Keep only PDF pages containing the text "ABC"

[magnussandstrom]

Hi all,

I’m looking for a way to extract pages from a PDF based on whether they contain a specific text string.

The text string would be variable and could appear on one or multiple pages in the incoming PDF. If a page contains the matching text, I would like that page to be extracted or saved as a separate PDF.

Has anyone solved something similar in Enfocus Switch, PitStop Server, or using another workflow component?

Any suggestions would be appreciated.

Thanks!

[JimmyHartington]

My idea would be to split the pdf.
Have an action list with "Select text containg" the variable.
Log the selection as error.
Output the errors from Pitstop Server configurator.

[freddyp]

Try "Strategy - Action List" of this app: https://www0.enfocus.com/en/appstore/pr ... -pdf-pages. Make sure to read the documentation because the Action List has to have a specific behavior.

[laurentd]

Or you can run this action list, then split the remaining pages.

Remove pages not containing desired text.png (30.73 KiB) Viewed 764 times

[magnussandstrom]

Solved it with a Python script and Run Command, fast and easy:

Code: Select all

#!/usr/bin/env python3

import sys
import re
import fitz
from pathlib import Path


def normalize_text(value: str) -> str:
    return " ".join(value.split())


def main() -> int:
    if len(sys.argv) != 4:
        print("Usage: extract_printvis_assignment.py input.pdf output.pdf SOR414620-2", file=sys.stderr)
        return 1

    input_pdf = sys.argv[1]
    output_pdf = sys.argv[2]
    target_order_no = sys.argv[3].strip()

    if not target_order_no:
        print("ERROR: Empty target order number", file=sys.stderr)
        return 1

    pattern = re.compile(
        r"Order\s+No\.?\s+" + re.escape(target_order_no),
        re.IGNORECASE
    )

    matched_pages = []

    try:
        source_doc = fitz.open(input_pdf)
        output_doc = fitz.open()

        for page_index in range(source_doc.page_count):
            page = source_doc.load_page(page_index)

            # Search only in the top part of the page for performance.
            page_rect = page.rect
            header_rect = fitz.Rect(
                0,
                0,
                page_rect.width,
                page_rect.height * 0.18
            )

            page_text = page.get_text("text", clip=header_rect)
            normalized_text = normalize_text(page_text)

            if pattern.search(normalized_text):
                output_doc.insert_pdf(
                    source_doc,
                    from_page=page_index,
                    to_page=page_index
                )
                matched_pages.append(page_index + 1)

        if not matched_pages:
            print(f"ERROR: Did not find Order No. {target_order_no}", file=sys.stderr)
            return 3

        Path(output_pdf).parent.mkdir(parents=True, exist_ok=True)
        output_doc.save(output_pdf, garbage=3, deflate=True)
        output_doc.close()
        source_doc.close()

        print(f"Extracted Order No. {target_order_no}")
        print(f"Pages: {','.join(map(str, matched_pages))}")
        return 0

    except Exception as e:
        print(f"ERROR: Failed while processing PDF: {e}", file=sys.stderr)
        return 4


if __name__ == "__main__":
    sys.exit(main())