Text not extracted properly in 4.30.1, but works with 4.30.0

[VikParuchuri]

Checklist

• I confirm to be using an official package of pypdfium2 from PyPI or GitHub/pypdfium2-team.

Description

Text was extracted from the table on page 4 of this pdf properly with pypdfium2==4.30.0, but no longer works on 4.30.1. Only a partial set of characters are pulled from the table.

I'm not sure if this is an issue with the python bindings, or base pdfium - happy to close and open an issue there if it is.

Install Info

Python 3.11.3 (main, May 23 2023, 07:06:17) [Clang 14.0.3 (clang-1403.0.22.14.1)]

macOS-15.0.1-arm64-arm-64bit

pypdfium2 4.30.0

pdfium 126.0.6462.0 at pypdfium2_raw/libpdfium.dylib

Name: pypdfium2
Version: 4.30.0
Summary: Python bindings to PDFium
Home-page: https://github.com/pypdfium2-team/pypdfium2
Author: pypdfium2-team
Author-email: [email protected]
License: (Apache-2.0 OR BSD-3-Clause) AND LicenseRef-PdfiumThirdParty
Location: lib/python3.11/site-packages
Requires:

[mara004]

I ran the following commands on our current pypdfium2 source tree:

# assuming an editable install (python3 -m pip install -e .)
./run emplace auto:6462
pypdfium2 extract-text "p25-1144.pdf" --pages 4 --strategy range > pdfium_6462_range.txt
pypdfium2 extract-text "p25-1144.pdf" --pages 4 --strategy bounded > pdfium_6462_bounded.txt
./run emplace auto:6899
pypdfium2 extract-text "p25-1144.pdf" --pages 4 --strategy range > pdfium_6899_range.txt
pypdfium2 extract-text "p25-1144.pdf" --pages 4 --strategy bounded > pdfium_6899_bounded.txt

diff pdfium_6462_range.txt pdfium_6899_range.txt

85,90c85,90
< Under 18 years. . . . . . . . . . . . . . . 73.6 74.0 75.7 77.1 78.2 80.1 6.5 8.8
< 18 to 44 years. . . . . . . . . . . . . . . . 116.0 119.2 125.0 126.4 129.6 132.7 16.7 14.4
< 45 to 64 years. . . . . . . . . . . . . . . . 84.3 83.4 81.3 89.1 95.4 97.0 12.7 15.1
< 65 years and over. . . . . . . . . . . . 49.2 56.1 73.1 80.8 85.7 94.7 45.4 92.3
< 85 years and over. . . . . . . . . . . . 6.4 6.7 9.1 14.4 18.6 19.0 12.6 198.1
< 100 years and over. . . . . . . . . . . 0.1 0.1 0.1 0.2 0.4 0.6 0.5 618.3
---
> Under 18 years 73.6 74.0 75.7 77.1 78.2 80.1 6.5 8.8
> 18 to 44 years 116.0 119.2 125.0 126.4 129.6 132.7 16.7 14.4
> 45 to 64 years 84.3 83.4 81.3 89.1 95.4 97.0 12.7 15.1
> 65 years and over 49.2 56.1 73.1 80.8 85.7 94.7 45.4 92.3
> 85 years and over 6.4 6.7 9.1 14.4 18.6 19.0 12.6 198.1
> 100 years and over 0.1 0.1 0.1 0.2 0.4 0.6 0.5 618.3

diff pdfium_6462_bounded.txt pdfium_6899_bounded.txt

84,90c84,90
<    Total population . . . . . . 323.1 332.6 355.1 373.5 388.9 404.5 81.4 25.2
< Under 18 years. . . . . . . . . . . . . . . 73.6 74.0 75.7 77.1 78.2 80.1 6.5 8.8
< 18 to 44 years. . . . . . . . . . . . . . . . 116.0 119.2 125.0 126.4 129.6 132.7 16.7 14.4
< 45 to 64 years. . . . . . . . . . . . . . . . 84.3 83.4 81.3 89.1 95.4 97.0 12.7 15.1
< 65 years and over. . . . . . . . . . . . 49.2 56.1 73.1 80.8 85.7 94.7 45.4 92.3
< 85 years and over. . . . . . . . . . . . 6.4 6.7 9.1 14.4 18.6 19.0 12.6 198.1
< 100 years and over. . . . . . . . . . . 0.1 0.1 0.1 0.2 0.4 0.6 0.5 618.3
---
>    Total population . . . . . .  323.1 332.6 355.1 373.5 388.9 404.5 81.4 25.
> Under 18 years. . . . . . . . . . . . . . .  73.6 74
> 18 to 44 years. . . . . . . . . . . . . . . .  116.0 119.2 1
> 45 to 64 years. . . . . . . . . . . . . . . .  84.3 83
> 65 years and over. . . . . . . . . . . .  49.2 56.1 73.1 
> 85 years and over. . . . . . . . . . . .  6.4 6.7 9.1 1
> 100 years and over. . . . . . . . . . .  0.1 0.1 0.1

That does look like a regression in pdfium concerning the GetTextBounded extraction strategy.
pypdfium2 is only forwarding what pdfium returns, so yes, please file an issue about this upstream. Thanks!

FWIW, the python helpers have barely changed between 4.30.0 and 4.30.1 (see diff) – this was mostly setup bug fixes and a pdfium update.

[mara004]

Might be interesting to bisect this to narrow down the commit range...

[mara004]

OK, could narrow this down to 6721 (good) - 6844 (bad) [commit log]
Unfortunately, there are no pdfium-binaries builts in between, so this is still a rather large span.

[mara004]

I went ahead and filed https://issues.chromium.org/issues/387277993 for this now.
Feel free to comment on the thread if you have anything to add. Thanks!