.download(username, filename, size))
if download_id:
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# Enhanced context storage with Spotify clean titles (GUI parity)
enhanced_context = individual_track_context.copy()
@ -6925,7 +6978,7 @@ def start_matched_download():
download_id = run_async(soulseek_client.download(username, filename, size))
if download_id:
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# Create context with FULL Spotify track metadata (like Download Missing Tracks modal)
matched_downloads_context[context_key] = {
@ -6985,7 +7038,7 @@ def start_matched_download():
download_id = run_async(soulseek_client.download(username, filename, size))
if download_id:
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# THE FIX: We preserve the spotify_album context if it was provided.
# For a regular single, spotify_album will be None.
@ -13559,7 +13612,7 @@ def _run_post_processing_worker(task_id, batch_id):
# Try to get context for generating the correct final filename
task_basename = extract_filename(task_filename)
context_key = f"{task_username}::{task_basename}"
context_key = _make_context_key(task_username, task_filename)
expected_final_filename = None
print(f"🔍 [Post-Processing] Looking up context with key: {context_key}")
@ -14266,7 +14319,7 @@ def _attempt_download_with_candidates(task_id, candidates, track, batch_id=None)
if download_id:
# Store context for post-processing with complete Spotify metadata (GUI PARITY)
context_key = f"{username}::{extract_filename(filename)}"
context_key = _make_context_key(username, filename)
with matched_context_lock:
# Create WebUI equivalent of GUI's SpotifyBasedSearchResult data structure
enhanced_payload = download_payload.copy()
@ -14791,8 +14844,8 @@ def _build_batch_status_data(batch_id, batch, live_transfers_lookup):
task_filename = task.get('filename') or _ti.get('filename')
task_username = task.get('username') or _ti.get('username')
if task_filename and task_username:
lookup_key = f"{task_username}::{extract_filename(task_filename)}"
lookup_key = _make_context_key(task_username, task_filename)
if lookup_key in live_transfers_lookup:
live_info = live_transfers_lookup[lookup_key]
state_str = live_info.get('state', 'Unknown')

Write Preview
Loading…
Cancel
Save