word-analysis

1# Word Analysis — .docx / .doc
2 
3## Environment
4 
5```python
6from docx import Document
7import os
8 
9# python-docx is available; for .doc (old format) convert via libreoffice first
10def load_doc(path):
11    """Load .docx directly; convert .doc to .docx first if needed."""
12    if path.lower().endswith('.doc'):
13        import subprocess
14        out_dir = os.path.dirname(path)
15        subprocess.run(
16            ['libreoffice', '--headless', '--convert-to', 'docx', '--outdir', out_dir, path],
17            check=True, capture_output=True
18        )
19        path = path.rsplit('.', 1)[0] + '.docx'
20    return Document(path)
21```
22 
23---
24 
25## Core Method 1: Full Text Extraction
26 
27```python
28def extract_full_text(doc_path):
29    """Extract all text: paragraphs + table cells, in document order."""
30    doc = load_doc(doc_path)
31    lines = []
32 
33    # Iterate paragraphs and tables in body order
34    from docx.oxml.ns import qn
35    for block in doc.element.body:
36        tag = block.tag.split('}')[-1]
37        if tag == 'p':
38            # Paragraph
39            from docx.text.paragraph import Paragraph
40            para = Paragraph(block, doc)
41            text = para.text.strip()
42            if text:
43                lines.append(text)
44        elif tag == 'tbl':
45            # Table
46            from docx.table import Table
47            tbl = Table(block, doc)
48            for row in tbl.rows:
49                row_text = '\t'.join(cell.text.strip() for cell in row.cells)
50                if row_text.strip():
51…