predragnikolic / inlinefold Goto Github PK

With the changes to the Python default package, it is now possible to fold the non-summary parts of a docstring, i.e. everything but the first line. However, when the summary line is on the second line, it means that we have folded regions on both sides of a line that include newlines, which makes for a somewhat whacky experience when navigating the caret within the docstring. See the apng in the following PR for an example:

deathaxe/sublime-packages#2 (comment)

You can reproduce this by using the Python package from the Package's master branch and the following config:

    "inline_fold.rules":
    [
        {
            "fold_selector": "comment.block.documentation.python - punctuation.definition.comment",
        }
    ],

My proposal for this situation would be to add a configuration for context line where folding would not be applied for regions that are within n lines of the caret instead of on the same line as the caret.

Another idea related to this would be to unfold any regions that share a line with the currently active region (i.e. a region that would be folded if the caret wasn't inside it) and/or consider the additional config option here.

I've been using the package as a replacement for Fold Python Docstrings for quite a while now (and also testing the HTML class folding configuration) and I think it's definitely usable in its current form.

What do you think of publishing it? I just stubled upon this package again as I was comparing my installed (in PC configuration) vs actual packages list (on disk).

InlineFold.zip

Not sure if this is even possible but some tools reload files after they are saved (for example, prettier), and when that occurs all the area suddenly fold again which is a disturbing DX. Any ways to improve it or workaround it?

This isn't a huge priority, but I wanted to list it here as food for thought.

I'd like to only hide the class attribute text if it doesn't stand on its own line already, because that indicates that I am having exactly one attribute per line. In order to do that, I could match for \S\s+class="$, but the current preceding_text setting does not support this kind of flexibility.

Example:

                <span class="text-danger">
                    some text be here. the class above should be folded
                </span>

                <span 
                    class="text-danger"
                >
                    more text. here, the class should not be folded
                </span>

So doing the fold command (Ctrl+Shift+[) won't work anymore. It also breaks on non-xml files (JS, TS).

Hello 👋

Feel free to share your "inline_fold.rules" with others!
Add an image so we can see it in action.
Use reactions.

Only write recipes in comments :) - I will hide any unrelated comment