PDF OCR Extractor#

Use this skill to extract text from scanned PDFs or image-based PDFs that lack a native text layer. It's completely free, doesn't utilize third-party APIs, and offers unlimited usage. It renders PDF pages to images and runs optical character recognition (OCR).

Dependencies#

This skill requires:

1. System Binary: tesseract (along with required language data packs like chi_sim or eng).
2. Python Packages: pypdfium2, pytesseract, and Pillow.

Note: Do not run automated pip install commands at runtime. Rely on the user or the environment to pre-install the dependencies defined in the metadata block.

Quick Start#

Create a Python script (e.g., extract.py) in a temporary directory to handle the extraction safely:

python

import pypdfium2 as pdfium
import pytesseract
from PIL import Image
import sys
import os

def extract(pdf_path):
    doc = pdfium.PdfDocument(pdf_path)
    full_text = []
    for i, page in enumerate(doc):
        # Render page to a high-resolution image
        bitmap = page.render(scale=2)
        tmp_img = f"/tmp/page_{i}.png"
        bitmap.to_pil().save(tmp_img)

        # Run OCR (assuming English and Simplified Chinese packs are installed)
        text = pytesseract.image_to_string(Image.open(tmp_img), lang='chi_sim+eng')
        full_text.append(text)

        # Cleanup temporary file
        os.remove(tmp_img)

    return "\n".join(full_text)

if __name__ == "__main__":
    if len(sys.argv) > 1:
        print(extract(sys.argv[1]))

Then execute the script:

bash

python3 extract.py /path/to/document.pdf

Security & Sandbox Constraints#

• Write temporary images only to /tmp/ and clean them up immediately after extraction.
• Do not attempt to dynamically download or install language packs via shell commands; notify the user if a specific language is missing.