Do not generate dropdown in sidebar. #1771
Conversation
| as well as the list of links to but in a dropdown. | ||
|
|
||
| Returns: | ||
| - HTML str for the navbar |
There was a problem hiding this comment.
I think it would be a simple refactor that might help make this function easier to understand if it returns
- list of HTML str for navbar
- list of HTML str for the dropdown
There was a problem hiding this comment.
I think better than that, the cached function should not return html but the data necessary to build the html. But that require some refactor.
I'll make an issue to suggest doing a refactor for later.
Edit: for completeness this was done as issue #1773
src/pydata_sphinx_theme/theme/pydata_sphinx_theme/sections/sidebar-primary.html
Show resolved
Hide resolved
src/pydata_sphinx_theme/theme/pydata_sphinx_theme/sections/sidebar-primary.html
Outdated
Show resolved
Hide resolved
src/pydata_sphinx_theme/utils.py
Outdated
| @@ -32,7 +32,9 @@ def config_provided_by_user(app: Sphinx, key: str) -> bool: | |||
| return any(key in ii for ii in [app.config.overrides, app.config._raw_config]) | |||
|
|
|||
|
|
|||
| def traverse_or_findall(node: Node, condition: str, **kwargs) -> Iterator[Node]: | |||
| def traverse_or_findall( | |||
| node: Node, condition: Union[Callable, type], **kwargs | |||
There was a problem hiding this comment.
Oops this doesn't seem related to the other changes. Could you move this to a different PR please?
trallard
added
tag: CSS
This should fix pydata#1735, it is complementary to pydata#1564 but with a different approach. Instead of generating the same navbar as the fullpage width one, it generate one with no dropdowns, with this, any css that tries to make downtown looks like normal links in html becomes unnecessary. Co-authored-by: gabalafou <[email protected]>
|
unrelated type change removed and pulled in their own PR. |
|
BTW, every time I push the workflows re-need approval are the workflow settings a bit too agressive ? I though approval were only a one-time thing. |
On top of pydata#1771, should fix pydata#1773 (once fininished) This try to split the function into two parts: one that wrangle the data, one that generate html. The generating the data part should be cacheable, while the other can be made more flexible to be reusable in different part of the theme (typically generating or not dropdowns in the sidebar or navbar).
That's a good question. I don't have access to the workflow settings. @drammock do you? |
|
In the meantime, I do have permission to approve and run workflows for a particular PR, so I went ahead and re-ran the workflows on this PR |
I just gave @Carreau write access to this repo, which should make the CIs run automatically on his PRs in future. Then I remembered that we usually have a process for this --- opening a PR nominating someone to have access, giving folks time to weigh in. Sorry for subverting the process! I can still open that PR if folks want, and revoke access if @Carreau gets vetoed, WDYT? FWIW both @trallard and I know @Carreau personally and I trust him to not do anything dangerous here so if folks want to skip the formal vote that's also OK with me. |
Thanks, much appreciated. and it's ok if you wish to remove me.
I'm not too sure, I'll try to understand pytest-regression and how to select the right things in the sidebar. |
On top of pydata#1771, should fix pydata#1773 (once fininished) This try to split the function into two parts: one that wrangle the data, one that generate html. The generating the data part should be cacheable, while the other can be made more flexible to be reusable in different part of the theme (typically generating or not dropdowns in the sidebar or navbar).
|
Ok, there was no test for the nav of the sidebar, I think I added one in the right place. This should also take care of #1799 |
|
@gabalafou you requested changes here, can you take another look and merge if satisfied? |
On top of pydata#1771, should fix pydata#1773 (once fininished) This try to split the function into two parts: one that wrangle the data, one that generate html. The generating the data part should be cacheable, while the other can be made more flexible to be reusable in different part of the theme (typically generating or not dropdowns in the sidebar or navbar).
|
Thanks ! |
On top of pydata#1771, should fix pydata#1773 (once fininished) This try to split the function into two parts: one that wrangle the data, one that generate html. The generating the data part should be cacheable, while the other can be made more flexible to be reusable in different part of the theme (typically generating or not dropdowns in the sidebar or navbar).
* Do not generate dropdown in sidebar. This should fix pydata#1735, it is complementary to pydata#1564 but with a different approach. Instead of generating the same navbar as the fullpage width one, it generate one with no dropdowns, with this, any css that tries to make downtown looks like normal links in html becomes unnecessary. Co-authored-by: gabalafou <[email protected]> * apply review * Add failing test for dropdown sidebar --------- Co-authored-by: gabalafou <[email protected]>
This should fix #1735, it is complementary to #1564 but with a different approach.
Instead of generating the same navbar as the fullpage width one, it generate one with no dropdowns, with this, any css that tries to make downtown looks like normal links in html becomes unnecessary.