Package Details: python-pdfplumber 0.10.4-1

Git Clone URL: https://aur.archlinux.org/python-pdfplumber.git (read-only, click to copy)
Package Base: python-pdfplumber
Description: Plumb a PDF for detailed information about each char, rectangle, line, et cetera.
Upstream URL: https://github.com/jsvine/pdfplumber
Licenses: MIT
Submitter: falko17
Maintainer: falko17
Last Packager: falko17
Votes: 0
Popularity: 0.000000
First Submitted: 2021-07-14 21:20 (UTC)
Last Updated: 2024-02-15 13:54 (UTC)

Latest Comments

martinc commented on 2022-05-10 08:58 (UTC)

Thank you. All working again.

falko17 commented on 2022-05-09 21:52 (UTC)

Thanks again for your comment. As mentioned here or here by the repository owner, the dependency to pdfminer.six will always require a version that will become out-of-date at some point -- even though using a newer version most likely won't cause any trouble, which is why I kept the dependency here >= rather than ==.

This worked until now, but due to some change (I'm not exactly sure where), when paper2remarkable imports pikepdf which imports pkg_resources, this requirement mismatch is caught by the requirement resolver and the whole thing crashes on an Exception.

Since I don't want to require an older version here though, I've now resolved this by loosening the requirement within pdfplumber too, so it should work again (and future updates hopefully shouldn't cause the same errors to happen).

martinc commented on 2022-05-09 16:13 (UTC)

Hi, with everything up-to-date (python-pdfminer.six replaced with python-pdfminer), running p2r barfs with

...
File "/usr/lib/python3.10/site-packages/pkg_resources/__init__.py", line 777, in resolve
    raise DistributionNotFound(req, requirers)
pkg_resources.DistributionNotFound: The 'pdfminer.six==20220319' distribution was not found and is required by pdfplumber

Any idea what might be causing this?

falko17 commented on 2021-12-14 23:21 (UTC) (edited on 2022-05-06 12:39 (UTC) by falko17)

Thanks for your comment! At the very least, pdfplumber doesn't seem to work well with python-pdfminer, so I've replaced the dependency with python-pdfminer.six.

That said, while the 20200517-version of python-pdfminer.six is listed in their requirements.txt, requiring that exact version would make this package much more cumbersome to install — according to this comment of the repository owner, it will probably also be the case that pdfplumber will always require an exact version of pdfminer.six instead of just a minimum version. As said in the same comment, the newest pdfminer.six version will still work in "most" situations though (specifically, comparing test results across the two versions, it seems that converting a PDF to CSV and functionality described in this issue won't work), so I think keeping this dependency as a minimum instead of exact requirement would be preferred in most cases. I've added a short pre_install message notifying the user of the situation.

EDIT: As upstream has now updated their dependency on pdfminer.six to version 20211012, the problem has been resolved (for now). Since python-pdfminer.six has now been moved into community's python-pdfplumber, I've also updated the dependencies to reflect that.

martinc commented on 2021-12-13 15:54 (UTC)

Hi, just installed paper2remarkable (thanks!) and found pdfplumber complaining that it wanted an exact (20200517) version of python-pdfminer.six. A quick google suggests you've already had a conversation with upstream (and I fixed things by installing the required version).

But I can't get pdfplumber to work with python-pdfminer (or a newer version of -six) which the dependencies here suggest should be possible. Is it?