Python exceptions when PDF to TXT conversion

[fhanzlik]

When I tried convert PDF document to text, no output was created and I got error:

$ ebook-convert V2510.pdf V2510_ebook.txt --enable-heuristics
Conversion options changed from defaults:
  enable_heuristics: True
1% Converting input to HTML...
InputFormatPlugin: PDF Input running
on /home/OTHER/data/dos/diskd/UctoFH_doklady/faDosle.fh/t-mobile/Vyuctovani_55052935_2510.pdf
pdftohtml log:
Page-1
Page-2
Page-3
Traceback (most recent call last):
  File "/usr/bin/ebook-convert", line 21, in <module>
    sys.exit(main())
             ~~~~^^
  File "/usr/lib64/calibre/calibre/ebooks/conversion/cli.py", line 429, in main
    plumber.run()
    ~~~~~~~~~~~^^
  File "/usr/lib64/calibre/calibre/ebooks/conversion/plumber.py", line 1089, in run
    self.oeb = self.input_plugin(stream, self.opts,
               ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^
                                self.input_fmt, self.log,
                                ^^^^^^^^^^^^^^^^^^^^^^^^^
                                accelerators, tdir)
                                ^^^^^^^^^^^^^^^^^^^
  File "/usr/lib64/calibre/calibre/customize/conversion.py", line 242, in __call__
    ret = self.convert(stream, options, file_ext,
                       log, accelerators)
  File "/usr/lib64/calibre/calibre/ebooks/conversion/plugins/pdf_input.py", line 66, in convert
    PDFDocument(xml, self.opts, self.log)
    ~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib64/calibre/calibre/ebooks/pdf/reflow.py", line 1476, in __init__
    self.find_header_footer()
    ~~~~~~~~~~~~~~~~~~~~~~~^^
  File "/usr/lib64/calibre/calibre/ebooks/pdf/reflow.py", line 1877, in find_header_footer
    if self.pages[head_page].texts \
       ~~~~~~~~~~^^^^^^^^^^^
IndexError: list index out of range

with calibre-8.0.1-5.fc42.x86_64 on Fedora 42 x86_64 Linux.
Same result is without --enable-heuristics option.

Calibre was installed with dependencies:
calibre-8.0.1-5.fc42.x86_64
optipng-7.9.1-1.fc42.x86_64
podofo-0.10.5-1.fc42.x86_64
python3-lxml-html-clean-0.4.2-1.fc42.noarch
python3-pyqt6-webengine-6.9.0-0.1.fc42.x86_64
python3-xxhash-3.6.0-1.fc42.x86_64
qt6-qtimageformats-6.9.3-1.fc42.x86_64
qt6-qtwebview-6.9.3-1.fc42.x86_64
libwebp-tools-1.5.0-2.fc42.x86_64
mathjax3-3.2.2-7.fc42.noarch
python3-apsw-3.47.2.0-2.fc42.x86_64
python3-css-parser-1.0.10-3.fc42.noarch
python3-html2text-2024.2.26-5.fc42.noarch
python3-html5-parser-0.4.12-5.fc42.x86_64
python3-mechanize-0.4.10-4.fc42.noarch
python3-pychm-0.8.6-16.fc42.x86_64
python3-regex-2024.11.6-1.fc42.x86_64
chmlib-0.40-45.fc42.x86_64

Know someone where problem could be?
(I'm quite new to Calibre/ebook-convert)
Thanks, Franta Hanzlik

[kovidgoyal]

You are using a very old and unsupported version of calibre. Uninstall it, and install the official binary from https://calibre-ebook.com/download_linux if the error still occurs with that, then follow the instructions in: https://www.mobileread.com/forums/sh...d.php?t=186697

[fhanzlik]

Hello Kovid, thanks for Your help!
I think I solved my problem by using pdftotext in the meantime. But I installed the latest version of calibre as you recommended. Unfortunately, the error still seems to occur (and the output is not produced):

$  ebook-convert 2041.pdf 2041.txt
1% Converting input to HTML...
InputFormatPlugin: PDF Input running
on /home/OTHER/tmp/2041.pdf
pdftohtml log:
Page-1
Page-2
Traceback (most recent call last):
  File "runpy.py", line 198, in _run_module_as_main
  File "runpy.py", line 88, in _run_code
  File "site.py", line 47, in <module>
  File "site.py", line 43, in main
  File "calibre/ebooks/conversion/cli.py", line 427, in main
  File "calibre/ebooks/conversion/plumber.py", line 1088, in run
  File "calibre/customize/conversion.py", line 242, in __call__
  File "calibre/ebooks/conversion/plugins/pdf_input.py", line 66, in convert
  File "calibre/ebooks/pdf/reflow.py", line 1474, in __init__
  File "calibre/ebooks/pdf/reflow.py", line 1885, in find_header_footer
IndexError: list index out of range

But I also found a file where the conversion went well (they are all accounting documents - invoices, from one organization).
If it would help, I can send you a PDF file where the error appears via e-mail (it is about 108kB in size).

$ ebook-convert --version
ebook-convert (calibre 8.16.2)
Created by: Kovid Goyal <kovid@kovidgoyal.net>

[kovidgoyal]

Yes, I will need the PDF to be able to help you further.

[fhanzlik]

Problematic file 2041.pdf should be uploaded.
Again, thank for Your effort!
Fr. Hanzlik

[kovidgoyal]

https://github.com/kovidgoyal/calibr...769a37e1bea083