language-processing-tool 0.2.7

A PDF language detection and OCR tool

Language Processing Tool

Language Processing Tool is a Python package that processes PDFs and detects languages in documents. It supports both text-based and scanned PDFs using OCR.

Installation

You can install the package from PyPI using:

pip install language-processing-tool

Features

• Extracts text from PDFs (both text-based and scanned).
• Uses OCR (Tesseract) for image-based PDFs.
• Detects the dominant language and provides a language distribution.
• Supports batch processing of multiple PDFs.
• CLI and programmatic usage available.

Usage

In Python Code

To use this package in your Python script:

from language_processing_tool.process_pdfs import process_pdfs, process_single_file

# Process a single PDF file
print(process_single_file("/path/to/file.pdf"))

# Process multiple PDFs from a directory (using CSV input)
process_pdfs("/path/to/input_folder/", "pdf_files.csv", "/path/to/output_folder/")

Command Line Usage

After installing the package, you can use the CLI tool:

Process a Single PDF

process-pdfs /path/to/file.pdf

Batch Processing (Using CSV)

process-pdfs /path/to/input_folder/ /path/to/csv_file.csv /path/to/output_folder/

CSV File Format

The CSV file should contain a column named filename, listing the PDF filenames (without extensions). Example:

filename
document1
document2
document3

Make sure the corresponding PDFs (document1.pdf, document2.pdf, etc.) are in the specified input folder.

Project Structure

language_processing_tool
├── LICENSE
├── README.md
├── setup.py
├── pyproject.toml
├── requirements.txt
├── language_processing_tool
│   ├── __init__.py
│   ├── process_pdfs.py
│   ├── sourcecode.py
├── tests
│   ├── test.py

Entry Points

The package includes a CLI entry point:

process-pdfs <arguments>

which maps to the main() function in process_pdfs.py.

Dependencies

This package requires:

• pytesseract
• langdetect
• pandas
• PyMuPDF
• icecream
• Pillow
• argparse

License

This project is licensed under the MIT License.

Author

• Harish Kumar S
• GitHub: Harish-nika
• Email: [e-mail]
• portfolio: Harish Kumar S - AI ML Engineer