File size: 13,740 Bytes
c089ca4 | 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 | #!/usr/bin/env python3
"""
Text Processor
Handles text cleaning, chunking, and preprocessing for the NZ Legislation Loophole Analysis.
Optimized for legal/legislative content with specialized cleaning and structuring.
"""
import re
from typing import List, Dict, Any, Optional, Tuple
import hashlib
import json
class TextProcessor:
"""Advanced text processing for legislation analysis"""
def __init__(self):
"""Initialize the text processor with legal-specific patterns"""
# Legal-specific patterns
self.section_pattern = re.compile(r'^(\d+):', re.MULTILINE)
self.act_name_pattern = re.compile(r'(\b\w+(?:\s+\w+)*)\s+Act\s+(\d{4})', re.IGNORECASE)
self.date_patterns = [
(r'(\d{1,2})\s*(?:January|February|March|April|May|June|July|August|September|October|November|December)\s*(\d{4})',
lambda m: f"{m.group(1)} {m.group(2)}"),
(r'(\d{1,2})/(\d{1,2})/(\d{4})', r'\1/\2/\3'),
(r'(\d{4})-(\d{2})-(\d{2})', r'\1-\2-\3')
]
# NZ-specific legal terms
self.nz_terms = {
'New Zealand': 'New Zealand',
'Parliament': 'Parliament',
'Crown': 'Crown',
'Government': 'Government',
'Treaty of Waitangi': 'Treaty of Waitangi',
'NZB': 'NZB',
'Her Majesty': 'Her Majesty',
'Governor-General': 'Governor-General'
}
def clean_text(self, text: str, preserve_structure: bool = True) -> str:
"""
Clean and normalize text for better processing, optimized for legal content
Args:
text: Raw text to clean
preserve_structure: Whether to preserve legal document structure
Returns:
Cleaned text
"""
if not text:
return ""
# Preserve section numbers and legal structure if requested
if preserve_structure:
# Keep section numbers like "1:", "2:", etc.
text = self.section_pattern.sub(r'\1', text)
# Remove excessive whitespace but preserve paragraph structure
text = re.sub(r'[ \t]+', ' ', text) # Replace multiple spaces/tabs with single space
text = re.sub(r'\n\s*\n', '\n\n', text) # Preserve paragraph breaks but clean up
text = re.sub(r'\n{3,}', '\n\n', text) # Reduce excessive newlines to double
# Remove control characters but preserve legal formatting
text = re.sub(r'[\x00-\x08\x0b\x0c\x0e-\x1f\x7f-\x9f]', '', text)
# Handle legal-specific characters and formatting
allowed_chars = r'\w\s\.\,\!\?\;\:\-\(\)\[\]\{\}\"\'\/\@\#\$\%\^\&\*\+\=\~\`°§'
text = re.sub(r'[^' + allowed_chars + ']', '', text)
# Normalize quotes and apostrophes for legal text
text = re.sub(r'[""]', '"', text)
text = re.sub(r"['']", "'", text)
text = re.sub(r'`', "'", text)
# Clean up legal numbering and references
text = re.sub(r'section\s+(\d+)', r'section \1', text, flags=re.IGNORECASE)
# Normalize date formats
for pattern, replacement in self.date_patterns:
if callable(replacement):
text = re.compile(pattern, re.IGNORECASE).sub(replacement, text)
else:
text = re.compile(pattern, re.IGNORECASE).sub(replacement, text)
# Normalize act names with years
text = self.act_name_pattern.sub(r'\1 Act', text)
# Clean up amendment references
text = re.sub(r'[Aa]mendment\(s\)\s+incorporated\s+in\s+the\s+[Aa]ct\(s\)', 'Amendments incorporated', text)
# Normalize section references
text = re.sub(r'section\s+\d+\(\d+\)\([a-zA-Z]\)', lambda m: m.group(0).lower(), text)
# Generic pattern for legal document sections
text = re.sub(r'(\b(?:section|part|chapter|article|clause|subsection|paragraph))\s+(\d+[a-zA-Z]*)',
lambda m: f"{m.group(1)} {m.group(2)}", text, flags=re.IGNORECASE)
# NZ-specific legal enhancements
for term, normalized in self.nz_terms.items():
text = re.sub(re.escape(term), normalized, text, flags=re.IGNORECASE)
# Handle Maori-specific characters if present
maori_chars = 'āēīōūwhĀĒĪŌŪWH'
allowed_chars += maori_chars
text = re.sub(r'[^' + allowed_chars + ']', '', text)
# Remove empty lines and trim while preserving legal structure
lines = []
for line in text.split('\n'):
stripped = line.strip()
if stripped: # Keep non-empty lines
if preserve_structure and re.match(r'^\d+:', stripped):
lines.append(stripped) # Preserve section headers
else:
lines.append(stripped)
text = '\n'.join(lines)
return text.strip()
def chunk_text(self, text: str, chunk_size: int = 4096, overlap: int = 256,
method: str = "sentence") -> List[str]:
"""
Split text into overlapping chunks for processing
Args:
text: Text to chunk
chunk_size: Size of each chunk
overlap: Overlap between chunks
method: Chunking method ('sentence', 'word', 'character')
Returns:
List of text chunks
"""
if not text or len(text) <= chunk_size:
return [text] if text else []
chunks = []
if method == "sentence":
chunks = self._chunk_by_sentence(text, chunk_size, overlap)
elif method == "word":
chunks = self._chunk_by_word(text, chunk_size, overlap)
else: # character
chunks = self._chunk_by_character(text, chunk_size, overlap)
return chunks
def _chunk_by_sentence(self, text: str, chunk_size: int, overlap: int) -> List[str]:
"""Chunk text by sentence boundaries"""
# Split into sentences (rough approximation)
sentence_pattern = r'(?<=[.!?])\s+'
sentences = re.split(sentence_pattern, text)
chunks = []
current_chunk = ""
overlap_text = ""
for sentence in sentences:
if not sentence.strip():
continue
# Check if adding this sentence would exceed chunk size
potential_chunk = current_chunk + sentence + " "
if len(potential_chunk) > chunk_size and current_chunk:
# Save current chunk
chunks.append(current_chunk.strip())
# Start new chunk with overlap
if overlap > 0 and len(current_chunk) > overlap:
overlap_text = current_chunk[-overlap:].strip()
current_chunk = overlap_text + " " + sentence + " "
else:
current_chunk = sentence + " "
else:
current_chunk = potential_chunk
# Add the last chunk
if current_chunk.strip():
chunks.append(current_chunk.strip())
return chunks
def _chunk_by_word(self, text: str, chunk_size: int, overlap: int) -> List[str]:
"""Chunk text by word boundaries"""
words = text.split()
chunks = []
if not words:
return []
start = 0
while start < len(words):
end = start + 1
chunk_words = []
# Build chunk up to chunk_size
while end <= len(words):
potential_chunk = " ".join(words[start:end])
if len(potential_chunk) > chunk_size:
break
chunk_words = words[start:end]
end += 1
if chunk_words:
chunk = " ".join(chunk_words)
chunks.append(chunk)
# Move start position with overlap
overlap_words = max(0, min(overlap // 5, len(chunk_words))) # Rough word overlap
start = max(start + 1, end - overlap_words)
else:
break
return chunks
def _chunk_by_character(self, text: str, chunk_size: int, overlap: int) -> List[str]:
"""Chunk text by character count (simple fallback)"""
chunks = []
start = 0
while start < len(text):
end = min(start + chunk_size, len(text))
chunk = text[start:end]
chunks.append(chunk)
# Move start with overlap
start = end - overlap if end < len(text) else len(text)
return chunks
def extract_metadata(self, text: str) -> Dict[str, Any]:
"""Extract metadata from legislation text"""
metadata = {
'sections': [],
'acts_referenced': [],
'dates': [],
'word_count': len(text.split()),
'character_count': len(text),
'has_nz_references': False,
'has_maori_terms': False
}
# Extract section numbers
sections = self.section_pattern.findall(text)
metadata['sections'] = [int(s) for s in sections]
# Extract referenced acts
acts = self.act_name_pattern.findall(text)
metadata['acts_referenced'] = [f"{act[0]} Act" for act in acts]
# Check for NZ-specific references
nz_indicators = ['New Zealand', 'Parliament', 'Crown', 'Government', 'Treaty of Waitangi']
metadata['has_nz_references'] = any(term in text for term in nz_indicators)
# Check for Maori terms
maori_indicators = ['ā', 'ē', 'ī', 'ō', 'ū', 'whakapapa', 'tangata whenua', 'mana']
metadata['has_maori_terms'] = any(term in text.lower() for term in maori_indicators)
# Extract dates (basic)
date_pattern = r'\b\d{1,2}[/-]\d{1,2}[/-]\d{4}\b'
dates = re.findall(date_pattern, text)
metadata['dates'] = dates
return metadata
def calculate_text_hash(self, text: str) -> str:
"""Calculate SHA-256 hash of text for caching"""
return hashlib.sha256(text.encode('utf-8')).hexdigest()
def get_chunk_statistics(self, chunks: List[str]) -> Dict[str, Any]:
"""Get statistics about text chunks"""
if not chunks:
return {
'total_chunks': 0,
'avg_chunk_size': 0,
'min_chunk_size': 0,
'max_chunk_size': 0,
'total_characters': 0
}
chunk_sizes = [len(chunk) for chunk in chunks]
return {
'total_chunks': len(chunks),
'avg_chunk_size': sum(chunk_sizes) / len(chunks),
'min_chunk_size': min(chunk_sizes),
'max_chunk_size': max(chunk_sizes),
'total_characters': sum(chunk_sizes)
}
def preprocess_legislation_json(self, json_data: Dict[str, Any]) -> Dict[str, Any]:
"""Preprocess legislation data from JSON format"""
processed = {
'id': json_data.get('id', ''),
'title': json_data.get('title', ''),
'year': json_data.get('year', ''),
'source': json_data.get('source', ''),
'original_text': json_data.get('text', ''),
'cleaned_text': '',
'chunks': [],
'metadata': {},
'processing_stats': {}
}
# Clean the text
raw_text = json_data.get('text', '')
processed['cleaned_text'] = self.clean_text(raw_text)
# Extract metadata
processed['metadata'] = self.extract_metadata(processed['cleaned_text'])
return processed
def batch_process_texts(self, texts: List[str], chunk_size: int = 4096,
overlap: int = 256) -> List[Dict[str, Any]]:
"""Process multiple texts in batch"""
results = []
for text in texts:
cleaned = self.clean_text(text)
chunks = self.chunk_text(cleaned, chunk_size, overlap)
metadata = self.extract_metadata(cleaned)
stats = self.get_chunk_statistics(chunks)
result = {
'original_text': text,
'cleaned_text': cleaned,
'chunks': chunks,
'metadata': metadata,
'processing_stats': stats
}
results.append(result)
return results
def validate_text_quality(self, text: str) -> Dict[str, Any]:
"""Validate and assess text quality for processing"""
quality = {
'is_valid': True,
'issues': [],
'score': 100,
'metrics': {}
}
# Check minimum length
if len(text.strip()) < 10:
quality['issues'].append("Text too short")
quality['score'] -= 50
# Check for excessive special characters
special_chars = len(re.findall(r'[^\w\s]', text))
special_ratio = special_chars / len(text) if text else 0
if special_ratio > 0.3:
quality['issues'].append("High special character ratio")
quality['score'] -= 20
# Check for legal content indicators
legal_indicators = ['section', 'act', 'law', 'regulation', 'clause', 'subsection']
has_legal_content = any(indicator in text.lower() for indicator in legal_indicators)
if not has_legal_content:
quality['issues'].append("May not be legal content")
quality['score'] -= 30
quality['is_valid'] = len(quality['issues']) == 0
quality['metrics'] = {
'length': len(text),
'word_count': len(text.split()),
'special_char_ratio': special_ratio,
'has_legal_content': has_legal_content
}
return quality
|