danvics/soulsync
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 2

diacritic normalization

Browse source
This commit is contained in:
Broque Thomas 2026-02-14 12:03:49 -08:00
parent 45500892aa
commit c48ac1d4f9
1 changed files with 15 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
database/music_database.py
View file

@ -2257,10 +2257,17 @@ class MusicDatabase:
def _generate_album_title_variations(self, title: str) -> List[str]: def _generate_album_title_variations(self, title: str) -> List[str]:
"""Generate variations of album title to handle edition matching""" """Generate variations of album title to handle edition matching"""
variations = [title] # Always include original variations = [title] # Always include original
# Add diacritic-normalized variation (fixes #101)
# SQLite LIKE is not Unicode-aware, so "găină" won't match "gaina"
# Adding the normalized form lets the SQL query catch both
normalized_title = self._normalize_for_comparison(title)
if normalized_title != title.lower():
variations.append(normalized_title)
# Clean up the title # Clean up the title
title_lower = title.lower().strip() title_lower = title.lower().strip()
# Define edition patterns and their variations # Define edition patterns and their variations
edition_patterns = { edition_patterns = {
r'\s*\(deluxe\s*edition?\)': ['deluxe', 'deluxe edition'], r'\s*\(deluxe\s*edition?\)': ['deluxe', 'deluxe edition'],
@ -2396,7 +2403,12 @@ class MusicDatabase:
def _generate_track_title_variations(self, title: str) -> List[str]: def _generate_track_title_variations(self, title: str) -> List[str]:
"""Generate variations of track title for better matching""" """Generate variations of track title for better matching"""
variations = [title] # Always include original variations = [title] # Always include original
# Add diacritic-normalized variation (fixes #101)
normalized_title = self._normalize_for_comparison(title)
if normalized_title != title.lower():
variations.append(normalized_title)
# IMPORTANT: Generate bracket/dash style variations for better matching # IMPORTANT: Generate bracket/dash style variations for better matching
# Convert "Track - Instrumental" to "Track (Instrumental)" and vice versa # Convert "Track - Instrumental" to "Track (Instrumental)" and vice versa
if ' - ' in title: if ' - ' in title: