Reusing PdfMerger after write generates PDF with extra pages

[bashirmindee]

I was trying to merge the same PDF with itself multiple number of times. 1st I want to have the original PDF, then I want to have the PDF duplicated, then I want to duplicate it three times, and so forth.

Environment

ubuntu 20.04
Python==3.8.12+
Package Version

---

pip==21.1.1
PyPDF2==2.10.5
setuptools==56.0.0
typing-extensions==4.3.0

$ python -m platform
Linux-5.11.0-40-generic-x86_64-with-glibc2.29
$ python -c "import PyPDF2;print(PyPDF2.__version__)"
2.10.5

Code + PDF

This is a minimal, complete example that shows the issue:

##script.py

from PyPDF2 import PdfReader, PdfMerger

merger = PdfMerger()
reader = PdfReader("blank.pdf")

for j in range(9):
    merger.append(reader)
    merger.write(f"generated_pdfs/{len(merger.pages)}.pdf")

Here is the blank.pdf that causes the issue.

Expected behavior

1.pdf: must contain 1 page but contains 1 page ✅
2.pdf: must contain 2 page but contains 2 page ❌
3.pdf: must contain 3 page but contains 3 page ❌
4.pdf: must contain 4 page but contains 4 page ❌
5.pdf: must contain 5 page but contains 5 page ❌
6.pdf: must contain 6 page but contains 6 page ❌

[MartinThoma]

Interesting. For PyPDF2==2.10.9

• 1.pdf contains 1 page
• 2.pdf contains 3 pages (+2)
• 3.pdf contains 6 pages (+3)
• 4.pdf contains 10 pages (+4)
• 5.pdf contains 15 pages (+5)
• ...

I'm not sure why ...

[pubpub-zz]

@bashirmindee
a change part of PR #1371 in f9d7d19 should fix it. The other commits should not be required.
the change is small you should be able to copy it, if you want to try

[pubpub-zz]

@bashirmindee
with the lastest version of pypdf

##script.py

from PyPDF2 import PdfReader, PdfWriter

writer = PdfWriter()
reader = PdfReader("blank.pdf")

for j in range(9):
    writer.append(reader)    
    writer.write(f"generated_pdfs/{len(writer.pages)}.pdf")
    writer.reset_translation(reader)  # to append independent pages

[pubpub-zz]

I close this as solved