Spaces:
Build error
Build error
File size: 9,717 Bytes
d5eabda | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 | """
Candidate Post-Processing Component
Handles deduplication, filtering, and diversity enforcement for candidate tracks.
Extracted from the monolithic UnifiedCandidateGenerator for better modularity.
"""
from typing import List, Dict, Any, Set, Optional
import structlog
from collections import defaultdict
from ...models.metadata_models import UnifiedTrackMetadata
class CandidateProcessor:
"""
Processes and refines candidate tracks generated by strategies.
Responsibilities:
- Deduplication
- Quality filtering
- Diversity enforcement
- Source confidence weighting
"""
def __init__(self):
"""Initialize the candidate processor."""
self.logger = structlog.get_logger(__name__)
def process_candidates(
self,
candidates: List[Dict[str, Any]],
enforce_diversity: bool = True,
min_confidence: float = 0.1,
diversity_threshold: float = 0.7
) -> List[Dict[str, Any]]:
"""
Process and refine candidate tracks.
Args:
candidates: Raw candidate tracks from strategies
enforce_diversity: Whether to enforce artist/genre diversity
min_confidence: Minimum source confidence to keep tracks
diversity_threshold: Artist diversity threshold (0.0-1.0)
Returns:
Processed and refined candidate tracks
"""
if not candidates:
return []
self.logger.info(f"🔧 PROCESSING CANDIDATES: {len(candidates)} initial candidates")
# Step 1: Filter by confidence
filtered_candidates = self._filter_by_confidence(candidates, min_confidence)
self.logger.debug(f"After confidence filtering: {len(filtered_candidates)} candidates")
# Step 2: Deduplicate
deduplicated_candidates = self._deduplicate_candidates(filtered_candidates)
self.logger.debug(f"After deduplication: {len(deduplicated_candidates)} candidates")
# Step 3: Enforce diversity if requested
if enforce_diversity:
diverse_candidates = self._enforce_diversity(deduplicated_candidates, diversity_threshold)
self.logger.debug(f"After diversity enforcement: {len(diverse_candidates)} candidates")
else:
diverse_candidates = deduplicated_candidates
# Step 4: Sort by quality metrics
sorted_candidates = self._sort_by_quality(diverse_candidates)
self.logger.info(f"🔧 PROCESSING COMPLETE: {len(sorted_candidates)} final candidates")
return sorted_candidates
def _filter_by_confidence(self, candidates: List[Dict[str, Any]], min_confidence: float) -> List[Dict[str, Any]]:
"""Filter candidates by minimum source confidence."""
filtered = []
for candidate in candidates:
confidence = candidate.get('source_confidence', 0.5)
if confidence >= min_confidence:
filtered.append(candidate)
else:
self.logger.debug(
f"Filtered out low confidence candidate",
track=candidate.get('name', 'Unknown'),
confidence=confidence
)
return filtered
def _deduplicate_candidates(self, candidates: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
"""
Remove duplicate tracks based on multiple criteria.
Uses artist-title combinations and track IDs for deduplication.
"""
seen_tracks: Set[str] = set()
seen_ids: Set[str] = set()
deduplicated = []
for candidate in candidates:
# Create deduplication keys
artist = candidate.get('artist', '').lower().strip()
title = candidate.get('name', '').lower().strip()
track_key = f"{artist}::{title}"
# Check track ID if available
track_id = candidate.get('track_id') or candidate.get('id')
# Skip if we've seen this track before
if track_key in seen_tracks or (track_id and track_id in seen_ids):
self.logger.debug(
f"Skipping duplicate track",
track=candidate.get('name', 'Unknown'),
artist=candidate.get('artist', 'Unknown')
)
continue
# Add to seen sets
seen_tracks.add(track_key)
if track_id:
seen_ids.add(track_id)
deduplicated.append(candidate)
return deduplicated
def _enforce_diversity(self, candidates: List[Dict[str, Any]], diversity_threshold: float) -> List[Dict[str, Any]]:
"""
Enforce artist and genre diversity in candidates.
Args:
candidates: Deduplicated candidates
diversity_threshold: Maximum proportion for any single artist (0.0-1.0)
Returns:
Diversity-enforced candidates
"""
if not candidates:
return candidates
# Group candidates by artist
artist_groups: Dict[str, List[Dict[str, Any]]] = defaultdict(list)
for candidate in candidates:
artist = candidate.get('artist', 'Unknown Artist').lower().strip()
artist_groups[artist].append(candidate)
# Calculate maximum tracks per artist
max_tracks_per_artist = max(1, int(len(candidates) * diversity_threshold))
diverse_candidates = []
artist_track_counts = defaultdict(int)
# Sort candidates by quality for selection
sorted_candidates = self._sort_by_quality(candidates)
# Select candidates while enforcing diversity
for candidate in sorted_candidates:
artist = candidate.get('artist', 'Unknown Artist').lower().strip()
if artist_track_counts[artist] < max_tracks_per_artist:
diverse_candidates.append(candidate)
artist_track_counts[artist] += 1
else:
self.logger.debug(
f"Skipping track for diversity",
track=candidate.get('name', 'Unknown'),
artist=candidate.get('artist', 'Unknown'),
artist_count=artist_track_counts[artist],
max_per_artist=max_tracks_per_artist
)
# Log diversity statistics
unique_artists = len(set(c.get('artist', '').lower() for c in diverse_candidates))
self.logger.debug(
f"Diversity enforcement complete",
total_tracks=len(diverse_candidates),
unique_artists=unique_artists,
max_per_artist=max_tracks_per_artist
)
return diverse_candidates
def _sort_by_quality(self, candidates: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
"""
Sort candidates by quality metrics.
Prioritizes:
1. Source confidence
2. Popularity (if available)
3. Source priority
"""
def quality_score(candidate: Dict[str, Any]) -> float:
confidence = candidate.get('source_confidence', 0.5)
popularity = candidate.get('popularity', 50) / 100.0 # Normalize to 0-1
# Source priority weights
source_weights = {
'target_artist': 1.0,
'similar_artist': 0.9,
'genre_focused': 0.8,
'mood_filtered': 0.8,
'genre_exploration': 0.7,
'underground_gems': 0.6,
'serendipitous_discovery': 0.5,
'mood_based_serendipity': 0.5,
'random_genre_exploration': 0.4
}
source = candidate.get('source', 'unknown')
source_weight = source_weights.get(source, 0.5)
# Combined quality score
quality = (confidence * 0.5) + (popularity * 0.3) + (source_weight * 0.2)
return quality
return sorted(candidates, key=quality_score, reverse=True)
def get_processing_stats(self, candidates: List[Dict[str, Any]]) -> Dict[str, Any]:
"""
Get statistics about the processed candidates.
Args:
candidates: Processed candidates
Returns:
Statistics dictionary
"""
if not candidates:
return {'total': 0}
# Source distribution
source_distribution = defaultdict(int)
confidence_scores = []
for candidate in candidates:
source = candidate.get('source', 'unknown')
source_distribution[source] += 1
confidence_scores.append(candidate.get('source_confidence', 0.5))
# Artist diversity
unique_artists = len(set(c.get('artist', '').lower() for c in candidates))
stats = {
'total': len(candidates),
'unique_artists': unique_artists,
'artist_diversity_ratio': unique_artists / len(candidates) if candidates else 0,
'source_distribution': dict(source_distribution),
'avg_confidence': sum(confidence_scores) / len(confidence_scores) if confidence_scores else 0,
'min_confidence': min(confidence_scores) if confidence_scores else 0,
'max_confidence': max(confidence_scores) if confidence_scores else 0
}
return stats |