File size: 17,536 Bytes
f74b782 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 | """
Workplace-Pulse-Telemetry: Local-First Multi-Format Document Parser
Engineered by Fabio Torres (neurodeveloper11)
Extracts clean, structured text and metadata from PDF, DOCX, XLSX, CSV, EML and TXT
without any cloud or heavy GPU dependencies.
"""
import os
import re
import csv
import email
from email import policy
from pathlib import Path
from datetime import datetime
from typing import List, Optional, Dict, Any, Tuple
from pydantic import BaseModel, Field
import pandas as pd
from pypdf import PdfReader
import docx
from src.schemas import ScannedDocumentItem
class ExtractedSection(BaseModel):
"""Structured segment extracted from a document before discursive chunking."""
section_title: Optional[str] = None
speaker_or_author: Optional[str] = None
timestamp: Optional[datetime] = None
content: str
section_type: str = "paragraph" # 'heading', 'paragraph', 'table_row', 'email_body', 'tabular_record'
class ExtractedDocument(BaseModel):
"""Complete structured extraction result of an ingested document."""
file_path: str
file_name: str
file_extension: str
department: str
detected_date: Optional[datetime] = None
author: Optional[str] = None
title: Optional[str] = None
sections: List[ExtractedSection] = Field(default_factory=list)
raw_full_text: str = ""
class LocalDocumentParser:
"""
On-premise zero-leakage document parser supporting:
- PDF (.pdf) via pypdf
- Word (.docx) via python-docx
- Excel & CSV (.xlsx, .csv) via pandas + openpyxl
- Emails & Memos (.eml, .txt) via standard email & io
"""
NARRATIVE_COLUMN_KEYWORDS = [
r"observaci[o贸]n", r"descripci[o贸]n", r"motivo", r"comentario",
r"justificaci[o贸]n", r"hecho", r"incidente", r"queja", r"descargo",
r"bit[a谩]cora", r"nota", r"detalle", r"asunto", r"narrativa",
r"situaci[o贸]n", r"reporte", r"comunicado", r"mensaje", r"intervenci[o贸]n"
]
def __init__(self):
self._narrative_col_regex = re.compile(
"|".join(self.NARRATIVE_COLUMN_KEYWORDS), re.IGNORECASE
)
def parse_document(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Parses a document according to its file extension."""
ext = item.file_extension.lower()
path = item.file_path
if ext == ".pdf":
return self._parse_pdf(item)
elif ext == ".docx":
return self._parse_docx(item)
elif ext in [".xlsx", ".xls"]:
return self._parse_excel(item)
elif ext == ".csv":
return self._parse_csv(item)
elif ext == ".eml":
return self._parse_eml(item)
elif ext == ".txt":
return self._parse_txt(item)
else:
raise ValueError(f"Unsupported document format: {ext}")
def _parse_pdf(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Extracts text page-by-page from PDF files."""
reader = PdfReader(item.file_path)
sections: List[ExtractedSection] = []
full_text_parts: List[str] = []
# Read document metadata
doc_title = None
doc_author = None
doc_date = None
if reader.metadata:
doc_title = reader.metadata.title
doc_author = reader.metadata.author
creation_date_raw = reader.metadata.get("/CreationDate")
if creation_date_raw and isinstance(creation_date_raw, str):
# PDF date format: D:YYYYMMDDHHmmSS
m = re.search(r"D:(\d{4})(\d{2})(\d{2})", creation_date_raw)
if m:
try:
doc_date = datetime(int(m.group(1)), int(m.group(2)), int(m.group(3)))
except Exception:
pass
for page_idx, page in enumerate(reader.pages):
page_text = page.extract_text() or ""
clean_page_text = page_text.strip()
if clean_page_text:
full_text_parts.append(clean_page_text)
sections.append(
ExtractedSection(
section_title=f"P谩gina {page_idx + 1}",
speaker_or_author=doc_author,
timestamp=doc_date,
content=clean_page_text,
section_type="paragraph",
)
)
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=doc_date,
author=doc_author,
title=doc_title or item.file_name,
sections=sections,
raw_full_text="\n\n".join(full_text_parts),
)
def _parse_docx(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Extracts paragraphs, headers, and tables from Word (.docx) documents."""
doc = docx.Document(item.file_path)
sections: List[ExtractedSection] = []
full_text_parts: List[str] = []
doc_author = doc.core_properties.author or None
doc_date = doc.core_properties.created or None
doc_title = doc.core_properties.title or None
current_heading = "Introducci贸n"
# 1. Process paragraphs
for para in doc.paragraphs:
text = para.text.strip()
if not text:
continue
style_name = (para.style.name or "").lower()
if "heading" in style_name or "t铆tulo" in style_name:
current_heading = text
sections.append(
ExtractedSection(
section_title=current_heading,
speaker_or_author=doc_author,
timestamp=doc_date,
content=text,
section_type="heading",
)
)
else:
sections.append(
ExtractedSection(
section_title=current_heading,
speaker_or_author=doc_author,
timestamp=doc_date,
content=text,
section_type="paragraph",
)
)
full_text_parts.append(text)
# 2. Process tables
for table_idx, table in enumerate(doc.tables):
headers = []
for row_idx, row in enumerate(table.rows):
row_cells = [cell.text.strip() for cell in row.cells]
if row_idx == 0:
headers = row_cells
else:
row_pairs = []
for h_idx, cell_val in enumerate(row_cells):
col_name = headers[h_idx] if h_idx < len(headers) and headers[h_idx] else f"Col_{h_idx+1}"
if cell_val:
row_pairs.append(f"{col_name}: {cell_val}")
if row_pairs:
row_summary = " | ".join(row_pairs)
sections.append(
ExtractedSection(
section_title=f"Tabla {table_idx + 1}",
speaker_or_author=doc_author,
timestamp=doc_date,
content=row_summary,
section_type="table_row",
)
)
full_text_parts.append(row_summary)
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=doc_date,
author=doc_author,
title=doc_title or item.file_name,
sections=sections,
raw_full_text="\n\n".join(full_text_parts),
)
def _parse_excel(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Extracts free-text columns from Excel workbooks (.xlsx)."""
sections: List[ExtractedSection] = []
full_text_parts: List[str] = []
excel_file = pd.ExcelFile(item.file_path, engine="openpyxl")
sheet_names = excel_file.sheet_names
for sheet in sheet_names:
df = pd.read_excel(excel_file, sheet_name=sheet)
sheet_sections, sheet_texts = self._process_dataframe_text_rows(df, context_label=sheet)
sections.extend(sheet_sections)
full_text_parts.extend(sheet_texts)
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=None,
author=None,
title=item.file_name,
sections=sections,
raw_full_text="\n\n".join(full_text_parts),
)
def _parse_csv(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Extracts free-text columns from CSV files with encoding detection."""
encodings_to_try = ["utf-8", "latin1", "cp1252"]
df = None
for enc in encodings_to_try:
try:
# Detect separator
with open(item.file_path, "r", encoding=enc, errors="replace") as f:
sample = f.read(2048)
sep = ","
if sample.count(";") > sample.count(","):
sep = ";"
elif sample.count("\t") > sample.count(","):
sep = "\t"
df = pd.read_csv(item.file_path, encoding=enc, sep=sep)
break
except Exception:
continue
if df is None:
df = pd.DataFrame()
sections, full_text_parts = self._process_dataframe_text_rows(df, context_label="CSV")
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=None,
author=None,
title=item.file_name,
sections=sections,
raw_full_text="\n\n".join(full_text_parts),
)
def _process_dataframe_text_rows(self, df: pd.DataFrame, context_label: str) -> Tuple[List[ExtractedSection], List[str]]:
"""Scans a DataFrame to locate free-text narrative columns and turns each significant row into an ExtractedSection."""
sections: List[ExtractedSection] = []
full_text_parts: List[str] = []
if df.empty:
return sections, full_text_parts
# Clean column names
df.columns = [str(c).strip() for c in df.columns]
# 1. Identify date column
date_col = None
for col in df.columns:
if re.search(r"\b(fecha|date|timestamp|dia)\b", col, re.IGNORECASE):
date_col = col
break
# 2. Identify author / speaker column
author_col = None
for col in df.columns:
if re.search(r"\b(empleado|colaborador|autor|usuario|nombre|persona|remitente|user)\b", col, re.IGNORECASE):
author_col = col
break
# 3. Identify narrative columns
narrative_cols = []
for col in df.columns:
if col in [date_col, author_col]:
continue
# Check by name
if self._narrative_col_regex.search(col):
narrative_cols.append(col)
continue
# Check by average word count of string entries
str_series = df[col].dropna().astype(str)
if not str_series.empty:
avg_words = str_series.apply(lambda x: len(x.split())).mean()
if avg_words >= 3.0:
narrative_cols.append(col)
if not narrative_cols:
# Fallback: take all object/string columns
narrative_cols = [c for c in df.select_dtypes(include=["object"]).columns if c not in [date_col, author_col]]
# 4. Extract rows
for idx, row in df.iterrows():
row_date = None
if date_col and pd.notna(row[date_col]):
try:
row_date = pd.to_datetime(row[date_col]).to_pydatetime()
except Exception:
pass
row_author = str(row[author_col]).strip() if author_col and pd.notna(row[author_col]) else None
row_texts = []
for col in narrative_cols:
val = row[col]
if pd.notna(val):
val_str = str(val).strip()
if len(val_str) > 5:
row_texts.append(f"[{col}]: {val_str}")
if row_texts:
content = " | ".join(row_texts)
sections.append(
ExtractedSection(
section_title=f"{context_label} - Fila {idx + 1}",
speaker_or_author=row_author,
timestamp=row_date,
content=content,
section_type="tabular_record",
)
)
full_text_parts.append(content)
return sections, full_text_parts
def _parse_eml(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Extracts email headers, sender, date, subject and plain text body from .eml files."""
with open(item.file_path, "rb") as f:
msg = email.message_from_binary_file(f, policy=policy.default)
sender = msg.get("From", "")
recipients = msg.get("To", "")
subject = msg.get("Subject", "")
date_str = msg.get("Date", "")
msg_date = None
if date_str:
try:
msg_date = email.utils.parsedate_to_datetime(date_str)
if msg_date.tzinfo:
msg_date = msg_date.replace(tzinfo=None)
except Exception:
pass
# Extract body
body_text = ""
if msg.is_multipart():
for part in msg.walk():
content_type = part.get_content_type()
content_disp = str(part.get("Content-Disposition", ""))
if content_type == "text/plain" and "attachment" not in content_disp:
payload = part.get_payload(decode=True)
if payload:
charset = part.get_content_charset() or "utf-8"
body_text += payload.decode(charset, errors="replace") + "\n"
else:
payload = msg.get_payload(decode=True)
if payload:
charset = msg.get_content_charset() or "utf-8"
body_text = payload.decode(charset, errors="replace")
sections: List[ExtractedSection] = []
header_summary = f"De: {sender} | Para: {recipients} | Asunto: {subject}"
sections.append(
ExtractedSection(
section_title="Encabezados de Correo",
speaker_or_author=sender,
timestamp=msg_date,
content=header_summary,
section_type="email_header",
)
)
clean_body = body_text.strip()
if clean_body:
sections.append(
ExtractedSection(
section_title=f"Asunto: {subject}",
speaker_or_author=sender,
timestamp=msg_date,
content=clean_body,
section_type="email_body",
)
)
raw_full = f"{header_summary}\n\n{clean_body}"
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=msg_date,
author=sender,
title=subject or item.file_name,
sections=sections,
raw_full_text=raw_full,
)
def _parse_txt(self, item: ScannedDocumentItem) -> ExtractedDocument:
"""Parses raw text files (.txt) with robust encoding fallback."""
encodings = ["utf-8", "latin1", "cp1252"]
content = ""
for enc in encodings:
try:
with open(item.file_path, "r", encoding=enc) as f:
content = f.read()
break
except Exception:
continue
paragraphs = [p.strip() for p in content.split("\n\n") if p.strip()]
sections = [
ExtractedSection(
section_title=f"Secci贸n {i+1}",
speaker_or_author=None,
timestamp=None,
content=p,
section_type="paragraph",
)
for i, p in enumerate(paragraphs)
]
return ExtractedDocument(
file_path=item.file_path,
file_name=item.file_name,
file_extension=item.file_extension,
department=item.department,
detected_date=None,
author=None,
title=item.file_name,
sections=sections,
raw_full_text=content,
)
|