pdf2odt

pdf2odt is a Python command-line utility and library that converts PDF documents into LibreOffice Writer (.odt) documents.

Key Features

• 100% Pure Python & Self-Contained: No need to install external system tools such as Poppler, Tesseract, or LibreOffice. Everything is installed via pip.
• High-Quality Page Rendering: Uses PyMuPDF (MuPDF) for fast and pixel-perfect rendering to PNG images with configurable DPI resolution.
• Smart Hybrid Text Extraction & OCR:
  • Extracts native vector text directly from digital PDFs with 100% accuracy and near-zero latency.
  • Automatically runs RapidOCR (ONNX Runtime) on scanned images or bitmap pages to recognize text.
• Character-Anchored Images (as-char): Page images are embedded in the ODT document using odfdo with proportional dimensions and anchored as characters, ensuring consistent layout in LibreOffice Writer.
• Fast & Multi-Threaded: Uses Python thread pooling to process pages concurrently across all available CPU cores.

Installation

Install pdf2odt using pip:

pip install pdf2odt

Or using Poetry:

poetry add pdf2odt

Command-Line Usage

1. Basic Conversion

Convert a PDF into an ODT document (renders pages at default 300 DPI):

pdf2odt --pdf document.pdf output.odt

2. Conversion with Text Extraction / OCR

Extract native text and perform OCR on images, inserting the text below each page image in the ODT document:

pdf2odt --pdf document.pdf --ocr output.odt

3. Custom Image Resolution

Set a custom image resolution in DPI (default is 300 DPI; use lower values like 150 DPI for smaller file sizes):

pdf2odt --pdf document.pdf --resolution 150 output.odt

4. Full Options Reference

usage: pdf2odt [-h] [--version] --pdf PDF [--resolution RESOLUTION] [--ocr] output

Converts a pdf to a LibreOffice Writer document with pages as images

positional arguments:
  output                Output odt file

options:
  -h, --help            show this help message and exit
  --version             show program's version number and exit
  --pdf PDF             PDF file to convert
  --resolution RESOLUTION
                        Sets DPI image resolution. Default is 300
  --ocr                 Extracts text with page.get_text() or OCR and inserts result after image in ODT document

Python API Usage

You can also use pdf2odt directly in your Python applications:

from pdf2odt.core import main_command

# Convert a PDF to ODT with 300 DPI and OCR enabled
main_command(
    pdf="path/to/document.pdf",
    resolution=300,
    ocr=True,
    output="path/to/output.odt"
)

Dependencies

pdf2odt relies on the following Python packages:

• PyMuPDF: High-performance PDF rendering and native text extraction.
• odfdo: Pure Python OpenDocument (.odt) document generator.
• rapidocr-onnxruntime: Lightweight ONNX-powered OCR engine.
• Pillow: Image dimension and format processing.
• tqdm: Console progress bar.
• colorama: Colored terminal output.

Development & Testing

This project uses Poetry and Poe the Poet for development tasks.

Run Tests

poetry run poe test

Run Coverage Report

poetry run poe coverage

Update Translations

poetry run poe translate

License

Distributed under the GPL-3.0 License.