local variable 'cm' referenced before assignment

[thelazydogsback]

Trying to extract text from page.
Tested in Win11 & Linux container.
pypdf==4.2.0, crypt_provider=('cryptography', '42.0.5'), PIL=none

Traceback

File "/usr/local/lib/python3.10/site-packages/pypdf/_page.py", line 2052, in extract_text
    return self._layout_mode_text(
  File "/usr/local/lib/python3.10/site-packages/pypdf/_page.py", line 1950, in _layout_mode_text
    fonts = self._layout_mode_fonts()
  File "/usr/local/lib/python3.10/site-packages/pypdf/_page.py", line 1902, in _layout_mode_fonts
    *cmap, font_dict_obj = build_char_map(font_name, 200.0, self)
  File "/usr/local/lib/python3.10/site-packages/pypdf/_cmap.py", line 33, in build_char_map
    font_subtype, font_halfspace, font_encoding, font_map = build_char_map_from_dict(
  File "/usr/local/lib/python3.10/site-packages/pypdf/_cmap.py", line 58, in build_char_map_from_dict
    map_dict, space_code, int_entry = parse_to_unicode(ft, space_code)
  File "/usr/local/lib/python3.10/site-packages/pypdf/_cmap.py", line 235, in parse_to_unicode
    cm = prepare_cm(ft)
  File "/usr/local/lib/python3.10/site-packages/pypdf/_cmap.py", line 260, in prepare_cm
    if isinstance(cm, str):
UnboundLocalError: local variable 'cm' referenced before assignment

[pubpub-zz]

please provide code and input file

[pubpub-zz]

@thelazydogsback please update the issue with code and input file, else we will have to close the issue as "can't reproduce"

[pubpub-zz]

@thelazydogsback
Please update the issue with code and input file.
Else, the issue will be closed as can not be reproduced

[pubpub-zz]

I close this dead issue

[bazinga014]

I encountered the same issue! Here is the code and the PDF file.

for idx, page in enumerate(PdfReader(pdf_path).pages):
        page_content = ""
        text = page.extract_text() # UnboundLocalError: local variable 'cm' referenced before assignment

and there is some infomation I can provide when I running code: Ignoring wrong pointing object 1758 0 (offset 12558676) Ignoring wrong pointing object 1759 0 (offset 12561201) Object 548 0 not defined. Object 557 0 not defined. Object 562 0 not defined. [Uploading train_a.pdf…]()

[pubpub-zz]

@bazinga014
the file seems to be missing.

[bazinga014]

train_a.pdf

[bazinga014]

@pubpub-zz I just uploaded the file in the issue

[pubpub-zz]

@bazinga014
at first glanc,e, looking at the file the errors opening it with pypdf and confirmed with Acrobat Reader, I can observe some errors being reported this could explain the issue.

[bazinga014]

@pubpub-zz Yes, I can open it normally with Chrome's built-in PDF parser, but there are errors when opening it with Acrobat Reader. So how should I handle this situation? How can I parse it with pypdf?

[stefan6419846]

Apparently, opening with pypdf is already possible - there just is some issue with the text extraction if this is about the undefined cm variable. I am going to re-open this issue for now to evaluate if there is a suitable solution for this case.

[pubpub-zz]

@bazinga014

in _cmap.py
can you try this patch :

def prepare_cm(ft: DictionaryObject) -> bytes:
    tu = ft["/ToUnicode"]
    cm: bytes
    if isinstance(tu, StreamObject):
        cm = b_(cast(DecodedStreamObject, ft["/ToUnicode"]).get_data())
    elif (tu is None) or (isinstance(tu, str) and tu.startswith("/Identity")):             ###### <- line to be replaced
        # the full range 0000-FFFF will be processed
        cm = b"beginbfrange\n<0000> <0001> <0000>\nendbfrange"

and confirm the output is ok

[bazinga014]

@pubpub-zz OK! I will try it! thanks a lot!

[thelazydogsback]

Thanks @bazinga014 for following up -- I was unable to provide the file because it is a customer file that I was unable to share.

[pubpub-zz]

@thelazydogsback can you also try the patch ?

[naktinis]

@pubpub-zz also had the same issue (with a customer's private file) and the patch seems to help!