read-docx

name	read-docx
description	Read Word DOCX files and extract text content with proper encoding handling.
metadata	{"clawdbot":{"emoji":"📘","os":["linux","darwin","win32"]}}

Usage

CLI

# Read single file or wildcard pattern
python3 /path/to/your/.openclaw/workspace/skills/read-docx/read-docx.py "path/*.docx"

# Verbose mode
python3 read-docx.py "exams/*.docx" --verbose

# Examples
python3 read-docx.py "exams/HDMI*.docx"
python3 read-docx.py "/path/to/document.docx" -v

Python Code

from docx import Document
import glob
import os

# Read DOCX files (use glob to handle Chinese filename encoding issues)
for f in glob.glob("*.docx"):
    try:
        doc = Document(f)
        # Fast extraction using generator
        text = '\n'.join(p.text for p in doc.paragraphs if p.text.strip())
        
        # Alternative: extract from runs directly (faster for large docs)
        # text = '\n'.join(run.text for para in doc.paragraphs for run in para.runs if run.text.strip())
        
        print(f"--- {os.path.basename(f)} ---")
        print(text)
    except Exception as e:
        print(f"Failed to process {f}: {e}")

Why Not read Tool?

• ❌ OpenClaw's read tool only supports: text files, images (jpg/png/gif/webp)
• ❌ DOCX is a ZIP archive containing XML - needs specialized parsing
• ✅ python-docx handles DOCX structure properly

Important Notes

Chinese Filename Encoding

Chinese filenames may have encoding issues when accessed directly. Always use glob wildcards:

# ✅ Good - glob handles encoding
import glob
for f in glob.glob("考试*.docx"):
    doc = Document(f)

# ❌ Bad - may fail with Chinese filenames
doc = Document("考试.docx")  # Might not find the file

Dependencies

# Already installed
pip3 install python-docx

Performance Tips

1. Use wildcards - Avoids encoding issues with Chinese filenames
2. Generator expression - '\n'.join(p.text for p in ...) avoids intermediate list
3. Filter empty paragraphs - if p.text.strip() reduces output noise
4. For large docs - Extract from runs directly (see code above)

File Structure

read-docx/
├── SKILL.md          # This file
├── read-docx.py      # CLI script
└── _meta.json        # Skill metadata

待补充描述 /path/to/your/.openclaw/workspace/skills/read-docx