Adding a transformation to a page that has content linked from a table of contents removes the content from the page

[glynam1]

When scaling down content to fit it onto an a4 page, we noticed an issue where if a user has used the table of contents functionality to generate links to pages, each page that contains one of the "destination" links, will end up blank after the transformation.

Environment

reproducible on python 3.9 and pypdf 3.17.1

Code + PDF

import tempfile

from pypdf import PdfReader, PdfWriter, Transformation


def scale_pdf():
    _, temp_output_path = tempfile.mkstemp(suffix=".pdf")
    reader = PdfReader("example.pdf")
    writer = PdfWriter()
    for page in reader.pages:
        t = Transformation()
        t = t.scale(1, 1)
        page.add_transformation(t)
        writer.add_page(page)
    writer.write(temp_output_path)
    shutil.move(temp_output_path, "result.pdf")

if __name__ == '__main__':
    scale_pdf()

example.pdf

result.pdf

[glynam1]

This is a regression from 3.11.0, testing on 3.10.0 fixes the issue

[pubpub-zz]

using add_page you are loosing the link between pages that do exist in the source.
this is the way I propose to do it:

import pypdf
w=pypdf.PdfWriter(clone_from="example.pdf")
for p in w.pages:
    t=pypdf.Transformation().scale(1,1)
    p.add_transformation(t)
w.write("result.pdf")

[glynam1]

using add_page you are loosing the link between pages that do exist in the source. this is the way I propose to do it:

import pypdf
w=pypdf.PdfWriter(clone_from="example.pdf")
for p in w.pages:
    t=pypdf.Transformation().scale(1,1)
    p.add_transformation(t)
w.write("result.pdf")

This works great, thanks for that!