Spaces:
Running
Running
File size: 9,714 Bytes
09801ca | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 | """
Query Decomposer - Break complex queries into sub-queries
==========================================================
Handles multi-part questions by:
1. Detecting if query is complex
2. Breaking into sub-queries
3. Merging results back together
Examples:
- "What is total revenue and who are top customers?" -> 2 sub-queries
- "Compare Q1 vs Q2 and show trends" -> 2 sub-queries
"""
import re
from typing import List, Dict, Tuple, Optional
from dataclasses import dataclass
import logging
logger = logging.getLogger(__name__)
@dataclass
class SubQuery:
"""A decomposed sub-query"""
query: str
query_type: str # factual, aggregation, comparison, etc.
priority: int = 1
depends_on: Optional[int] = None # Index of query this depends on
@dataclass
class DecomposedQuery:
"""Result of query decomposition"""
original: str
is_complex: bool
sub_queries: List[SubQuery]
merge_strategy: str # 'concat', 'compare', 'summarize'
def is_complex_query(query: str) -> bool:
"""
Detect if a query needs decomposition.
Complex queries contain:
- Multiple questions (and, also, then)
- Comparisons (vs, versus, compare)
- Multi-step requests (first...then, after that)
"""
query_lower = query.lower()
# Multi-part indicators
multi_part_patterns = [
r'\band\b.*\?', # "X and Y?"
r'\balso\b', # "also show"
r'\bthen\b', # "then compare"
r'\bfirst\b.*\bthen\b', # "first X then Y"
r'\b(?:as well|additionally|moreover)\b',
]
for pattern in multi_part_patterns:
if re.search(pattern, query_lower):
return True
# Multiple question marks
if query.count('?') > 1:
return True
# Long queries often need decomposition
if len(query.split()) > 25:
return True
return False
def decompose_query(query: str) -> DecomposedQuery:
"""
Decompose a complex query into simpler sub-queries.
Args:
query: The original user query
Returns:
DecomposedQuery with sub-queries and merge strategy
"""
if not is_complex_query(query):
return DecomposedQuery(
original=query,
is_complex=False,
sub_queries=[SubQuery(query=query, query_type="simple")],
merge_strategy="single"
)
sub_queries = []
merge_strategy = "concat"
# Split by common conjunctions
query_lower = query.lower()
# Handle "X and Y" patterns
if ' and ' in query_lower:
parts = re.split(r'\s+and\s+', query, flags=re.IGNORECASE)
for i, part in enumerate(parts):
part = part.strip().rstrip('?').strip()
if len(part) > 5: # Meaningful part
query_type = _detect_query_type(part)
sub_queries.append(SubQuery(
query=part + "?",
query_type=query_type,
priority=i + 1
))
# Handle comparison patterns
elif any(p in query_lower for p in ['compare', ' vs ', 'versus', 'difference between']):
merge_strategy = "compare"
# Extract entities to compare
compare_match = re.search(
r'(?:compare|difference between)\s+(.+?)\s+(?:and|vs|versus|with)\s+(.+?)(?:\?|$)',
query, re.IGNORECASE
)
if compare_match:
entity1 = compare_match.group(1).strip()
entity2 = compare_match.group(2).strip()
sub_queries.append(SubQuery(
query=f"What are the details of {entity1}?",
query_type="lookup",
priority=1
))
sub_queries.append(SubQuery(
query=f"What are the details of {entity2}?",
query_type="lookup",
priority=1
))
sub_queries.append(SubQuery(
query=f"Compare {entity1} vs {entity2}",
query_type="comparison",
priority=2,
depends_on=0 # Depends on first two
))
# Handle "Compare X vs Y over time" (Trend Comparison)
trend_match = re.search(r'(?:compare|how did)\s+(.+?)\s+(?:vs|versus|and)\s+(.+?)\s+(?:perform|change|grow|trend)\s+(?:over time|historically)', query_lower)
if trend_match:
entity1 = trend_match.group(1).strip()
entity2 = trend_match.group(2).strip()
sub_queries.append(SubQuery(
query=f"Show trend for {entity1}",
query_type="trend",
priority=1
))
sub_queries.append(SubQuery(
query=f"Show trend for {entity2}",
query_type="trend",
priority=1
))
sub_queries.append(SubQuery(
query=f"Compare trends of {entity1} vs {entity2}",
query_type="comparison_trend",
priority=2
))
merge_strategy = "compare"
# Handle "first...then" patterns
elif 'first' in query_lower and 'then' in query_lower:
merge_strategy = "sequential"
parts = re.split(r'\s*(?:,\s*)?then\s+', query, flags=re.IGNORECASE)
for i, part in enumerate(parts):
part = part.replace('first', '').strip()
if len(part) > 5:
sub_queries.append(SubQuery(
query=part,
query_type=_detect_query_type(part),
priority=i + 1,
depends_on=i - 1 if i > 0 else None
))
# Fallback: treat as single query
if not sub_queries:
sub_queries.append(SubQuery(query=query, query_type="complex"))
return DecomposedQuery(
original=query,
is_complex=True,
sub_queries=sub_queries,
merge_strategy=merge_strategy
)
def _detect_query_type(query: str) -> str:
"""Detect the type of a sub-query"""
query_lower = query.lower()
if any(w in query_lower for w in ['total', 'sum', 'count', 'how many', 'how much']):
return "aggregation"
elif any(w in query_lower for w in ['top', 'best', 'highest', 'lowest', 'worst']):
return "ranking"
elif any(w in query_lower for w in ['trend', 'over time', 'growth', 'change']):
return "trend"
elif any(w in query_lower for w in ['compare', 'vs', 'versus', 'difference']):
return "comparison"
elif any(w in query_lower for w in ['why', 'reason', 'cause', 'explain']):
return "analytical"
elif any(w in query_lower for w in ['list', 'show', 'give me', 'what are']):
return "listing"
elif any(w in query_lower for w in ['predict', 'forecast', 'will', 'future']):
return "prediction"
else:
return "factual"
def merge_results(results: List[str], strategy: str = "concat") -> str:
"""
Merge results from multiple sub-queries.
Args:
results: List of answers from sub-queries
strategy: How to merge ('concat', 'compare', 'summarize')
Returns:
Merged answer string
"""
if not results:
return "No results found."
if len(results) == 1:
return results[0]
if strategy == "compare":
# Format as comparison
merged = "## Comparison\n\n"
for i, result in enumerate(results):
merged += f"**Option {i + 1}:**\n{result}\n\n"
return merged
elif strategy == "sequential":
# Format as steps
merged = ""
for i, result in enumerate(results):
merged += f"**Step {i + 1}:**\n{result}\n\n"
return merged
else: # concat
return "\n\n---\n\n".join(results)
def get_query_intent(query: str) -> Dict[str, any]:
"""
Analyze query intent for routing decisions.
Returns:
Dict with intent, entities, and confidence
"""
query_lower = query.lower()
intent = {
'type': 'unknown',
'entities': [],
'confidence': 0.5,
'needs_data': True,
'needs_chart': False
}
# Detect chart needs
if any(w in query_lower for w in ['chart', 'graph', 'plot', 'visualize', 'show me']):
intent['needs_chart'] = True
# Detect intent type
if any(w in query_lower for w in ['hi', 'hello', 'hey']):
intent['type'] = 'greeting'
intent['needs_data'] = False
intent['confidence'] = 0.95
elif any(w in query_lower for w in ['total', 'sum', 'how much', 'how many']):
intent['type'] = 'aggregation'
intent['confidence'] = 0.85
elif any(w in query_lower for w in ['top', 'best', 'highest']):
intent['type'] = 'ranking'
intent['confidence'] = 0.85
elif any(w in query_lower for w in ['compare', 'vs', 'difference']):
intent['type'] = 'comparison'
intent['confidence'] = 0.85
elif any(w in query_lower for w in ['predict', 'forecast', 'future']):
intent['type'] = 'prediction'
intent['confidence'] = 0.8
elif any(w in query_lower for w in ['list', 'show all', 'give me all']):
intent['type'] = 'listing'
intent['confidence'] = 0.8
else:
intent['type'] = 'general'
intent['confidence'] = 0.6
# Extract potential entities (capitalized words)
entities = re.findall(r'\b([A-Z][a-z]+(?:\s+[A-Z][a-z]+)*)\b', query)
intent['entities'] = [e for e in entities if e.lower() not in
['what', 'who', 'when', 'where', 'how', 'why', 'show', 'list', 'give']]
return intent
|