docx-extractor 1.0.0

Recursive DOCX extractor with embedded files support

DOCX Extractor

Recursive DOCX extractor that handles embedded files (Excel, Word, SQL/TXT) with compact TOON output format.

Extracts:

• ✅ DOCX paragraphs & tables (with full cell data)
• ✅ Embedded Excel sheets (all rows/columns)
• ✅ Nested DOCX (recursive up to 5 levels)
• ✅ SQL/TXT files
• ✅ Filters empty content

🚀 Quick Start

Install

pip install docx-extractor

Extract DOCX to TOON format

docx-extract input.docx -o output.toon

With options

docx-extract input.docx -o result.toon --depth 3 -v

📄 Sample TOON Output

FILE input.docx .docx L0 PARA 1 Project Summary Report TABLE 1 3rows x 4cols ROW T1.R1 Project|Q1|Q2|Q3|Total ROW T1.R2 ProjectA|100|150|200|450 ROW T1.R3 ProjectB|80|120|180|380 EMBEDDED 1 FILE sheet1.xlsx .xlsx L1 SHEET Sales 15rows ROW Sales.1 Product|Units|Price|Total ROW Sales.2 WidgetA|100|25.50|2550.00

🛠️ Installation

From PyPI

pip install docx-extractor

From Source

git clone https://github.com/yourusername/docx-extractor.git cd docx-extractor pip install -e .

🎯 Usage

docx-extract [OPTIONS] <docx_file>

Options:

-o, --output Output TOON file (default: extraction.toon) -d, --depth Max recursion depth (default: 5) -v, --verbose Show preview -h, --help Show help

🏗️ Package Structure

docx_extractor/ ├── core.py # Main extraction logic ├── processors.py # Excel/TXT handlers ├── toon_formatter.py # TOON output formatter └── init.py

🔧 Programmatic Usage

from docx_extractor import extract_docx_recursive from docx_extractor.toon_formatter import format_toon

Extract

result = extract_docx_recursive("input.docx", max_depth=3)

TOON format

toon_lines = format_toon(result)

Save

with open("output.toon", "w") as f: f.write("\n".join(toon_lines))

📊 TOON Format Specification

Tab-separated, hierarchical, token-efficient:

TYPE IDENTIFIER DATA FILE filename.docx .docx L0 PARA 1 Document title TABLE 1 3rows x 4cols ROW T1.R1 Project|Q1|Q2|Total SHEET Sales 15rows ROW Sales.1 Product|100|25.50 TEXT 1250chars First 1000 chars...

🧪 Features Tested

• DOCX paragraphs (non-empty only)
• DOCX tables (full cell data)
• Embedded Excel (.xlsx/.xls)
• Nested DOCX (recursive)
• SQL/TXT files
• OLE package stream extraction
• Empty content filtering
• TOON compact output

🔍 Dependencies

Package	Purpose
python-docx	DOCX parsing
openpyxl	Excel reading
olefile	OLE embedded extraction

⚙️ Development

Install dev dependencies

pip install -e .[dev]

Build package

python -m build

Test

pytest tests/

Publish

twine upload dist/*

📄 License

MIT License - see LICENSE for details.

🙏 Acknowledgments

• python-docx
• openpyxl
• olefile

---

Made with ❤️ for document automation