Spaces:
Running
Running
File size: 16,558 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 |
# DEPENDENCIES
from typing import List
from typing import Tuple
from typing import Optional
from config.models import DocumentChunk
from config.logging_config import get_logger
from chunking.token_counter import TokenCounter
# Setup Logging
logger = get_logger(__name__)
class OverlapManager:
"""
Manages overlapping regions between chunks : ensures smooth context transitions and optimal retrieval
"""
def __init__(self, overlap_tokens: int = 50):
"""
Initialize overlap manager
Arguments:
----------
overlap_tokens { int } : Target overlap in tokens
"""
self.overlap_tokens = overlap_tokens
self.token_counter = TokenCounter()
self.logger = logger
def add_overlap(self, chunks: List[DocumentChunk], overlap_tokens: Optional[int] = None) -> List[DocumentChunk]:
"""
Add overlap to existing chunks
Arguments:
----------
chunks { list } : List of chunks without overlap
overlap_tokens { int } : Override default overlap
Returns:
--------
{ list } : List of chunks with overlap
"""
if (not chunks or (len(chunks) < 2)):
return chunks
overlap = overlap_tokens or self.overlap_tokens
overlapped_chunks = list()
for i, chunk in enumerate(chunks):
if (i == 0):
# First chunk: no prefix, add suffix from next
new_text = chunk.text
if (i + 1 < len(chunks)):
suffix = self._get_overlap_text(text = chunks[i + 1].text,
overlap_tokens = overlap,
from_start = True,
)
new_text = new_text + " " + suffix
elif (i == len(chunks) - 1):
# Last chunk: add prefix from previous, no suffix
prefix = self._get_overlap_text(text = chunks[i - 1].text,
overlap_tokens = overlap,
from_start = False,
)
new_text = prefix + " " + chunk.text
else:
# Middle chunk: add both prefix and suffix
prefix = self._get_overlap_text(text = chunks[i - 1].text,
overlap_tokens = overlap,
from_start = False,
)
suffix = self._get_overlap_text(text = chunks[i + 1].text,
overlap_tokens = overlap,
from_start = True,
)
new_text = prefix + " " + chunk.text + " " + suffix
# Create new chunk with overlapped text
overlapped_chunk = DocumentChunk(chunk_id = chunk.chunk_id,
document_id = chunk.document_id,
text = new_text,
chunk_index = chunk.chunk_index,
start_char = chunk.start_char,
end_char = chunk.end_char,
page_number = chunk.page_number,
section_title = chunk.section_title,
token_count = self.token_counter.count_tokens(new_text),
metadata = chunk.metadata,
)
overlapped_chunks.append(overlapped_chunk)
self.logger.debug(f"Added overlap to {len(chunks)} chunks")
return overlapped_chunks
def _get_overlap_text(self, text: str, overlap_tokens: int, from_start: bool) -> str:
"""
Extract overlap text from beginning or end
Arguments:
----------
text { str } : Source text
overlap_tokens { int } : Number of tokens to extract
from_start { bool } : True for start, False for end
Returns:
--------
{ str } : Overlap text
"""
total_tokens = self.token_counter.count_tokens(text)
if (total_tokens <= overlap_tokens):
return text
if from_start:
# Get first N tokens
return self.token_counter.truncate_to_tokens(text = text,
max_tokens = overlap_tokens,
suffix = "",
)
else:
# Get last N tokens using token counter's boundary finding
char_pos, overlap_text = self.token_counter.find_token_boundaries(text = text,
target_tokens = overlap_tokens,
)
# Take from the end instead of beginning
if (char_pos < len(text)):
return text[-char_pos:] if (char_pos > 0) else text
return overlap_text
def remove_overlap(self, chunks: List[DocumentChunk]) -> List[DocumentChunk]:
"""
Remove overlap from chunks (get core content only)
Arguments:
----------
chunks { list } : List of chunks with overlap
Returns:
--------
{ list } : List of chunks without overlap
"""
if (not chunks or (len(chunks) < 2)):
return chunks
core_chunks = list()
for i, chunk in enumerate(chunks):
if (i == 0):
# First chunk: remove suffix
core_text = self._remove_suffix_overlap(text = chunk.text,
next_text = chunks[i + 1].text if i + 1 < len(chunks) else "",
)
elif (i == len(chunks) - 1):
# Last chunk: remove prefix
core_text = self._remove_prefix_overlap(text = chunk.text,
previous_text = chunks[i - 1].text,
)
else:
# Middle chunk: remove both
temp_text = self._remove_prefix_overlap(text = chunk.text,
previous_text = chunks[i - 1].text,
)
core_text = self._remove_suffix_overlap(text = temp_text,
next_text = chunks[i + 1].text,
)
core_chunk = DocumentChunk(chunk_id = chunk.chunk_id,
document_id = chunk.document_id,
text = core_text,
chunk_index = chunk.chunk_index,
start_char = chunk.start_char,
end_char = chunk.end_char,
page_number = chunk.page_number,
section_title = chunk.section_title,
token_count = self.token_counter.count_tokens(core_text),
metadata = chunk.metadata,
)
core_chunks.append(core_chunk)
return core_chunks
def _remove_prefix_overlap(self, text: str, previous_text: str) -> str:
"""
Remove overlap with previous chunk
"""
if not text or not previous_text:
return text
words = text.split()
prev_words = previous_text.split()
# Find longest common suffix-prefix match
max_overlap = 0
for overlap_size in range(1, min(len(words), len(prev_words)) + 1):
if (words[:overlap_size] == prev_words[-overlap_size:]):
max_overlap = overlap_size
if (max_overlap > 0):
return " ".join(words[max_overlap:])
return text
def _remove_suffix_overlap(self, text: str, next_text: str) -> str:
"""
Remove overlap with next chunk
"""
# Find common suffix
words = text.split()
next_words = next_text.split()
common_length = 0
for i in range(1, min(len(words), len(next_words)) + 1):
if (words[-i] == next_words[i - 1]):
common_length += 1
else:
break
if (common_length > 0):
return " ".join(words[:-common_length])
return text
def calculate_overlap_percentage(self, chunks: List[DocumentChunk]) -> float:
"""
Calculate average overlap percentage
Arguments:
----------
chunks { list } : List of chunks
Returns:
--------
{ float } : Average overlap percentage
"""
if (len(chunks) < 2):
return 0.0
overlaps = list()
for i in range(len(chunks) - 1):
overlap = self._measure_overlap(chunks[i].text, chunks[i + 1].text)
overlaps.append(overlap)
return sum(overlaps) / len(overlaps) if overlaps else 0.0
def _measure_overlap(self, text1: str, text2: str) -> float:
"""
Measure overlap between two texts
Arguments:
----------
text1 { str } : First text
text2 { str } : Second text
Returns:
--------
{ float } : Overlap percentage (0-100)
"""
words1 = set(text1.lower().split())
words2 = set(text2.lower().split())
if (not words1 or not words2):
return 0.0
common = words1 & words2
overlap_pct = (len(common) / min(len(words1), len(words2))) * 100
return overlap_pct
def optimize_overlaps(self, chunks: List[DocumentChunk], target_overlap: int, tolerance: int = 10) -> List[DocumentChunk]:
"""
Optimize overlap sizes to target
Arguments:
----------
chunks { list } : List of chunks
target_overlap { int } : Target overlap in tokens
tolerance { int } : Acceptable deviation in tokens
Returns:
--------
{ list } : Optimized chunks
"""
if (len(chunks) < 2):
return chunks
# Validate target_overlap is reasonable
if (target_overlap <= 0):
self.logger.warning("Target overlap must be positive, using default")
target_overlap = self.overlap_tokens
optimized = list()
for i in range(len(chunks)):
chunk = chunks[i]
# Check current overlap with next chunk
if (i < len(chunks) - 1):
current_overlap = self._count_overlap_tokens(text1 = chunk.text,
text2 = chunks[i + 1].text,
)
# Adjust if outside tolerance
if (abs(current_overlap - target_overlap) > tolerance):
# Add or remove text to reach target
if (current_overlap < target_overlap):
# Need more overlap
additional = self._get_overlap_text(text = chunks[i + 1].text,
overlap_tokens = target_overlap - current_overlap,
from_start = True,
)
new_text = chunk.text + " " + additional
else:
# Need less overlap
new_text = self.token_counter.truncate_to_tokens(text = chunk.text,
max_tokens = self.token_counter.count_tokens(chunk.text) - (current_overlap - target_overlap),
)
chunk = DocumentChunk(chunk_id = chunk.chunk_id,
document_id = chunk.document_id,
text = new_text,
chunk_index = chunk.chunk_index,
start_char = chunk.start_char,
end_char = chunk.end_char,
page_number = chunk.page_number,
section_title = chunk.section_title,
token_count = self.token_counter.count_tokens(new_text),
metadata = chunk.metadata,
)
optimized.append(chunk)
return optimized
def _count_overlap_tokens(self, text1: str, text2: str) -> int:
"""
Count overlapping tokens between two texts
"""
# Find longest common substring at the boundary
words1 = text1.split()
words2 = text2.split()
max_overlap = 0
for i in range(1, min(len(words1), len(words2)) + 1):
if (words1[-i:] == words2[:i]):
overlap_text = " ".join(words1[-i:])
max_overlap = self.token_counter.count_tokens(overlap_text)
return max_overlap
def get_overlap_statistics(self, chunks: List[DocumentChunk]) -> dict:
"""
Get statistics about overlaps
Arguments:
----------
chunks { list } : List of chunks
Returns:
--------
{ dict } : Statistics dictionary
"""
if (len(chunks) < 2):
return {"num_chunks" : len(chunks),
"num_overlaps" : 0,
"avg_overlap_tokens" : 0,
"avg_overlap_percentage" : 0,
}
overlap_tokens = list()
overlap_percentages = list()
for i in range(len(chunks) - 1):
tokens = self._count_overlap_tokens(chunks[i].text, chunks[i + 1].text)
pct = self._measure_overlap(chunks[i].text, chunks[i + 1].text)
overlap_tokens.append(tokens)
overlap_percentages.append(pct)
return {"num_chunks" : len(chunks),
"num_overlaps" : len(overlap_tokens),
"avg_overlap_tokens" : sum(overlap_tokens) / len(overlap_tokens) if overlap_tokens else 0,
"min_overlap_tokens" : min(overlap_tokens) if overlap_tokens else 0,
"max_overlap_tokens" : max(overlap_tokens) if overlap_tokens else 0,
"avg_overlap_percentage" : sum(overlap_percentages) / len(overlap_percentages) if overlap_percentages else 0,
} |