# Word Analysis

Word (.docx/.doc) 文档全量解析。覆盖：正文/段落文本提取、表格数据提取、高亮/颜色格式读取、多文件汇总对比、嵌入图片转 caption。

- Repository: https://github.com/OpenSenseNova/SenseNova-Skills (6002 stars)
- Category: Documents
- License: MIT
- Last updated: 2026-10-29
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add OpenSenseNova/SenseNova-Skills --skill word-analysis -g -a claude-code -y`
Codex: `npx skills add OpenSenseNova/SenseNova-Skills --skill word-analysis -g -a codex -y`
Cursor: `npx skills add OpenSenseNova/SenseNova-Skills --skill word-analysis -g -a cursor -y`
Gemini CLI: `npx skills add OpenSenseNova/SenseNova-Skills --skill word-analysis -g -a gemini-cli -y`

## Description

Word (.docx/.doc) 文档全量解析。覆盖：正文/段落文本提取、表格数据提取、高亮/颜色格式读取、多文件汇总对比、嵌入图片转 caption。

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Word Analysis, .docx / .doc

## Environment

```python
from docx import Document
import os

# python-docx is available; for .doc (old format) convert via libreoffice first
def load_doc(path):
    """Load .docx directly; convert .doc to .docx first if needed."""
    if path.lower().endswith('.doc'):
        import subprocess
        out_dir = os.path.dirname(path)
        subprocess.run(
            ['libreoffice', '--headless', '--convert-to', 'docx', '--outdir', out_dir, path],
            check=True, capture_output=True
        )
        path = path.rsplit('.', 1)[0] + '.docx'
    return Document(path)
```

---

## Core Method 1: Full Text Extraction

```python
def extract_full_text(doc_path):
    """Extract all text: paragraphs + table cells, in document order."""
    doc = load_doc(doc_path)
    lines = []

    # Iterate paragraphs and tables in body order
    from docx.oxml.ns import qn
    for block in doc.element.body:
        tag = block.tag.split('}')[-1]
        if tag == 'p':
            # Paragraph
            from docx.text.paragraph import Paragraph
            para = Paragraph(block, doc)
            text = para.text.strip()
            if text:
                lines.append(text)
        elif tag == 'tbl':
            # Table
            from docx.table import Table
            tbl = Table(block, doc)
            for row in tbl.rows:
                row_text = '\t'.join(cell.text.strip() for cell in row.cells)
                if row_text.strip():
                    lines.append(row_text)

    return '\n'.join(lines)

# Usage
text = extract_full_text("/mnt/data/doc.docx")
print(text[:2000])  # preview first 2000 chars
```

---

## Core Method 2: Table Extraction (Structured)

```python
import pandas as pd

def extract_all_tables(doc_path):
    """Extract all tables from a Word document as list of DataFrames."""
    doc = load_doc(doc_path)
    tables = []

The rest: https://github.com/OpenSenseNova/SenseNova-Skills/blob/HEAD/skills/sn-da-non-spreadsheet-analysis/capability/word-analysis/SKILL.md