RunOnSave causing an issue when using autosave.

[donk3ylee]

I wrote this RFC on headwinds github page but after seeing the package is not well maintained I figured I would use your fork instead in the hope of getting this issue fixed. My apologies if some of the settings described here are not named the same but I'm sure you will get the idea:

RunOnSave feature is causing an issue when using vscode autosave. The issue is that while I'm typing my tailwind classes, vscode autosaves triggering headwind to reorder. As this reorder applies the cursor does not follow the original classname and I end up writing half of my classname in the center of another, corrupting that classname and leaving the classname I'm trying to declare half written and lost somewhere else in the string.

Could you add another cofiguration setting say headwind.triggerPoint where the user can choose options "runOnSave" or "runOnLineOut". The runOnLineOut would be an option whereby the reordering would happen when the user focuses away from the line or maybe when the cursor focuses away from the class value declaration string (anything within the ""). Or just move the cursor with the reordering of possible, you get the idea.

It's a very handy extension for code readability and therefore productivity. It just becomes unusable when users have vscode autosave enabled and it gets triggered every 5 seconds, corrupting their tailwind classnames.

Thanks,
Donk3yLee.

[Trapfether]

@donk3ylee That sounds very frustrating. Definitely worth fixing. I'll dig into this later this week. My approaches will generally prefer automatic detection so when people install the extension it "just works". So I'll try and detect active editing and disable reordering. My first approach will be detecting if the selection (cursor is a zero width selection) lies within the detected text range to be reordered and skip just that range. Another approach I might try is attempting to stick the cursor to its position relative to class it currently is within / on the end of as you suggested. If I can't make either of those work, I will investigate the addition of another configuration option.

Aside from those, I think that maintaining cursor position is a generally good idea and so I will make sure that when we change the length of className - such as removing duplicates or extra whitespace - we update the cursor position accordingly.

Thank you for taking the time to submit this issue and bring this to my attention!

[donk3ylee]

No worries Andrew thanks for the quick reply. It's a useful plugin. I think if you can stick the cursor to the class name your typing as it is reordered that would be thebest fix like you said. Cheers, Donk3ylee

…

On Tue, 27 Feb 2024, 19:34 Andrew Trefethen, ***@***.***> wrote: @donk3ylee <https://github.com/donk3ylee> That sounds very frustrating. Definitely worth fixing. I'll dig into this later this week. My approaches will generally prefer automatic detection so when people install the extension it "just works". So I'll try and detect active editing and disable reordering. My first approach will be detecting if the selection (cursor is a zero width selection) lies within the detected text range to be reordered and skip just that range. Another approach I might try is attempting to stick the cursor to its position relative to class it currently is within / on the end of as you suggested. If I can't make either of those work, I will investigate the addition of another configuration option. Aside from those, I think that maintaining cursor position is a generally good idea and so I will make sure that when we change the length of className - such as removing duplicates or extra whitespace - we update the cursor position accordingly. Thank you for taking the time to submit this issue and bring this to my attention! — Reply to this email directly, view it on GitHub <#14 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ATWRJY3CWQ3AV7E2QTT3VJLYVYYLPAVCNFSM6AAAAABD4QGBKKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNRXGQ2TQMRUGI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[SantanuDatta]

I have been searching for a solution for so long now, I hope the work is still going on this, and an update soon, thanks