Shahriar Heidrich activity https://gitlab.com/smheidrich 2025-06-07T13:31:57Z tag:gitlab.com,2025-06-04:4324179454 Shahriar Heidrich pushed new project tag v0.7.3 at Shahriar Heidrich / full-apply 2025-06-04T19:49:44Z smheidrich Shahriar Heidrich

Shahriar Heidrich (ddc9a235) at 04 Jun 19:49

Bump version to 0.7.3

tag:gitlab.com,2025-06-04:4324179045 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / full-apply 2025-06-04T19:49:33Z smheidrich Shahriar Heidrich

Shahriar Heidrich (ddc9a235) at 04 Jun 19:49

Bump version to 0.7.3

... and 1 more commit

tag:gitlab.com,2025-02-05:4006651227 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / conv-ng 2025-02-05T23:41:14Z smheidrich Shahriar Heidrich

Shahriar Heidrich (357ed3dc) at 05 Feb 23:41

Add unmark command

... and 3 more commits

tag:gitlab.com,2025-02-03:4000027154 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / conv-ng 2025-02-03T22:39:04Z smheidrich Shahriar Heidrich

Shahriar Heidrich (57e5c328) at 03 Feb 22:39

Fix ls . bug (raised exception on any mark)

... and 1 more commit

tag:gitlab.com,2025-02-03:3999940525 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / conv-ng 2025-02-03T21:39:31Z smheidrich Shahriar Heidrich

Shahriar Heidrich (d6f0ec97) at 03 Feb 21:39

Fix fs-aware mark typing in manager

tag:gitlab.com,2025-02-02:3997121460 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / conv-ng 2025-02-02T23:46:40Z smheidrich Shahriar Heidrich

Shahriar Heidrich (cee6af42) at 02 Feb 23:46

Update TODO

... and 4 more commits

tag:gitlab.com,2025-01-26:3979972990 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / conv-ng 2025-01-26T00:48:25Z smheidrich Shahriar Heidrich

Shahriar Heidrich (fd4c7c5b) at 26 Jan 00:48

Fix export root bug

... and 2 more commits

tag:gitlab.com,2025-01-04:3929396828 Shahriar Heidrich pushed to project branch main at Shahriar Heidrich / recursive-mtime-cache 2025-01-04T20:49:34Z smheidrich Shahriar Heidrich

Shahriar Heidrich (a1046af6) at 04 Jan 20:49

Comment out dependency on fs-reduction-tree-cache

tag:gitlab.com,2024-12-24:3913780461 Shahriar Heidrich pushed to project branch fix-most-doc-warnings at Shahriar Heidrich / KiCad Footprint Generator 2024-12-24T00:52:37Z smheidrich Shahriar Heidrich

Shahriar Heidrich (f9e134bf) at 24 Dec 00:52

Fix most warnings during doc build

... and 1689 more commits

tag:gitlab.com,2024-12-24:3913780032 Shahriar Heidrich commented on merge request !1356 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-24T00:51:45Z smheidrich Shahriar Heidrich

After thinking about it some more, my vote is for just leaving the docstrings themselves as they are for now by enabling sphinx.ext.napoleon, because maybe good reasons to make them consistent in one direction or the other will come up in the future.

Completely agree with everything else you wrote.

tag:gitlab.com,2024-12-16:3894983301 Shahriar Heidrich commented on merge request !1356 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-16T02:36:14Z smheidrich Shahriar Heidrich

I had a look again as well and what I get is:

  • 22 Google-style (Args:) docstrings.
  • 38 docstrings with a :params: (plural!) field list, which I can't find documented anywhere in Sphinx's docs, followed by a standard bulleted list for each param - almost definitely a non-standard way of documenting parameters, although valid ReST.
  • 69 docstrings with ReST-style :param parameter_name: docstrings.
    • But a lot of these deviate from the standard in how they document types, e.g. :param low_limit: (``int``) -- ....
      • According to Sphinx docs the type should be either documented as :param int low_limit: ... or :param low_limit: ... followed by :type low_limit: int.
      • And of course, in modern Python, it's usually better to put types into type hints and leave them out of docstrings.

So I guess eventually, a significant fraction of the non-Google-style docstrings should be updated anyway, no matter which style is preferred, which means making them consistent in any direction shouldn't be that different in terms of effort.

But if you guys prefer ReST-style docstrings regardless of the existing ones and would e.g. use this style for a brand new project as well, it still makes sense to update them in that direction, of course.

tag:gitlab.com,2024-12-15:3894202794 Shahriar Heidrich opened issue #699: ReadTheDocs build hasn't run in 4 years at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-15T02:28:49Z smheidrich Shahriar Heidrich

As far as I can tell, this project's online documentation is meant to be built automatically by ReadTheDocs. But looking at its project page over there, it seems the current version is the one from 4 years ago. And indeed, new additions to the docs like the one from #687 aren't visible in the online docs yet.

Maybe some configuration needs to be updated? Instructions for letting RTD configure the GitLab project on behalf of the user can be found on the How to connect your Read the Docs account to your Git provider page of the RTD docs, but if that doesn't work, it can also be done by manually adding a webhook to the GitLab project.

tag:gitlab.com,2024-12-15:3894165503 Shahriar Heidrich commented on merge request !1356 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-15T00:59:22Z smheidrich Shahriar Heidrich

Not sure if this was the right choice: Most of the docstrings use ReST syntax like :param ...: ..., but at least a handful use Google docstring syntax (Args:).

The Napoleon extension lets Sphinx parse the latter correctly, but maybe it would be better to turn them all into ReST-style docstrings instead for consistency?

I think most people prefer the Google syntax, but making it consistent the other way round by turning all ReST-style docstrings into Google-style ones would be way too much effort.

tag:gitlab.com,2024-12-15:3894164107 Shahriar Heidrich opened merge request !1356: Fix most warnings during doc build at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-15T00:55:24Z smheidrich Shahriar Heidrich

The docs build currently outputs quite a few warnings. This MR fixes the ones that are easy and sensible (IMO) to fix.

tag:gitlab.com,2024-12-15:3894161628 Shahriar Heidrich pushed new project branch fix-most-doc-warnings at Shahriar Heidrich / KiCad Footprint Generator 2024-12-15T00:49:47Z smheidrich Shahriar Heidrich

Shahriar Heidrich (7afe42e6) at 15 Dec 00:49

Fix most warnings during doc build

... and 16 more commits

tag:gitlab.com,2024-12-13:3890605874 Shahriar Heidrich commented on merge request !1344 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-13T02:53:20Z smheidrich Shahriar Heidrich

Since this now installs a package named scripts into the Python library dir, maybe it has become more important to point out to people that they should be doing this in a venv (of course that's a good idea in general but now even moreso due to the higher likelihood of a name collision with something else given the generic name).

tag:gitlab.com,2024-12-12:3890390510 Shahriar Heidrich commented on merge request !1344 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-12T23:22:56Z smheidrich Shahriar Heidrich

@smheidrich you inspired this, would you like to sanity check it?

Other than the comments i left, looks good to me 👍

tag:gitlab.com,2024-12-12:3890380226 Shahriar Heidrich commented on merge request !1344 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-12T23:14:14Z smheidrich Shahriar Heidrich

Same here as in the README comment regarding that it should be fine to reduce this to just pip install .[dev].

tag:gitlab.com,2024-12-12:3890378371 Shahriar Heidrich commented on merge request !1344 at KiCad / KiCad Libraries / KiCad Footprint and 3D model Generator 2024-12-12T23:12:29Z smheidrich Shahriar Heidrich

I think for this repo it wouldn't even have been strictly necessary to replace setup.py with pyproject.toml, but it's probably a good idea anyway 👍