copyable-pdf

A lightweight, dependency-minimal bash script to convert scanned PDFs into searchable PDFs using Tesseract OCR.

copyable-pdf takes a PDF input, converts each page to an image, performs OCR (Optical Character Recognition) using Tesseract, and merges them back into a single, searchable PDF document.

Features

• OCR: Make scanned documents searchable and copyable.
• Parallel Processing: Uses multiple cores for faster OCR.
• Dependency Check: Automatically checks for missing tools.
• Customizable: Set language and DPI.

Installation

Via Homebrew

brew install maxgfr/tap/copyable-pdf

Manual Installation

1. Clone the repository:

   git clone https://github.com/maxgfr/copyable-pdf.git
   cd copyable-pdf

2. Make the script executable:

   chmod +x script.sh

3. (Optional) Move to your bin directory:

   mv script.sh /usr/local/bin/copyable-pdf

Dependencies

Ensure you have the following installed:

• tesseract: For OCR.
• poppler: For pdftoppm and pdfunite.

On macOS (Homebrew):

brew install tesseract poppler

On Ubuntu/Debian:

sudo apt-get install tesseract-ocr poppler-utils

Usage

copyable-pdf [options] input.pdf

Options

Option	Description	Default
-l, --lang <code>	Language code (e.g., fra, eng)	eng
-o, --output <path>	Custom output file path	input_ocr.pdf
-d, --dpi <num>	DPI resolution for OCR	300
-j, --jobs <num>	Number of parallel jobs	Auto-detect
-t, --text	Generate an additional .txt file	false
-m, --markdown	Generate an additional .md file	false
-k, --keep	Keep temporary files (debug)	false
-v, --verbose	Verbose output	false
-h, --help	Show help message	-

Examples

Basic usage:

copyable-pdf document.pdf

Specify language (French) and higher DPI:

copyable-pdf -l fra -d 600 document.pdf

Explicitly set output filename:

copyable-pdf -o searchable_doc.pdf scan.pdf

License

MIT