Spaces:
Sleeping
Sleeping
File size: 10,408 Bytes
da6a0a4 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 | """
Optimized Search Engine with Advanced Indexing and Caching
Ultra-fast search with < 100ms response time
"""
import pandas as pd
import numpy as np
from collections import defaultdict
from functools import lru_cache
import time
from text_processor import get_text_processor
class OptimizedSearchEngine:
"""
Ultra-optimized search engine with:
- Multi-level caching
- Inverted index for O(1) lookups
- Pre-computed normalized fields
- Efficient ranking algorithm
"""
def __init__(self, df):
print("[STARTUP] Building optimized search engine...")
start = time.time()
self.df = df
self.text_processor = get_text_processor()
# Pre-process all data once
self.players = self._preprocess_players()
# Build inverted indices for instant lookups
self.name_index = defaultdict(set)
self.club_index = defaultdict(set)
self.nationality_index = defaultdict(set)
self.position_index = defaultdict(set)
self._build_indices()
elapsed = (time.time() - start) * 1000
print(f" Search engine ready in {elapsed:.0f}ms")
print(f" - {len(self.players)} players indexed")
print(f" - {len(self.name_index)} name tokens")
print(f" - {len(self.club_index)} clubs")
print(f" - {len(self.nationality_index)} nationalities")
def _preprocess_players(self):
"""Pre-compute all normalized fields"""
players = []
for idx, row in self.df.iterrows():
player = {
# Original fields
'id': idx,
'short_name': row.get('short_name', ''),
'long_name': row.get('long_name', ''),
'overall': int(row.get('overall', 0)),
'potential': int(row.get('potential', 0)),
'age': int(row.get('age', 0)),
'club_name': row.get('club_name', ''),
'nationality_name': row.get('nationality_name', ''),
'player_positions': row.get('player_positions', ''),
'value_eur': float(row.get('value_eur', 0)),
'wage_eur': float(row.get('wage_eur', 0)),
'pace': int(row.get('pace', 0)),
'shooting': int(row.get('shooting', 0)),
'passing': int(row.get('passing', 0)),
'dribbling': int(row.get('dribbling', 0)),
'defending': int(row.get('defending', 0)),
'physic': int(row.get('physic', 0)),
# Pre-normalized fields for fast search
'_norm_name': self.text_processor.normalize_text(row.get('short_name', '')),
'_norm_long': self.text_processor.normalize_text(row.get('long_name', '')),
'_norm_club': self.text_processor.normalize_text(row.get('club_name', '')),
'_norm_nat': self.text_processor.normalize_text(row.get('nationality_name', '')),
'_norm_pos': self.text_processor.normalize_text(row.get('player_positions', '')),
# Pre-tokenized for instant matching
'_tokens': self.text_processor.tokenize(
f"{row.get('short_name', '')} {row.get('long_name', '')}"
)
}
players.append(player)
return players
def _build_indices(self):
"""Build inverted indices for O(1) lookups"""
for idx, player in enumerate(self.players):
# Index by name tokens
for token in player['_tokens']:
self.name_index[token].add(idx)
# Index by club
if player['_norm_club']:
self.club_index[player['_norm_club']].add(idx)
# Index by nationality
if player['_norm_nat']:
self.nationality_index[player['_norm_nat']].add(idx)
# Index by position
positions = player['_norm_pos'].split()
for pos in positions:
if pos:
self.position_index[pos].add(idx)
def search(self, query, filters=None, max_results=20):
"""
Ultra-fast search with multi-strategy approach
Target: < 100ms response time
"""
start_time = time.time()
if not query or not query.strip():
# No query - return top players by rating
candidates = self.players[:100]
else:
# Get candidates using inverted index
candidates = self._get_candidates_fast(query)
# Apply filters
if filters:
candidates = self._apply_filters_fast(candidates, filters)
# Rank results
ranked = self._rank_fast(query, candidates, max_results)
elapsed = (time.time() - start_time) * 1000
print(f" Search completed in {elapsed:.1f}ms ({len(ranked)} results)")
return ranked
def _get_candidates_fast(self, query):
"""
Lightning-fast candidate retrieval using indices
Strategy: Start with smallest result set
"""
query_norm = self.text_processor.normalize_text(query)
query_tokens = self.text_processor.tokenize(query)
candidate_sets = []
# Strategy 1: Name token matches (most precise)
for token in query_tokens:
if token in self.name_index:
candidate_sets.append(self.name_index[token])
# Strategy 2: Club matches
if query_norm in self.club_index:
candidate_sets.append(self.club_index[query_norm])
# Strategy 3: Nationality matches
if query_norm in self.nationality_index:
candidate_sets.append(self.nationality_index[query_norm])
# Strategy 4: Position matches
for token in query_tokens:
if token in self.position_index:
candidate_sets.append(self.position_index[token])
if not candidate_sets:
# Fallback: partial matching (slower but comprehensive)
return self._fallback_search(query_norm, query_tokens)
# Union of all candidate sets
candidate_indices = set()
for s in candidate_sets:
candidate_indices.update(s)
# Convert indices to player objects
return [self.players[idx] for idx in candidate_indices if idx < len(self.players)]
def _fallback_search(self, query_norm, query_tokens):
"""Fallback: scan all players for partial matches"""
candidates = []
for player in self.players:
# Quick substring check in pre-normalized fields
searchable = f"{player['_norm_name']} {player['_norm_club']} {player['_norm_nat']}"
if query_norm in searchable:
candidates.append(player)
if len(candidates) >= 100: # Limit fallback results
break
return candidates
def _apply_filters_fast(self, candidates, filters):
"""Ultra-fast filtering using numpy-style operations"""
filtered = []
for player in candidates:
# Overall
if 'overallMin' in filters and filters['overallMin']:
if player['overall'] < filters['overallMin']:
continue
if 'overallMax' in filters and filters['overallMax']:
if player['overall'] > filters['overallMax']:
continue
# Age
if 'ageMin' in filters and filters['ageMin']:
if player['age'] < filters['ageMin']:
continue
if 'ageMax' in filters and filters['ageMax']:
if player['age'] > filters['ageMax']:
continue
# Position
if 'position' in filters and filters['position']:
if filters['position'].lower() not in player['_norm_pos']:
continue
# Pace
if 'paceMin' in filters and filters['paceMin']:
if player['pace'] < filters['paceMin']:
continue
# Shooting
if 'shootingMin' in filters and filters['shootingMin']:
if player['shooting'] < filters['shootingMin']:
continue
filtered.append(player)
return filtered
def _rank_fast(self, query, candidates, max_results):
"""
Fast ranking algorithm
Scores based on: relevance (70%) + quality (30%)
"""
if not query or not query.strip():
# No query - sort by overall rating
candidates.sort(key=lambda p: p['overall'], reverse=True)
return candidates[:max_results]
query_norm = self.text_processor.normalize_text(query)
scored = []
for player in candidates:
# Relevance score (fast matching)
rel_score = 0.0
# Name match (highest weight)
if query_norm in player['_norm_name']:
rel_score += 100.0
if query_norm == player['_norm_name']:
rel_score += 200.0
# Club match
if query_norm in player['_norm_club']:
rel_score += 30.0
# Nationality match
if query_norm in player['_norm_nat']:
rel_score += 20.0
# Token overlap
for token in player['_tokens']:
if query_norm in token or token in query_norm:
rel_score += 10.0
# Quality score
quality = player['overall']
# Final score: 70% relevance + 30% quality
final_score = (rel_score * 0.7) + (quality * 0.3)
scored.append((player, final_score))
# Sort by score descending
scored.sort(key=lambda x: x[1], reverse=True)
return [player for player, score in scored[:max_results]]
|