Imported from bighardperson/computer-science-skills-collection (skills/read-word/SKILL.md). Install upstream with npx skills add bighardperson/computer-science-skills-collection --skill read-word. Copyright stays with the author (MIT).

Read Word Document

A professional tool for reading Microsoft Word documents, supporting both modern .docx and legacy .doc formats with full Chinese language support.

Features

• Read .docx files - Word 2007 and later format
• Read .doc files - Word 97-2003 format via OLE parsing
• Auto format detection - Automatically identifies file type
• Full Chinese support - Handles Chinese encoding correctly
• Keyword search - Search for keywords across all paragraphs
• Export to text - Save as UTF-8 text files
• Document analysis - Get document statistics and info
• No Word required - Works without Microsoft Word installation

Installation

Prerequisites

pip install python-docx olefile

Install Skill

# Copy to your OpenClaw skills directory
cp -r read-word ~/.openclaw/skills/

Usage

Command Line

# Basic reading (shows first 100 paragraphs)
python ~/.openclaw/skills/read-word/read_word.py "document.docx"

# Show more content
python ~/.openclaw/skills/read-word/read_word.py "document.docx" --limit 200

# Search for keywords
python ~/.openclaw/skills/read-word/read_word.py "document.docx" --search "keyword1,keyword2"

# Save as text file
python ~/.openclaw/skills/read-word/read_word.py "document.docx" --output "output.txt"

# Show document info only
python ~/.openclaw/skills/read-word/read_word.py "document.docx" --info

Python API

# Method 1: Import functions
import sys
sys.path.insert(0, '~/.openclaw/skills/read-word')
from read_word import read_word_document, search_in_document

# Read document
paragraphs = read_word_document("document.docx")
for para in paragraphs:
    print(para)

# Search keywords
results = search_in_document("document.docx", ["keyword1", "keyword2"])

Examples

Example 1: Read and Analyze

from read_word import read_word_document

paragraphs = read_word_document("report.docx")
print(f"Document has {len(paragraphs)} paragraphs")

# Show first 10 paragraphs
for i, p in enumerate(paragraphs[:10]):
    print(f"{i+1}. {p}")

Example 2: Search Keywords

from read_word import search_in_document

# Find paragraphs containing "kitchen" or "feng shui"
results = search_in_document("book.docx", ["kitchen", "feng shui"])
for r in results:
    print(r)

Example 3: Batch Processing

from pathlib import Path
from read_word import read_word_document

desktop = Path.home() / "Desktop"
for doc_file in desktop.glob("*.docx"):
    paragraphs = read_word_document(doc_file)
    print(f"{doc_file.name}: {len(paragraphs)} paragraphs")

API Reference

read_word_document(filepath)

Read a Word document and return a list of paragraphs.

Parameters:

• filepath (str|Path): Path to the Word document

Returns:

• list: List of paragraph strings

Raises:

• FileNotFoundError: If file doesn't exist
• ValueError: If file format is not supported

search_in_document(filepath, keywords)

Search for keywords in a Word document.

Parameters:

• filepath (str|Path): Path to the Word document
• keywords (list): List of keywords to search for

Returns:

• list: Matching paragraphs with format "[Paragraph N] content"

save_as_text(paragraphs, output_path)

Save paragraphs to a UTF-8 text file.

Parameters:

• paragraphs (list): List of paragraph strings
• output_path (str|Path): Output file path

analyze_document(filepath)

Analyze document and return statistics.

Returns:

• dict: Contains filename, size, paragraphs count, total characters

Troubleshooting

Error: ModuleNotFoundError: No module named 'docx'

Solution: pip install python-docx

Error: Legacy .doc file shows garbled text

Reason: OLE parsing has limitations with complex formatting Solution: Convert .doc to .docx using Microsoft Word, then read

Error: Chinese characters display incorrectly

Reason: Terminal encoding issue Solution: Use --output to save to file, then open with editor

File Support

Permissions

• Read: User-specified Word documents
• Write (optional): Output .txt files when using --output
• Network: None

Security

Risk Level: LOW - Local file operations only, no network access, original files are never modified.

Changelog

v1.0.0 (2026-03-20)

• Initial release
• Support .docx and .doc formats
• Keyword search functionality
• Text export capability
• Chinese encoding support

Author

叶文洁 (Ye Wenjie) - Created for reading Feng Shui books and Word documents

License

MIT License