Spaces:
Running
Running
File size: 15,102 Bytes
0a4529c | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 | # DEPENDENCIES
import re
from typing import List
from typing import Dict
from typing import Tuple
from typing import Optional
from collections import defaultdict
from config.models import DocumentChunk
from config.models import ChunkWithScore
from config.logging_config import get_logger
from utils.error_handler import CitationError
from utils.error_handler import handle_errors
# Setup Logging
logger = get_logger(__name__)
class CitationTracker:
"""
Citation tracking and management: Tracks source citations in generated text and provides citation formatting and validation
"""
def __init__(self):
"""
Initialize citation tracker
"""
self.logger = logger
self.citation_pattern = re.compile(r'\[(\d+)\]')
def extract_citations(self, text: str) -> List[int]:
"""
Extract citation numbers from text
Arguments:
----------
text { str } : Text containing citations
Returns:
--------
{ list } : List of citation numbers found in text
"""
if not text:
return []
try:
matches = self.citation_pattern.findall(text)
citation_numbers = [int(match) for match in matches]
# Remove duplicates and sort
unique_citations = sorted(set(citation_numbers))
self.logger.debug(f"Extracted {len(unique_citations)} citations from text")
return unique_citations
except Exception as e:
self.logger.error(f"Citation extraction failed: {repr(e)}")
return []
def validate_citations(self, text: str, sources: List[ChunkWithScore]) -> Tuple[bool, List[int]]:
"""
Validate that all citations in text reference existing sources
Arguments:
----------
text { str } : Text containing citations
sources { list } : List of available sources
Returns:
--------
{ Tuple[bool, List[int]] } : (is_valid, invalid_citations)
"""
citation_numbers = self.extract_citations(text = text)
if not citation_numbers:
return True, []
# Check if all citation numbers are within valid range
max_valid = len(sources)
invalid_citations = [num for num in citation_numbers if (num < 1) or (num > max_valid)]
if invalid_citations:
self.logger.warning(f"Invalid citations found: {invalid_citations}. Valid range: 1-{max_valid}")
return False, invalid_citations
return True, []
def format_citations(self, sources: List[ChunkWithScore], style: str = "numeric") -> str:
"""
Format citations as reference list
Arguments:
----------
sources { list } : List of sources to format
style { str } : Citation style ('numeric', 'verbose')
Returns:
--------
{ str } : Formatted citation text
"""
if not sources:
return ""
try:
citations = list()
for i, source in enumerate(sources, 1):
if (style == "verbose"):
citation = self._format_verbose_citation(source = source,
number = i,
)
else:
citation = self._format_numeric_citation(source = source,
number = i,
)
citations.append(citation)
citation_text = "\n".join(citations)
self.logger.debug(f"Formatted {len(citations)} citations in {style} style")
return citation_text
except Exception as e:
self.logger.error(f"Citation formatting failed: {repr(e)}")
return ""
def _format_numeric_citation(self, source: ChunkWithScore, number: int) -> str:
"""
Format citation in numeric style with sanitization
Arguments:
----------
source { ChunkWithScore } : Source to format
number { int } : Citation number
Returns:
--------
{ str } : Formatted citation
"""
chunk = source.chunk
parts = [f"[{number}]"]
# Add source information with proper sanitization
if (hasattr(chunk, 'metadata') and chunk.metadata):
if ('filename' in chunk.metadata):
# Sanitize filename more thoroughly
filename = str(chunk.metadata['filename'])
# Remove problematic characters that could break citation parsing: Keep only alphanumeric, spaces, dots, hyphens, underscores
filename = re.sub(r'[^\w\s\.\-]', '_', filename)
# Limit length to prevent overflow
if (len(filename) > 50):
filename = filename[:47] + "..."
parts.append(f"Source: {filename}")
if chunk.page_number:
parts.append(f"Page {chunk.page_number}")
if chunk.section_title:
# Sanitize section title similarly
section = str(chunk.section_title)
section = re.sub(r'[^\w\s\.\-]', '_', section)
if (len(section) > 40):
section = section[:37] + "..."
parts.append(f"Section: {section}")
# Add relevance score if available
if (source.score > 0):
parts.append(f"(Relevance: {source.score:.2f})")
return " ".join(parts)
def _format_verbose_citation(self, source: ChunkWithScore, number: int) -> str:
"""
Format citation in verbose style - SAFER VERSION
Arguments:
----------
source { ChunkWithScore } : Source to format
number { int } : Citation number
Returns:
--------
{ str } : Formatted citation
"""
chunk = source.chunk
parts = [f"Citation {number}:"]
# Document information with sanitization
if (hasattr(chunk, 'metadata')):
meta = chunk.metadata
if ('filename' in meta):
filename = str(meta['filename'])
filename = re.sub(r'[^\w\s\.\-]', '_', filename)
if (len(filename) > 50):
filename = filename[:47] + "..."
parts.append(f"Document: {filename}")
if ('title' in meta):
title = str(meta['title'])
title = re.sub(r'[^\w\s\.\-]', '_', title)
if (len(title) > 60):
title = title[:57] + "..."
parts.append(f"Title: {title}")
if ('author' in meta):
author = str(meta['author'])
author = re.sub(r'[^\w\s\.\-]', '_', author)
if (len(author) > 40):
author = author[:37] + "..."
parts.append(f"Author: {author}")
# Location information
location_parts = list()
if chunk.page_number:
location_parts.append(f"page {chunk.page_number}")
if chunk.section_title:
section = str(chunk.section_title)
section = re.sub(r'[^\w\s\.\-]', '_', section)
if (len(section) > 40):
section = section[:37] + "..."
location_parts.append(f"section '{section}'")
if location_parts:
parts.append("(" + ", ".join(location_parts) + ")")
# Relevance information
if (source.score > 0):
parts.append(f"[Relevance score: {source.score:.3f}]")
return " ".join(parts)
def generate_citation_map(self, sources: List[ChunkWithScore]) -> Dict[int, Dict]:
"""
Generate mapping from citation numbers to source details
Arguments:
----------
sources { list } : List of sources
Returns:
--------
{ dict } : Dictionary mapping citation numbers to source details
"""
citation_map = dict()
for i, source in enumerate(sources, 1):
chunk = source.chunk
citation_map[i] = {'chunk_id' : chunk.chunk_id,
'document_id' : chunk.document_id,
'score' : source.score,
'text_preview' : chunk.text[:200] + "..." if (len(chunk.text) > 200) else chunk.text,
'metadata' : getattr(chunk, 'metadata', {}),
'page_number' : chunk.page_number,
'section_title' : chunk.section_title,
}
return citation_map
def replace_citation_markers(self, text: str, citation_map: Dict[int, str]) -> str:
"""
Replace citation markers with formatted citations - FIXED
Arguments:
----------
text { str } : Text containing citation markers
citation_map { dict } : Mapping of citation numbers to formatted strings
Returns:
--------
{ str } : Text with replaced citations
"""
def replacement(match):
try:
citation_num = int(match.group(1))
# Get replacement text and sanitize it
replacement_text = citation_map.get(citation_num, match.group(0))
return str(replacement_text)
except (ValueError, IndexError):
# Return original match if parsing fails
return match.group(0)
try:
return self.citation_pattern.sub(replacement, text)
except Exception as e:
self.logger.error(f"Citation replacement failed: {repr(e)}")
# Return original text on error
return text
def get_citation_statistics(self, text: str, sources: List[ChunkWithScore]) -> Dict:
"""
Get statistics about citations in text
Arguments:
----------
text { str } : Text containing citations
sources { list } : List of sources
Returns:
--------
{ dict } : Citation statistics
"""
citation_numbers = self.extract_citations(text = text)
if not citation_numbers:
return {"total_citations": 0}
# Calculate citation distribution
citation_counts = defaultdict(int)
for num in citation_numbers:
if 1 <= num <= len(sources):
source = sources[num - 1]
doc_id = source.chunk.document_id
citation_counts[doc_id] += 1
return {"total_citations" : len(citation_numbers),
"unique_citations" : len(set(citation_numbers)),
"citation_distribution": dict(citation_counts),
"citations_per_source" : {i: citation_numbers.count(i) for i in set(citation_numbers)},
}
def ensure_citation_consistency(self, text: str, sources: List[ChunkWithScore]) -> str:
"""
Ensure citation numbers are consistent and sequential
Arguments:
----------
text { str } : Text containing citations
sources { list } : List of sources
Returns:
--------
{ str } : Text with consistent citations
"""
is_valid, invalid_citations = self.validate_citations(text, sources)
if not is_valid:
self.logger.warning("Invalid citations found, attempting to fix consistency")
# Extract current citations and create mapping
current_citations = self.extract_citations(text = text)
if not current_citations:
return text
# Create mapping from old to new citation numbers
citation_mapping = dict()
for i, old_num in enumerate(sorted(set(current_citations)), 1):
if (old_num <= len(sources)):
citation_mapping[old_num] = i
# Replace citations in text
def consistent_replacement(match):
old_num = int(match.group(1))
new_num = citation_mapping.get(old_num, old_num)
return f"[{new_num}]"
fixed_text = self.citation_pattern.sub(consistent_replacement, text)
self.logger.info(f"Fixed citation consistency: {current_citations} -> {list(citation_mapping.values())}")
return fixed_text
return text
# Global citation tracker instance
_citation_tracker = None
def get_citation_tracker() -> CitationTracker:
"""
Get global citation tracker instance (singleton)
Returns:
--------
{ CitationTracker } : CitationTracker instance
"""
global _citation_tracker
if _citation_tracker is None:
_citation_tracker = CitationTracker()
return _citation_tracker
@handle_errors(error_type = CitationError, log_error = True, reraise = False)
def extract_and_validate_citations(text: str, sources: List[ChunkWithScore]) -> Tuple[List[int], bool]:
"""
Convenience function for citation extraction and validation
Arguments:
----------
text { str } : Text containing citations
sources { list } : List of sources
Returns:
--------
{ Tuple[List[int], bool] } : (citation_numbers, is_valid)
"""
tracker = get_citation_tracker()
citations = tracker.extract_citations(text = text)
is_valid, _ = tracker.validate_citations(text = text,
sources = sources,
)
return citations, is_valid |