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

adjust album matching

Browse source
This commit is contained in:
Broque Thomas 2025-07-26 23:41:33 -07:00
parent 212d7d3cac
commit c40b85c558
3 changed files with 4412 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

BIN
core/__pycache__/matching_engine.cpython-312.pyc
View file

Binary file not shown.

23
core/matching_engine.py
View file

@ -48,16 +48,27 @@ class MusicMatchingEngine:
def normalize_string(self, text: str) -> str: def normalize_string(self, text: str) -> str:
""" """
Normalizes string by converting to ASCII, lowercasing, and removing Normalizes string by handling common stylizations, converting to ASCII,
specific punctuation while keeping alphanumeric characters. lowercasing, and replacing separators with spaces.
""" """
if not text: if not text:
return "" return ""
text = unidecode(text) text = unidecode(text)
text = text.lower() text = text.lower()
# Keep alphanumeric, spaces, and hyphens, but remove other punctuation like '.' or ','
text = re.sub(r'[^\w\s-]', '', text) # --- IMPROVEMENT V4 ---
# The user correctly pointed out that replacing '$' with 's' was incorrect
# as it breaks searching for stylized names like A$AP Rocky.
# The new approach is to PRESERVE the '$' symbol during normalization.
# Replace common separators with spaces to preserve word boundaries.
text = re.sub(r'[._/]', ' ', text)
# Keep alphanumeric characters, spaces, hyphens, AND the '$' sign.
text = re.sub(r'[^a-z0-9\s$-]', '', text)
# Consolidate multiple spaces into one
text = re.sub(r'\s+', ' ', text).strip() text = re.sub(r'\s+', ' ', text).strip()
return text return text
@ -97,6 +108,8 @@ class MusicMatchingEngine:
# Common album suffixes to remove # Common album suffixes to remove
album_patterns = [ album_patterns = [
# Add pattern to remove trailing info after a hyphen, common for remasters/editions.
r'\s-\s.*',
r'\s*\(deluxe\s*edition?\)', r'\s*\(deluxe\s*edition?\)',
r'\s*\(expanded\s*edition?\)', r'\s*\(expanded\s*edition?\)',
r'\s*\(remastered?\)', r'\s*\(remastered?\)',
@ -363,4 +376,4 @@ class MusicMatchingEngine:
if best_confidence >= 0.8: # High threshold for album matching if best_confidence >= 0.8: # High threshold for album matching
return best_match, best_confidence return best_match, best_confidence
else: else:
return None, best_confidence return None, best_confidence

4394
logs/app.log
View file

File diff suppressed because it is too large Load diff