feat: continue single-line comments #8664
Conversation
…n1991/helix into feature/continue-comment
seanchen1991
changed the title
| /// | ||
| /// Return None otherwise. | ||
| pub fn get_comment_token<'a>(line: &RopeSlice, tokens: &'a [String]) -> Option<&'a str> { | ||
| // TODO: don't continue shebangs |
There was a problem hiding this comment.
This TODO was carried over from #1937, but I'm not sure what it means.
There was a problem hiding this comment.
About shebangs in general: https://en.wikipedia.org/wiki/Shebang_(Unix)
We shouldn't continue shebangs as if they were regular comments since they only take up one line
|
#4718 also adds support for multiple tokens. We will most likely merge that PR before we can give this a proper look. If you're resubmitting work from another PR it's customary to give the original author credit, either by re-using their commits without changes or by giving them co-author on the commits where you re-use their code: https://docs.github.com/en/pull-requests/committing-changes-to-your-project/creating-and-editing-commits/creating-a-commit-with-multiple-authors |
Is there a way to apply these changes to commits retroactively? |
|
Rebasing / ammending old commits can be complicated though. We could instead add a co-authored-by tag at merge-time if we squash + merge |
Sounds good to me 🙂 |
|
@seanchen1991 just a little reminder that the toogle-block-comments PR is merged if you haven't noticed it yet :) |
|
This change seems to be quite far along, what additional work is necessary to get it merged? edit: ah i see its continued in #10996 |
|
Closing in favor of #10996 |
Closes: #1730
This is a re-attempt at implementing continue-comment functionality, following the earlier attempt in #1937; I lifted most of @antoyo's logic from his initial PR.
This PR only implements continuation of single-line comments; block comments are left as a TODO. Additionally, I attempted to implement proper indentation in new comment lines as per this comment, but was finding it really finicky to get right, and so I'm opening this PR without proper handling of indentation in continued comments for now. I'll try to open a subsequent PR that adds that logic at a later time.
The functionality is behind a
continue-commentsconfiguration option that defaults to enabled. Additionally, acomment-tokensfield was added for each language inlanguages.tomlthat accepts an array of strings in order to specify which comment tokens to look for. I think there's two ways to handle this: introduce the newcomment-tokensfield and deprecate the oldcomment-tokenfield, or implement some logic that parses a string of comma-separated tokens in order to not introduce a breaking change. Which would be the preferred route here?