Arch Linux User Repository

paperless-ngx-venv 2.13.5-2 doesn't seem to build with Python 3.13.1 because dependencies httptools and uvloop have GCC compilation errors against Python 3.13:

Failed to build httptools uvloop
ERROR: ERROR: Failed to build installable wheels for some pyproject.toml based projects (httptools, uvloop)

Upstream feature request

For httptools:

  In file included from /usr/include/python3.13/longobject.h:107,
                   from /usr/include/python3.13/Python.h:81,
                   from httptools/parser/parser.c:22:
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  httptools/parser/parser.c: In function ‘__Pyx_PyInt_As_int’:
  httptools/parser/parser.c:10134:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  10134 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
        |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  error: command '/usr/bin/gcc' failed with exit code 1
  [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for httptools

For uvloop:

  In file included from /usr/include/python3.13/longobject.h:107,
                   from /usr/include/python3.13/Python.h:81,
                   from uvloop/loop.c:22:
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  uvloop/loop.c: In function ‘__Pyx_PyInt_As_uint64_t’:
  uvloop/loop.c:177765:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  177765 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
         |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  uvloop/loop.c: In function ‘__Pyx_PyInt_As_long’:
  uvloop/loop.c:177999:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  177999 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
         |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  uvloop/loop.c: In function ‘__Pyx_PyInt_As_unsigned_long’:
  uvloop/loop.c:178271:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  178271 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
         |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  uvloop/loop.c: In function ‘__Pyx_PyInt_As_unsigned_int’:
  uvloop/loop.c:178467:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  178467 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
         |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  uvloop/loop.c: In function ‘__Pyx_PyInt_As_size_t’:
  uvloop/loop.c:178739:27: error: too few arguments to function ‘_PyLong_AsByteArray’
  178739 |                 int ret = _PyLong_AsByteArray((PyLongObject *)v,
         |                           ^~~~~~~~~~~~~~~~~~~
  /usr/include/python3.13/cpython/longobject.h:111:17: note: declared here
    111 | PyAPI_FUNC(int) _PyLong_AsByteArray(PyLongObject* v,
        |                 ^~~~~~~~~~~~~~~~~~~
  error: command '/usr/bin/gcc' failed with exit code 1
  [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for uvloop

Yeah, nltk-data is out of date. Even better: it is orphaned, while in extra! :(

After switching from paperless-ngx 2.11.6-1 to paperless-ngx-venv 2.13.5-1, I encountered the following error:

[2024-11-26 17:07:48,861] [WARNING] [paperless.tasks] Classifier error:
**********************************************************************
  Resource #[93mpunkt_tab#[0m not found.
  Please use the NLTK Downloader to obtain the resource:
  #[31m>>> import nltk
  >>> nltk.download('punkt_tab')
  #[0m
  For more information see: https://www.nltk.org/data.html
  Attempted to load #[93mtokenizers/punkt_tab/italian/#[0m
  Searched in:
PosixPath('/usr/share/nltk_data')
**********************************************************************

I resolved the issue by downloading the folder punkt_tab from https://www.nltk.org/nltk_data/ and saving it to /usr/share/nltk_data/tokenizers.

When running this behind a proxy with Header authentikation it seems like i get a ModuleNotFoundError:

journalctl -xeu paperless-webserver results in:

Nov 15 16:41:03 seenas gunicorn[46190]:   File "<frozen importlib._bootstrap>", line 1360, in _find_and_load
Nov 15 16:41:03 seenas gunicorn[46190]:   File "<frozen importlib._bootstrap>", line 1331, in _find_and_load_unlocked
Nov 15 16:41:03 seenas gunicorn[46190]:   File "<frozen importlib._bootstrap>", line 935, in _load_unlocked
Nov 15 16:41:03 seenas gunicorn[46190]:   File "<frozen importlib._bootstrap_external>", line 995, in exec_module
Nov 15 16:41:03 seenas gunicorn[46190]:   File "<frozen importlib._bootstrap>", line 488, in _call_with_frames_removed
Nov 15 16:41:03 seenas gunicorn[46190]:   File "/usr/share/paperless/src/paperless/asgi.py", line 16, in <module>
Nov 15 16:41:03 seenas gunicorn[46190]:     from paperless.urls import websocket_urlpatterns  # noqa: E402
Nov 15 16:41:03 seenas gunicorn[46190]:     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Nov 15 16:41:03 seenas gunicorn[46190]:   File "/usr/share/paperless/src/paperless/urls.py", line 20, in <module>
Nov 15 16:41:03 seenas gunicorn[46190]:     from documents.views import AcknowledgeTasksView
Nov 15 16:41:03 seenas gunicorn[46190]:   File "/usr/share/paperless/src/documents/views.py", line 165, in <module>
Nov 15 16:41:03 seenas gunicorn[46190]:     from paperless_mail.oauth import PaperlessMailOAuth2Manager
Nov 15 16:41:03 seenas gunicorn[46190]:   File "/usr/share/paperless/src/paperless_mail/oauth.py", line 7, in <module>
Nov 15 16:41:03 seenas gunicorn[46190]:     from httpx_oauth.clients.google import GoogleOAuth2
Nov 15 16:41:03 seenas gunicorn[46190]: ModuleNotFoundError: No module named 'httpx_oauth'

A bit confusing to me since it seems like httpx-oauth is explicitly listed in the requirements.txt. Any idea what this could be? I'd love to help you out in triage if needed.

nevermind, my oversight

@shtrophic sorry, but what the heck? Did you just force-push stuff? Why?

I pushed some patches about the warnings caused by whoosh. OK?

I suspect it is issue of whoosh, last release was in 2016 and seems to be abandoned.

Not really... The only possibly breaking changes in the patched PKGBUILD are:

find "$pkgdir/usr/lib/paperless" -type f -exec sed -i "s|$srcdir/venv|/usr/lib/paperless|g" {} +
find "$pkgdir/usr/lib/paperless" -type d -name "__pycache__" | xargs rm -rf

in package().

I just pushed 2 fixes