* prevent scrolling the page when using arrow keys in
* update changelog
* bump prettier
Does GitHub Actions have an incorrect cache somehow?
* use Active LTS in CI
* convert dialog in playground to use Dialog.Panel
* convert `tabs-in-dialog` example to use `Dialog.Panel`
* add scrollable dialog example to the playground
* simplify `outside click` behaviour
Here is a little story. We used to use the `click` event listener on the
window to try and detect whether we clicked outside of the main area we
are working in.
This all worked fine, until we got a bug report that it didn't work
properly on Mobile, especially iOS. After a bit of debugging we switched
this behaviour to use `pointerdown` instead of the `click` event
listener. Worked great! Maybe...
The reason the `click` didn't work was because of another bug fix. In
React if you render a `<form><Dialog></form>` and your `Dialog` contains
a button without a type, (or an input where you press enter) then the
form would submit... even though we portalled the `Dialog` to a
different location, but it bubbled the event up via the SyntethicEvent
System. To fix this, we've added a "simple" `onClick(e) { e.stopPropagation() }`
to make sure that click events didn't leak out.
Alright no worries, but, now that we switched to `pointerdown` we got
another bug report that it didn't work on older iOS devices. Fine, let's
add a `mousedown` next to the `pointerdown` event. Now this works all
great! Maybe...
This doesn't work quite as we expected because it could happen that both
events fire and then the `onClose` of the Dialog component would fire
twice. In fact, there is an open issue about this: #1490 at the time of
writing this commit message.
We tried to only call the close function once by checking if those
events happen within the same "tick", which is not always the case...
Alright, let's ignore that issue for a second, there is another issue
that popped up... If you have a Dialog that is scrollable (because it is
greater than the current viewport) then a wild scrollbar appears (what a
weird Pokémon). The moment you try to click the scrollbar or drag it the
Dialog closes. What in the world...?
Well... turns out that `pointerdown` gets fired if you happen to "click"
(or touch) on the scrollbar. A click event does not get fired. No
worries we can fix this! Maybe...
(Narrator: ... nope ...)
One thing we can try is to measure the scrollbar width, and if you
happen to click near the edge then we ignore this click. You can think
of it like `let safeArea = viewportWidth - scrollBarWidth`. Everything
works great now! Maybe...
Well, let me tell you about macOS and "floating" scrollbars... you can't
measure those... AAAAAAAARGHHHH
Alright, scratch that, let's add an invisible 20px gap all around the
viewport without measuring as a safe area. Nobody will click in the 20px
gap, right, right?! Everything works great now! Maybe...
Mobile devices, yep, Dialogs are used there as well and usually there is
not a lot of room around those Dialogs so you almost always hit the
"safe area". Should we now try and detect the device people are
using...?
/me takes a deep breath...
Inhales... Exhales...
Alright, time to start thinking again... The outside click with a
"simple" click worked on Menu and Listbox not on the Dialog so this
should be enough right?
WAIT A MINUTE
Remember this piece of code from earlier:
```js
onClick(event) {
event.stopPropagation()
}
```
The click event never ever reaches the `window` so we can't detect the
click outside...
Let's move that code to the `Dialog.Panel` instead of on the `Dialog`
itself, this will make sure that we stop the click event from leaking
if you happen to nest a Dialog in a form and have a submitable
button/input in the `Dialog.Panel`. But if you click outside of the
`Dialog.Panel` the "click" event will bubble to the `window` so that we
can detect a click and check whether it was outside or not.
Time to start cleaning:
- ☑️ Remove all the scrollbar measuring code...
- Closing works on mobile now, no more safe area hack
- ☑️ Remove the pointerdown & mousedown event
- Outside click doesn't fire twice anymore
- ☑️ Use a "simple" click event listener
- We can click the scrollbar and the browser ignores it for us
All issues have been fixed! (Until the next one of course...)
* ensure a `Dialog.Panel` exists
* cleanup unnecessary code
* use capture phase for outside click behaviour
* further improve outside click
We added event.preventDefault() & event.defaultPrevented checks to make
sure that we only handle 1 layer at a time.
E.g.:
```js
<Dialog>
<Menu>
<Menu.Button>Button</Menu.Button>
<Menu.Items>...</Menu.Items>
</Menu>
</Dialog>
```
If you open the Dialog, then open the Menu, pressing `Escape` will close
the Menu but not the Dialog, pressing `Escape` again will close the
Dialog.
Now this is also applied to the outside click behaviour.
If you open the Dialog, then open the Menu, clicking outside will close
the Menu but not the Dialog, outside again will close the Dialog.
* add explicit `enabled` value to the `useOutsideClick` hook
* ensure outside click properly works with Poratl components
Usually this works out of the box, however our Portal components will
render inside the Dialog component "root" to ensure that it is inside
the non-inert tree and is inside the Dialog visually.
This means that the Portal is not in a separate container and
technically outside of the `Dialog.Panel` which means that it will close
when you click on a non-interactive item inside that Portal...
This fixes that and allows all Portal components.
* update changelog
* fix incorrect transitionend/transitioncancel events
Due to bubbling, the `Transition` component also "finished" when you had
children that uses `transition-colors` for example.
This commit ensures that we only care about transition events related to
the actual DOM node that we defined the transitions on...
* update changelog
* fix `slot` state of `RadioGroup` component
The `useEvent` is 1 tick too late (due to the update of the callback
happening in useEffect). This isn't a problem for event listeners, but
it is for functions that need to run "now".
We can change the `useLatestValue` hook to do something like:
```diff
export function useLatestValue<T>(value: T) {
let cache = useRef(value)
- useIsoMorphicEffect(() => {
- cache.current = value
- }, [value])
+ cache.current = value
return cache
}
```
But then we are mutating our refs in render which isn't ideal.
* update changelog
* add test to verify that the correct slot data is exposed
We had an issue where an open Dialog got hidden by css didn't properly
unmount because the Transition never "finished". We fixed this by
checking if the node was hidden by using `getBoundingClientRect`.
Today I learned that just *reading* those values (aka call
`node.getBoundingClientRect()`) it for whatever reason completely stops
the transition. This causes the enter transitions to completely stop
working.
Instead, we move this code so that we only check the existence of the
Node when we try to transition out because this means that the Node is
definitely there, just have to check its bounding rect.
* splitup CHANGELOG.md file
Scope each changelog per package
* simplify CHANGELOG.md files
We don't need to scope them anymore, they are already scoped.
* remove leftover code
This code existed before we had the option to make the first option the
"active" one.
This also contains a bug in the React code where pressing "ArrowDown" in
a closed Combobox opens the combobox and goes to the second item instead
of the first option.
* update changelog
* add `@headlessui/tailwindcss` plugin
* expose `data-headlessui-state="..."` data attribute
All components that expose boolean props in their render prop / v-slot
will receive a `data-headlessui-state="..."` attribute.
If it exposes boolean values but all are false, then there will be an
empty `data-headlessui-state=""`. If the current component is rendering
a `Fragment` then we don't expose those attributes.
* use tailwindcss in `playground-react` and `playground-vue`
We were using the CDN, but now that we have the
`@headlessui/tailwindcss` plugin, it's a bit easier to configure it
natively and import the plugin.
* ensure to build the `@headlessui/tailwindcss` package before starting the playground
* refactor `listbox` example to use the @headlessui/tailwindcss plugin
* update changelog
* bump Tailwind CSS to latest insiders version
* correctly generate types
* type `tailwind.config.js` files for playgrounds
* add todo for when `:has()` is available
In the cleanup PR, we added the `Data` and `Actions` type, but we
already had a `Actions` type so had to rename it to something. Chose
`Command` but this is now inconsistent with the rest of the codebase.
Instead, let's revert that change and use these shorthands:
- `Data` -> `_Data`
- `Actions` -> `_Actions`
- `Commands` -> `Actions`
- `CommandTypes` -> `ActionTypes`
The `_` prefix is a little bit strange, but it is a private type and not
exposed so fine for now.
* sort React imports
* improve type signature of the `useEvent` hook
* use more correct `useIsoMorphicEffect` check in `useEvent`
* refactor `useCallback` to cleaner `useEvent`
* convert `const` to `let`
Just for consistency..
* cleanup `Tabs` code
Created explicit functions that can be called from child components
instead of calling `dispatch` directly. Introduced a `useData` and
`useActions` hook to make child components easier.
The seperation of `useData` allows us to pass down props directly
instead of going via the `useReducer` hook and dispatching actions to
make values up to date.
* cleanup `Combobox` code
* cleanup `RadioGroup` code
* make the ref optional in the `Popover` component
We "required" the prop to calculate the `ownerDocument`. But if you
don't provide a ref, then we will use the `Popover.Button` to calculate
it. If that's not defined, then we can fallback to the default
`document`.
* update changelog
* improve scroll lock, scrollbarWidth
The idea is as follow:
If you currently have a scrollbar, and you open a Dialog then we enable
a "Scroll lock" so that you can't scroll in the background behind the
modal. We can achieve this by adding a `overflow: hidden;` to the
`html`.
The issue is that by doing this, we lose the scrollbar and therefore the
page will jump to right because now there is a bit more room.
To account for this, we set a `padding-right` on the `html` of the
scrollbarWidth in pixels. This counteracts the visual jump you would
see.
The issue with this approach is that there could *still* be a scrollbar
once we add the `overflow: hidden`. This can happen if you use new css
features like the `scrollbar-gutter: stable;`.
To take this into account, we will measure the scrollbar again after we
set the `overflow: hidden`. Now we will only apply that counteracting
offset if there would actually be a jump by measuring the before and
after widths and applying the diff if there is one.
* update changelog
* improve `Popover` keyboard usage
Use `TabSentinel` instead of intercepting the `Tab` keydown events.
* use Buttons in Popover example
* update changelog
* refactor `VisuallyHidden` to `Hidden` component
This new component will also make sure that it is visually hidden to
sighted users. However, it contains a few more features that are going
to be useful in other places as well. These features include:
1. Make visually hidden to sighted users (default)
2. Hide from assistive technology via `features={Features.Hidden}`
(will add `display: none;`)
3. Hide from assistive technology but make the element focusable via
`features={Features.Focusable}` (will add `aria-hidden="true"`)
* add `useEvent` hook
This will behave the same (roughly) as the new to be released `useEvent`
hook in React 18.X
This hook allows you to have a stable function that can "see" the latest
data it is using. We already had this concept using:
```js
let handleX = useLatestValue(() => {
// ...
})
```
But this returned a stable ref so you had to call `handleX.current()`.
This new hook is a bit nicer to work with but doesn't change much in the
end.
* add `useTabDirection` hook
This keeps track of the direction people are tabbing in. This returns a
ref so no re-renders happen because of this hook.
* add `useWatch` hook
This is similar to the `useEffect` hook, but only executes if values are
_actually_ changing... 😒
* add `microTask` util
* refactor `useFocusTrap` hook to `FocusTrap` component
Using a component directly allows us to simplify the focus trap logic
itself. Instead of intercepting the <kbd>Tab</kbd> keydown event and
figuring out the correct element to focus, we will now add 2 "guard"
buttons (hence why we require a component now). These buttons will
receive focus and if they do, redirect the focus to the first/last
element inside the focus trap.
The sweet part is that all the tabs in between those buttons will now be
handled natively by the browser. No need to find the first non disabled,
non hidden with correct tabIndex element!
* refactor the `Dialog` component to use the `FocusTrap` component
Also added a hidden button so that we know the correct "main" tree of
the application. Before this we were assuming the previous active
element which will still be correct in most cases but we don't have
access to that anymore since the logic is encapsulated inside the
FocusTrap component.
* ensure `<Portal />` properly cleans up
We make sure that the Portal is cleaning up its `element` properly.
We also make sure to call the `target.appendChild(element)`
conditionally because I ran into a super annoying bug where a focused
element got blurred because I believe that this re-mounts the element
instead of 'moving' it or just ignoring it, if it already is in the
correct spot.
* refactor: use `useEvent` instead of `useLatestValue`
Not really necessary, just cleaner.
* update changelog
* ignore `Escape` when event got prevented
Some external libraries only use `event.preventDefault()` and not
`event.stopPropagation()`. This means that the Dialog can still receive
an `Escape` keydown event which closes the Dialog.
We can also think about the `Escape` behaviour inside the modal as the
"default behaviour" once the Dialog is open. Therefore, we can also
check the `event.defaultPrevented` and ignore this event when this is
the case.
* update changelog
* cleanup README files
* ignore flakey tests for now
There are a handful of tests that are pretty flakey and fail every once
in a while on CI, I don't want to remove them yet, but rather ignore
them for now.
I am going to experiment with using Playwright/Puppeteer to use a real
browser instead.
* add explicit `multiple` prop to the `Combobox`
This allows you to set the value to a **tuple** in `single-value` mode,
which was not possible before the `multiple` prop was introduced,
because then it resulted in `multi-value` mode instead of `single-value`
mode.
* add explicit `multiple` prop to the `Listbox`
This allows you to set the value to a **tuple** in `single-value` mode,
which was not possible before the `multiple` prop was introduced,
because then it resulted in `multi-value` mode instead of `single-value`
mode.
* update changelog
* update playground to use `multiple` prop
* bump dev dependencies to React 18
* setup Jest to include `IS_REACT_ACT_ENVIRONMENT`
* prefer `useId` from React 18 if it exists
In React 16 & 17, where `useId` doesn't exist, we will fallback to our
implementation we have been using up until now.
The `useId` exposed by React 18, ensures stable references even in SSR
environments.
* update expected events
React 18 now uses the proper events:
- `blur` -> `focusout`
- `focus` -> `focusin`
* ensure to wait a bit longer
This is a bit unfortunate, but since React 18 now does an extra
unmount/remount in `StrictMode` to ensure that your code is
ConcurrentMode ready, it takes a bit longer to settle what the DOM sees.
That said, this is a temporary "hack". We are going to experiment with
using tools like Puppeteer/Playwright to run our tests in an actual
browser instead to eliminate all the weird details that we have to keep
in mind.
* prefer `.focus()` over `fireEvent.focus(el)`
* abstract `microTask` polyfill code
* prefer our `focus(el)` function over `el.focus()`
Internally we would still use `el.focus()`, but this allows us to have
more control over that `focus` function.
* add React 18 to the React Playground
* improve hooks for React 18
- Improving the cleanup of useEffect hooks
- useIsoMorphicEffect instead of normal useEffect, so that we can use
useLayoutEffect to be a bit quicker.
* improve disposables
- This allows us to add event listeners on a node, and get automatic
cleanup once `dispose` gets called.
- We also return all the `d.add` calls, so that we can cleanup specific
parts only instead of everything or nothing.
* reimplement the Transition component to be React 18 ready
* wait an additional frame for everything to settle
* update playground examples
* suppressConsoleLogs for RadioGroup components
* update changelog
* keep the `to` classes for a smoother transition
In the next transition we will remove _all_ classes provided and re-add
the once we need.
---
Some extra special thanks:
- Thanks @silvenon for your initial work on the `transition` events in #926
- Thanks @thecrypticace for doing late-night debugging sessions
Co-authored-by: =?UTF-8?q?Matija=20Marohni=C4=87?= <matija.marohnic@gmail.com>
Co-authored-by: Jordan Pittman <jordan@cryptica.me>
* implement `backspace` behaviour in tests
* add `Delete` Key
* implement `nullable` mode on Combobox in single value mode
If you pass a `nullable` prop to the Combobox, then it's possible to
unset the Combobox value by setting it to `null`.
This is triggered by removing all text from the input which will reset
the value itself as well.
* update changelog
Robin Malfait