Bug with error `TypeError: 'PDFObjRef' object is not iterable`

[pprados]

Bug report

With some PDF file, the code crash with TypeError: 'PDFObjRef' object is not iterable.

I propose a PR here

[pprados]

@pietermarsman or @euske

Can you integrate this PR and publish a new version?

pdfplumber crashes because of this. With LangChain, the leading LLM framework, it is not possible to use pdfminer or pdfplumber. Pdfplumber must force the use of an old version.

I want to help standardize the various parsers, but that's not possible without fixing this bug.

[bb-at-ss]

Yup. I'm still getting this bug too on Version: 20240706

TypeError: 'PDFObjRef' object is not iterable.

[pprados]

@pietermarsman or @euske

Can you integrate this PR and publish a new version?

[pietermarsman]

Will check it out somewhere this month

…

On Fri, 7 Mar 2025, 16:10 Philippe PRADOS, ***@***.***> wrote: @pietermarsman <https://github.com/pietermarsman> or @euske <https://github.com/euske> Can you integrate this PR and publish a new version? — Reply to this email directly, view it on GitHub <#1082 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA5H44FXLHYLNHBZT4KFPYD2TGZFNAVCNFSM6AAAAABX26FXPWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMBWGY4TCOJVHE> . You are receiving this because you were mentioned.Message ID: ***@***.***> [image: pprados]*pprados* left a comment (pdfminer/pdfminer.six#1082) <#1082 (comment)> @pietermarsman <https://github.com/pietermarsman> or @euske <https://github.com/euske> Can you integrate this PR and publish a new version? — Reply to this email directly, view it on GitHub <#1082 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA5H44FXLHYLNHBZT4KFPYD2TGZFNAVCNFSM6AAAAABX26FXPWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDOMBWGY4TCOJVHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[pietermarsman]

I can replicate this with the test pdf from the MR:

$ PYTHONPATH=. .venv/bin/python tools/pdf2txt.py ~/Downloads/annotations.pdf                                                                           08:05:56
Traceback (most recent call last):
  File "/home/pieter/projects/pdfminer.six/tools/pdf2txt.py", line 323, in <module>
    sys.exit(main())
             ~~~~^^
  File "/home/pieter/projects/pdfminer.six/tools/pdf2txt.py", line 317, in main
    outfp = extract_text(**vars(parsed_args))
  File "/home/pieter/projects/pdfminer.six/tools/pdf2txt.py", line 65, in extract_text
    pdfminer.high_level.extract_text_to_fp(fp, **locals())
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^
  File "/home/pieter/projects/pdfminer.six/pdfminer/high_level.py", line 133, in extract_text_to_fp
    for page in PDFPage.get_pages(
                ~~~~~~~~~~~~~~~~~^
        inf,
        ^^^^
    ...<3 lines>...
        caching=not disable_caching,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ):
    ^
  File "/home/pieter/projects/pdfminer.six/pdfminer/pdfpage.py", line 189, in get_pages
    for pageno, page in enumerate(cls.create_pages(doc)):
                        ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/pieter/projects/pdfminer.six/pdfminer/pdfpage.py", line 146, in create_pages
    yield cls(document, objid, tree, next(page_labels))
          ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/pieter/projects/pdfminer.six/pdfminer/pdfpage.py", line 71, in __init__
    resolve1(mediabox_param) for mediabox_param in self.attrs["MediaBox"]
                                                   ~~~~~~~~~~^^^^^^^^^^^^
TypeError: 'PDFObjRef' object is not iterable

[pietermarsman]

Closed by #1081

[pietermarsman]

Release in v20250324