New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Tree sitter parsers #35461
base: master
Are you sure you want to change the base?
Tree sitter parsers #35461
Conversation
Pull Request assignmentSubmitter: @msva dev-libs/tree-sitter-lua: @gentoo/proxy-maint (new package) Linked bugsBugs linked: 922146 In order to force reassignment and/or bug reference scan, please append Docs: Code of Conduct ● Copyright policy (expl.) ● Devmanual ● GitHub PRs ● Proxy-maint guide |
Thanks! Think that these new grammars need |
@MatthewGentoo done |
Please, leave a note, if all looks good for you, and I'll push it in gentoo myself (this PR is for review purposes only, I have push access myself, if any) |
Pull request CI reportReport generated at: 2024-02-21 11:08 UTC There are existing issues already. Please look into the report to make sure none of them affect the packages in question: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The packages look fine, but some of the tests in tree-sitter-markdown-inline
are failing.
8 failures:
expected / actual
1. Basic Wiki-link parsing.:
(inline
(shortcut_link
(link_text)))
(wiki_link
(link_destination)))
2. Wiki-link to a file:
(inline
(shortcut_link
(link_text)))
(wiki_link
(link_destination)))
3. Wiki-link to a heading in a note:
(inline
(shortcut_link
(link_text)))
(wiki_link
(link_destination)))
4. Wiki-link with title:
(inline
(shortcut_link
(wiki_link
(link_destination)
(link_text)))
5. Wiki-link version of Example 556:
(inline
(shortcut_link
(link_text)))
(wiki_link
(link_destination)))
6. Example 324 - https://github.github.com/gfm/#example-324:
(inline)
(inline
(tag)
(tag))
7. Example 638 - https://github.github.com/gfm/#example-638:
(inline)
(inline
(tag))
8. Tags are working:
(inline)
(inline
(tag)
(tag)
(tag))
I have dev-libs/tree-sitter-0.20.9-r1.
|
||
LICENSE="MIT" | ||
SLOT="0" | ||
KEYWORDS="~amd64 ~arm ~arm64 ~ppc ~ppc64" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please don't add them with other keywords: https://devmanual.gentoo.org/keywording/index.html#keywording-new-packages.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I did tested that it compiles fine on all of that arches.
Although, I only tested compilation itself, and didn't tested them in co-operation with "clients".
And also I tested ppc and ppc64 on qemu targets (arm and arm64 on "real" boards (orange pi zero
and pine a64+
)).
Isn't it enough?
If so - okay, I will remove anything except ~amd64
🤷
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No, it's not enough. We do keywording via arch teams via bugs. That's how it always works. You also run the test suite for them, not just building.
You have to ask for permission otherwise.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can I ask where you got the impression this was the done thing? This is covered in the quizzes and in the devmanual quite clearly?
Unfortunately, I have no idea what can I do with that 🤷 Can you give some advice, please? Maybe just drop markdown-inline grammar? |
First, you check if you can reproduce it. If you can, you poke at it, and report it upstream? Did you not run the tests for it before opening the PR? |
Yes, it reproduces here
I did. But this grammar passes many other tests, and can still be useful. Althouth, as it is not a neovim dependency (unlike non-inline markdown grammar), I not insisting on adding it to the gentoo repo until it will pass all tests, and I okay with dropping it from PR 🤷 |
...? Did you hit the failure, then just not mention it at all? That's bizarre? |
Well, it is really my fault, that I forgot to mention that (altough, that is the reason I created PR for: to other co-maintainers re-checked the things I forgot about). As "bad excuse" (that actually not excusinng, but sheds a light on reasons why did it happened) for that, I can mention that I'm working on this PR for about a month already (by minor steps when getting some spare hours and not-too-tired after work), so at current point I just forgot about this "speciality" of this not-so-necessary package (that was added mostly because it is co-packaged with not-inline version), and remembered that fact only when @MatthewGentoo mentioned that. And when I hit that with personal testing, IIRC (that was some time ago), I was guided by the logic like (I mean, "my thoughts was"):
So, yes, it is definitelly my fault in forgetting about it, but I don't agree with "bizzare". It is just "artifact" of a process how I worked on this, which was hopefuly catched by a process specifically designed for catching things like this 🤷 Also, as I said, this package (unlike others in the PR) is not a dependency of neovim, so it is not strictly needed. === Well, I already dropped it in local copy of branch, and will force-push commits without it 🤷 (I anyway planned to force-push as there is another change: |
also, I have dropped arches other than |
6967658
to
6021e8e
Compare
@MatthewGentoo I've re-pushed, can you, please, re-review? Also |
Bug: https://bugs.gentoo.org/922146 Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
Bug: https://bugs.gentoo.org/922146 Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
Bug: https://bugs.gentoo.org/922146 Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
Bug: https://bugs.gentoo.org/922146 Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
Bug: https://bugs.gentoo.org/922146 Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
dev-util/tree-sitter-cli is a conditional test dep for packages inheriting tree-sitter-grammar.eclass, but is written in rust, so mask it on rustless profiles. Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
Signed-off-by: Vadim Misbakh-Soloviov <mva@gentoo.org>
6021e8e
to
1893677
Compare
(re-pushed again because of lost (during previous force-push) "Bug" header in grammar commits) |
Pull request CI reportReport generated at: 2024-02-21 17:43 UTC There are existing issues already. Please look into the report to make sure none of them affect the packages in question: |
by the way, tree-sitter-markdown's upstream said to "try passing ALL_EXTENSIONS=1". So, I think, I'll leave the inline grammar and will not add an ebuild for it for now. |
regarding keywords: No, please follow the policy I cited at #35461 (comment). This package does not get special treatment. You get it merged, then you request (re)keywording via a bug. It is enormous hassle to deviate from that process, especially to handle keywording for things not even in-tree yet. |
In future, you should take care to run the test suites (again before pushing) and to mention any test failures immediately in a comment in the ebuild or similar for further investigation. We can't do much about people forgetting things - that's why it's good to not let some stuff run on for months indeed.
I am not sure why you're using 🤷 so much. It might appear dismissive. |
okay, I will, thanks ========= <offtop>
Well, I totally agree, that it will be good to do it another (better) way, but... As you can see, it wasn't done by anybody else during that time (for more than month since I mentioned that things in a bug on BZ). So I worked on it in "small steps" manner for that month+ (and would be okay if anybody finished this before me). Not sure if a month+ is enough time to make any conclusions, aspecially in that case, but I think, I can assume that others also have "timing issues". So, it looks like there is two possible ways:
// imho </offtop> |
Btw, @thesamesam, can you, please re-review, and approve the merge if all is okay? Github shows that you have requested changes, but says that it can't find corresponding things because of force-push |
By the way, I looked on one of my arm-boards (the one with ubuntu), and found that neovim is packaged with
So... Then there is again appears a question: should we package it or not. Without, for example, Lua grammar, Without Markdown grammars (both) And, if we decide to package (again) inline grammar, would it be okay to regenerate it with tree-sitter-cli (as upstream suggests) before building? It will definitelly make it unavailable (or failing tests) on @MatthewGentoo @sarnex WDYT? |
I think the main thing is you are happy it works with your editor :) Other grammars look good. |
CC: @00-matt @sarnex
New grammars.
Most of them also required in runtime by app-edittors/neovim-9999 (and it seems will be required by next non-patch release too)
Ref: https://github.com/neovim/neovim/blob/master/cmake.deps/deps.txt#L46 and below