Updated to support daisyui 4.0.0+ version.

[LeekJay]

No description provided.

[gorango]

Hey @LeekJay!

Are you still using this package? Running the /playground, I noticed an infinite loop of requests for hmr:

Related to unocss/issues/2039. Seems that cssnano is the culprit but I'm not sure where it's being invoked.

Update:
On closer inspection, removing the daisy.handler solves the hmr issue, so I'll investigate further.

[gorango]

Wow, really great find, thanks a lot!

P.S. there seems to be another call to parents.reverse() on line 115 but it doesn't seem to trigger the loop..

[LeekJay]

Hey @LeekJay!

Are you still using this package? Running the /playground, I noticed an infinite loop of requests for hmr:

Related to unocss/issues/2039. Seems that cssnano is the culprit but I'm not sure where it's being invoked.

Update: On closer inspection, removing the daisy.handler solves the hmr issue, so I'll investigate further.

Fixed.unocss-preset-component/pr/1

This issue has been identified:

Array.reverse() is an in-place operation, changing the original array.

If this array is in the module's global scope, its state (including the global scope's array state) will be retained when the module is read and cached by Vite.

Then when using Array.reverse(), since this is an in-place operation, the array state will be changed. This will be detected by Vite, which will consider the module to have changed, thus triggering a hot update.

When a hot update is triggered, Vite will not re-read and parse the module file, but will use the cached module state, which has been changed by Array.reverse().

Therefore, after each hot update Array.reverse() will be run again, and the array state changes again, which triggers a hot update, forming an infinite loop.

To fix this problem, we can avoid using Array.reverse() directly in the global scope. Or we can use Array.from(array).reverse() or [...array].reverse(), copy the array before each use, and then use Array.reverse(). This way the original array will not be changed.

[LeekJay]

Wow, really great find, thanks a lot!

P.S. there seems to be another call to parents.reverse() on line 115 but it doesn't seem to trigger the loop..

Wow, really great find, thanks a lot!

P.S. there seems to be another call to parents.reverse() on line 115 but it doesn't seem to trigger the loop..

Yes, maybe unocss already handles that in rules, but preflights doesn't.