This PR introduces a new `canonicalizeCandidates` function on the
internal Design System.
The big motivation to moving this to the core `tailwindcss` package is
that we can use this in various places:
- The Raycast extension
- The VS Code extension / language server
- 3rd party tools that use the Tailwind CSS design system APIs
> This PR looks very big, but **I think it's best to go over the changes
commit by commit**. Basically all of these steps already existed in the
upgrade tool, but are now moved to our core `tailwindcss` package.
Here is a list of all the changes:
- Added a new `canonicalizeCandidates` function to the design system
- Moved various migration steps to the core package. I inlined them in
the same file and because of that I noticed a specific pattern (more on
this later).
- Moved `printCandidate` tests to the `tailwindcss` package
- Setup tests for `canonicalizeCandidates` based on the existing tests
in the upgrade tool.
I noticed that all the migrations followed a specific pattern:
1. Parse the raw candidate into a `Candidate[]` AST
2. In a loop, try to migrate the `Candidate` to a new `Candidate` (this
often handled both the `Candidate` and its `Variant[]`)
3. If something changed, print the new `Candidate` back to a string, and
pass it to the next migration step.
While this makes sense in isolation, we are doing a lot of repeated work
by parsing, modifying, and printing the candidate multiple times. This
let me to introduce the `big refactor` commit. This changes the steps
to:
1. Up front, parse the raw candidate into a `Candidate[]` _once_.
2. Strip the variants and the important marker from the candidate. This
means that each migration step only has to deal with the base `utility`
and not care about the variants or the important marker. We can
re-attach these afterwards.
3. Instead of a `rawCandidate: string`, each migration step receives an
actual `Candidate` object (or a `Variant` object).
4. I also split up the migration steps for the `Candidate` and the
`Variant[]`.
All of this means that there is a lot less work that needs to be done.
We can also cache results between migrations. So `[@media_print]:flex`
and `[@media_print]:block` will result in `print:flex` and `print:block`
respectively, but the `[@media_print]` part is only migrated once across
both candidates.
One migration step relied on the `postcss-selector-parser` package to
parse selectors and attribute selectors. I didn't want to introduce a
package just for this, so instead used our own `SelectorParser` in the
migration and wrote a small `AttributeSelectorParser` that can parse the
attribute selector into a little data structure we can work with
instead.
If we want, we can split this PR up into smaller pieces, but since the
biggest chunk is moving existing code around, I think it's fairly doable
to review as long as you go commit by commit.
---
With this new API, we can turn:
```
[
'bg-red-500',
'hover:bg-red-500',
'[@media_print]:bg-red-500',
'hover:[@media_print]:bg-red-500',
'bg-red-500/100',
'hover:bg-red-500/100',
'[@media_print]:bg-red-500/100',
'hover:[@media_print]:bg-red-500/100',
'bg-[var(--color-red-500)]',
'hover:bg-[var(--color-red-500)]',
'[@media_print]:bg-[var(--color-red-500)]',
'hover:[@media_print]:bg-[var(--color-red-500)]',
'bg-[var(--color-red-500)]/100',
'hover:bg-[var(--color-red-500)]/100',
'[@media_print]:bg-[var(--color-red-500)]/100',
'hover:[@media_print]:bg-[var(--color-red-500)]/100',
'bg-(--color-red-500)',
'hover:bg-(--color-red-500)',
'[@media_print]:bg-(--color-red-500)',
'hover:[@media_print]:bg-(--color-red-500)',
'bg-(--color-red-500)/100',
'hover:bg-(--color-red-500)/100',
'[@media_print]:bg-(--color-red-500)/100',
'hover:[@media_print]:bg-(--color-red-500)/100',
'bg-[color:var(--color-red-500)]',
'hover:bg-[color:var(--color-red-500)]',
'[@media_print]:bg-[color:var(--color-red-500)]',
'hover:[@media_print]:bg-[color:var(--color-red-500)]',
'bg-[color:var(--color-red-500)]/100',
'hover:bg-[color:var(--color-red-500)]/100',
'[@media_print]:bg-[color:var(--color-red-500)]/100',
'hover:[@media_print]:bg-[color:var(--color-red-500)]/100',
'bg-(color:--color-red-500)',
'hover:bg-(color:--color-red-500)',
'[@media_print]:bg-(color:--color-red-500)',
'hover:[@media_print]:bg-(color:--color-red-500)',
'bg-(color:--color-red-500)/100',
'hover:bg-(color:--color-red-500)/100',
'[@media_print]:bg-(color:--color-red-500)/100',
'hover:[@media_print]:bg-(color:--color-red-500)/100',
'[background-color:var(--color-red-500)]',
'hover:[background-color:var(--color-red-500)]',
'[@media_print]:[background-color:var(--color-red-500)]',
'hover:[@media_print]:[background-color:var(--color-red-500)]',
'[background-color:var(--color-red-500)]/100',
'hover:[background-color:var(--color-red-500)]/100',
'[@media_print]:[background-color:var(--color-red-500)]/100',
'hover:[@media_print]:[background-color:var(--color-red-500)]/100'
]
```
Into their canonicalized form:
```
[
'bg-red-500',
'hover:bg-red-500',
'print:bg-red-500',
'hover:print:bg-red-500'
]
```
The list is also unique, so we won't end up with `bg-red-500 bg-red-500`
twice.
While the canonicalization itself is fairly fast, we still pay a **~1s**
startup cost for some migrations (once, and cached for the entire
lifetime of the design system). I would like to keep improving the
performance and the kinds of migrations we do, but I think this is a
good start.
The cost we pay is for:
1. Generating a full list of all possible utilities based on the
`getClassList` suggestions API.
2. Generating a full list of all possible variants.
The canonicalization step for this list takes **~2.9ms** on my machine.
Just for fun, if you use the `getClassList` API for intellisense that
generates all the suggestions and their modifiers, you get a list of
**263788** classes. If you canonicalize all of these, it takes
**~500ms** in total. So roughly **~1.9μs** per candidate.
This new API doesn't result in a performance difference for normal
Tailwind CSS builds.
The other potential concern is file size of the package. The generated
`tailwindcss.tgz` file changed like this:
```diff
- 684652 bytes (684.65 kB)
+ 749169 bytes (749.17 kB)
```
So the package increased by ~65 kB which I don't think is the end of the
world, but it is important for the `@tailwindcss/browser` build which we
don't want to grow unnecessarily. For this reason we remove some of the
code for the design system conditionally such that you don't pay this
cost in an environment where you will never need this API.
The `@tailwindcss/browser` build looks like this:
```shell
`dist/index.global.js 255.14 KB` (main)
`dist/index.global.js 272.61 KB` (before this change)
`dist/index.global.js 252.83 KB` (after this change, even smaller than on `main`)
```
Here is everything you need to know about this update. Please take a
good look at what changed and the test results before merging this pull
request.
### What changed?
#### ✳️ h3 (1.15.3 → 1.15.4) · [Repo](https://github.com/h3js/h3) ·
[Changelog](https://github.com/h3js/h3/blob/main/CHANGELOG.md)
<details>
<summary>Release Notes</summary>
<h4><a
href="https://github.com/h3js/h3/releases/tag/v1.15.4">1.15.4</a></h4>
<blockquote><p dir="auto"><a
href="https://bounce.depfu.com/github.com/h3js/h3/compare/v1.15.2...v1.15.4">compare
changes</a></p>
<h3 dir="auto">🩹 Fixes</h3>
<ul dir="auto">
<li>
<strong>getRequestHost:</strong> Return first host from <code
class="notranslate">x-forwarded-host</code> (<a
href="https://bounce.depfu.com/github.com/h3js/h3/pull/1175">#1175</a>)</li>
</ul>
<h3 dir="auto">💅 Refactors</h3>
<ul dir="auto">
<li>
<strong>useSession:</strong> Backport <code
class="notranslate">SessionManager</code> interface to fix types (<a
href="https://bounce.depfu.com/github.com/h3js/h3/pull/1058">#1058</a>)</li>
</ul>
<h3 dir="auto">🏡 Chore</h3>
<ul dir="auto">
<li>
<strong>docs:</strong> Fix typos (<a
href="https://bounce.depfu.com/github.com/h3js/h3/pull/1108">#1108</a>)</li>
</ul>
<h3 dir="auto">❤️ Contributors</h3>
<ul dir="auto">
<li>Pooya Parsa (<a
href="https://bounce.depfu.com/github.com/pi0">@pi0</a>)</li>
<li>Kricsleo (<a
href="https://bounce.depfu.com/github.com/kricsleo">@kricsleo</a>)</li>
<li>Izoukhai (<a
href="https://bounce.depfu.com/github.com/izoukhai">@izoukhai</a>)</li>
</ul></blockquote>
<p><em>Does any of this look wrong? <a
href="https://depfu.com/packages/npm/h3/feedback">Please let us
know.</a></em></p>
</details>
<details>
<summary>Commits</summary>
<p><a
href="5bd27a52d7...ec77d6bc14">See
the full diff on Github</a>. The new version differs by more commits
than we can show here.</p>
</details>
---
[Depfu](https://depfu.com) will automatically keep this PR
conflict-free, as long as you don't add any commits to this branch
yourself. You can also trigger a rebase manually by commenting with
`@depfu rebase`.
<details><summary>All Depfu comment commands</summary>
<blockquote><dl>
<dt>@depfu rebase</dt><dd>Rebases against your default branch and
redoes this update</dd>
<dt>@depfu recreate</dt><dd>Recreates this PR, overwriting any edits
that you've made to it</dd>
<dt>@depfu merge</dt><dd>Merges this PR once your tests are passing and
conflicts are resolved</dd>
<dt>@depfu cancel merge</dt><dd>Cancels automatic merging of this
PR</dd>
<dt>@depfu close</dt><dd>Closes this PR and deletes the branch</dd>
<dt>@depfu reopen</dt><dd>Restores the branch and reopens this PR (if
it's closed)</dd>
<dt>@depfu pause</dt><dd>Ignores all future updates for this dependency
and closes this PR</dd>
<dt>@depfu pause [minor|major]</dt><dd>Ignores all future minor/major
updates for this dependency and closes this PR</dd>
<dt>@depfu resume</dt><dd>Future versions of this dependency will
create PRs again (leaves this PR as is)</dd>
</dl></blockquote>
</details>
Co-authored-by: depfu[bot] <23717796+depfu[bot]@users.noreply.github.com>
Co-authored-by: Jordan Pittman <jordan@cryptica.me>
Closes#13694Closes#13591
# Source Maps Support for Tailwind CSS
This PR adds support for source maps to Tailwind CSS v4 allowing us to
track where styles come from whether that be user CSS, imported
stylesheets, or generated utilities. This will improve debuggability in
browser dev tools and gives us a good foundation for producing better
error messages. I'll go over the details on how end users can enable
source maps, any limitations in our implementation, changes to the
internal `compile(…)` API, and some details and reasoning around the
implementation we chose.
## Usage
### CLI
Source maps can be enabled in the CLI by using the command line argument
`--map` which will generate an inline source map comment at the bottom
of your CSS. A separate file may be generated by passing a file name to
`--map`:
```bash
# Generates an inline source map
npx tailwindcss -i input.css -o output.css --map
# Generates a separate source map file
npx tailwindcss -i input.css -o output.css --map output.css.map
```
### PostCSS
Source maps are supported when using Tailwind as a PostCSS plugin *in
development mode only*. They may or may not be enabled by default
depending on your build tool. If they are not you may be able to
configure them within your PostCSS config:
```jsonc
// package.json
{
// …
"postcss": {
"map": { "inline": true },
"plugins": {
"@tailwindcss/postcss": {},
},
}
}
```
### Vite
Source maps are supported when using the Tailwind CSS Vite plugin in
*development mode only* by enabling the `css.devSourcemap` setting:
```js
import tailwindcss from "@tailwindcss/vite";
import { defineConfig } from "vite";
export default defineConfig({
plugins: [tailwindcss()],
css: {
devSourcemap: true,
},
})
```
Now when a CSS file is requested by the browser it'll have an inline
source map comment that the browser can use.
## Limitations
- Production build source maps are currently disabled due to a bug in
Lightning CSS. See
https://github.com/parcel-bundler/lightningcss/pull/971 for more
details.
- In Vite, minified CSS build source maps are not supported at all. See
https://github.com/vitejs/vite/issues/2830 for more details.
- In PostCSS, minified CSS source maps are not supported. This is due to
the complexity required around re-associating every AST node with a
location in the generated, optimized CSS. This complexity would also
have a non-trivial performance impact.
## Testing
Here's how to test the source map functionality in different
environments:
### Testing the CLI
1. Setup typical project that the CLI can use and with sources to scan.
```css
@import "tailwindcss";
@utilty my-custom-utility {
color: red;
}
/* to test `@apply` */
.card {
@apply bg-white text-center shadow-md;
}
```
2. Build with source maps:
```bash
bun /path/to/tailwindcss/packages/@tailwindcss-cli/src/index.ts --input input.css -o output.css --map
```
3. Open Chrome DevTools, inspect an element with utility classes, and
you should see rules pointing to `input.css` or
`node_modules/tailwindcss/index.css`
### Testing with Vite
Testing in Vite will require building and installing necessary files
under `dist/*.tgz`.
1. Create a Vite project and enable source maps in `vite.config.js`:
```js
import tailwindcss from "@tailwindcss/vite";
import { defineConfig } from "vite";
export default defineConfig({
plugins: [tailwindcss()],
css: {
// This line is required for them to work
devSourcemap: true,
},
})
```
2. Add a component that uses Tailwind classes and custom CSS:
```jsx
// ./src/app.jsx
export default function App() {
return (
<div className="bg-blue-500 my-custom-class">
Hello World
</div>
)
}
```
```css
/* ./src/styles.css */
@import "tailwindcss";
@utilty my-custom-utility {
color: red;
}
/* to test `@apply` */
.card {
@apply bg-white text-center shadow-md;
}
```
3. Run `npm run dev`, open DevTools, and inspect elements to verify
source mapping works for both utility classes and custom CSS.
### Testing with PostCSS CLI
1. Create a test file and update your PostCSS config:
```css
/* input.css */
@import "tailwindcss";
@layer components {
.card {
@apply p-6 rounded-lg shadow-lg;
}
}
```
```jsonc
// package.json
{
// …
"postcss": {
"map": {
"inline": true
},
"plugins": {
"/path/to/tailwindcss/packages/packages/@tailwindcss-postcss/src/index.ts": {}
}
}
}
```
2. Run PostCSS through Bun:
```bash
bunx --bun postcss ./src/index.css -o out.css
```
3. Inspect the output CSS - it should include an inline source map
comment at the bottom.
### Testing with PostCSS + Next.js
Testing in Next.js will require building and installing necessary files
under `dist/*.tgz`. However, I've not been able to get CSS source maps
to work in Next.js without this hack:
```js
const nextConfig: NextConfig = {
// next.js overwrites config.devtool so we prevent it from doing so
// please don't actually do this…
webpack: (config) =>
Object.defineProperty(config, "devtool", {
get: () => "inline-source-map",
set: () => {},
}),
};
```
This is definitely not supported and also doesn't work with turbopack.
This can be used to test them temporarily but I suspect that they just
don't work there.
### Manual source map analysis
You can analyze source maps using Evan Wallace's [Source Map
Visualization](https://evanw.github.io/source-map-visualization/) tool
which will help to verify the accuracy and quality of source maps. This
is what I used extensively while developing this implementation.
It'll help verify that custom, user CSS maps back to itself in the
input, that generated utilities all map back to `@tailwind utilities;`,
that source locations from imported files are also handled correctly,
etc… It also highlights the ranges of stuff so it's easy to see if there
are off-by-one errors.
It's easiest to use inline source maps with this tool because you can
take the CSS file and drop it on the page and it'll analyze it while
showing the file content.
If you're using Vite you'll want to access the CSS file with `?direct`
at the end so you don't get a JS module back.
## Implementation
The source map implementation follows the ECMA-426 specification and
includes several key components to aid in that goal:
### Source Location Tracking
Each emittable AST node in the compilation pipeline tracks two types of
source locations:
- `src`: Original source location - [source file, start offset, end
offset]
- `dst`: Generated source location - [output file, start offset, end
offset]
This dual tracking allows us to maintain mappings between the original
source and generated output for things like user CSS, generated
utilities, uses of `@apply`, and tracking theme variables.
It is important to note that source locations for nodes _never overlap_
within a file which helps simplify source map generation. As such each
type of node tracks a specific piece of itself rather than its entire
"block":
| Node | What a `SourceLocation` represents |
| ----------- |
---------------------------------------------------------------- |
| Style Rule | The selector |
| At Rule | Rule name and params, includes the `@` |
| Declaration | Property name and value, excludes the semicolon |
| Comment | The entire comment, includes the start `/*` and end `*/`
markers |
### Windows line endings when parsing CSS
Because our AST tracks nodes through offsets we must ensure that any
mutations to the file do *not* change the lenth of the string. We were
previously replacing `\r\n` with `\n` (see [filter code
points](https://drafts.csswg.org/css-syntax/#css-filter-code-points)
from the spec) — which changes the length of the string and all offsets
may end up incorrect. The CSS parser was updated to handle the CRLF
token directly by skipping over the `\r` and letting remaining code
handle `\n` as it did previously. Some additional tweaks were required
when "peeking" the input but those changes were fairly small.
### Tracking of imports
Source maps need paths to the actual imported stylesheets but the
resolve step for stylesheets happens inside the call to `loadStylesheet`
which make the file path unavailable to us. Because of this the
`loadStylesheet` API was augmented such that it has to return a `path`
property that we can then use to identify imported sources. I've also
made the same change to the `loadModule` API for consistency but nothing
currently uses this property.
The `path` property likely makes `base` redundant but elminating that
(if we even want to) is a future task.
### Optimizing the AST
Our optimization pass may intoduce some nodes, for example, fallbacks we
create for `@property`. These nodes are linked back to `@tailwind
utilities` as ultimately that is what is responsible for creating them.
### Line Offset Tables
A key component to our source map generation is the line offset table,
which was inspired by some ESBuild internals. It stores a sorted list of
offsets for the start of each line allowing us to translate offsets to
line/column `Position`s in `O(log N)` time and from `Position`s to
offsets in `O(1)` time. Creation of the table takes `O(N)` time.
This means that we can store code point offsets for source locations and
not have to worry about computing or tracking line/column numbers during
parsing and serialization. Only when a source map is generated do these
offsets need to be computed. This ensures the performance penalty when
not using source maps is minimal.
### Source Map Generation
The source map returned by `buildSourceMap()` is designed to follow the
[ECMA-426 spec](https://tc39.es/ecma426). Because that spec is not
completely finalized we consider the result of `buildSourceMap()` to be
internal API that may change as the spec chamges.
The produces source map is a "decoded" map such that all sources and
mappings are in an object graph. A library like `source-map-js` must be
used to convert this to an encoded source map of the right version where
mappings are encoded with base 64 VLQs.
Any specific integration (Vite, PostCSS, etc…) can then use
`toSourceMap()` from `@tailwindcss/node` to convert from the internal
source map to an spec-compliant encoded source map that can be
understood by other tools.
### Handling minification in Lightning
Since we use Lightning CSS for optimization, and it takes in an input
map, we generate an encoded source map that we then pass to lightning.
The output source map *from lighting itself* is then passed back in
during the second optimization pass. The final map is then passed from
lightning to the CLI (but not Vite or PostCSS — see the limitations
section for details).
In some cases we have to "fix up" the output CSS. When this happens we
use `magic-string` to do the replacement in a way that is trackable and
`@amppproject/remapping` to map that change back onto the original
source map. Once the need for these fix ups disappear these dependencies
can go away.
Notes:
- The accuracy of source maps run though lightning is reduced as it only
tracks on a per-rule level. This is sufficient enough for browser dev
tools so should be fine.
- Source maps during optimization do not function properly at this time
because of a bug in Lightning CSS regarding license comments. Once this
bug is fixed they will start working as expected.
### How source locations flow through the system
1. During initial CSS parsing, source locations are preserved.
2. During parsing these source locations are also mapped to the
destinations which supports an optimization for when no utilities are
generated.
3. Throughout the compilation process, transformations maintain source
location data
4. Generated utilities are explicitly pointed to `@tailwind utilities`
unless generated by `@apply`.
5. When optimization is enabled, source maps are remapped through
lightningcss
6. Final source maps are written in the requested format (inline or
separate file)
Prepare the 4.0.16 release.
~~Also added a commit to mark the `--value('…')` and `--modifier('…')`
with literals strings as an experimental feature (aka not shipped in
this PR). But we can revert that commit if we still want to ship it in
4.0.16 instead of 4.1.~~
---------
Co-authored-by: Philipp Spiess <hello@philippspiess.com>
<!--
👋 Hey, thanks for your interest in contributing to Tailwind!
**Please ask first before starting work on any significant new
features.**
It's never a fun experience to have your pull request declined after
investing a lot of time and effort into a new feature. To avoid this
from happening, we request that contributors create an issue to first
discuss any significant new features. This includes things like adding
new utilities, creating new at-rules, or adding new component examples
to the documentation.
https://github.com/tailwindcss/tailwindcss/blob/master/.github/CONTRIBUTING.md
-->
<!--
👋 Hey, thanks for your interest in contributing to Tailwind!
**Please ask first before starting work on any significant new
features.**
It's never a fun experience to have your pull request declined after
investing a lot of time and effort into a new feature. To avoid this
from happening, we request that contributors create an issue to first
discuss any significant new features. This includes things like adding
new utilities, creating new at-rules, or adding new component examples
to the documentation.
https://github.com/tailwindcss/tailwindcss/blob/master/.github/CONTRIBUTING.md
-->
---------
Co-authored-by: Adam Wathan <adam.wathan@gmail.com>
<!--
👋 Hey, thanks for your interest in contributing to Tailwind!
**Please ask first before starting work on any significant new
features.**
It's never a fun experience to have your pull request declined after
investing a lot of time and effort into a new feature. To avoid this
from happening, we request that contributors create an issue to first
discuss any significant new features. This includes things like adding
new utilities, creating new at-rules, or adding new component examples
to the documentation.
https://github.com/tailwindcss/tailwindcss/blob/master/.github/CONTRIBUTING.md
-->
---------
Co-authored-by: Adam Wathan <4323180+adamwathan@users.noreply.github.com>
Resolves#15977
Our `@tailwindcss/browser` build is intended to run inside a `<script>`
tag inside browsers. Because of how variable assignment within
`<script>` tags work, all variables that were defined within that script
are currently assigned on the global namespace.
This is especially troublesome as eslint uses `$` as a valid mangling
character so importing the CDN build would now redefine `globalThis.$`
which collides with many very popular JavaScript libraries.
In order to avoid polluting the global namespace, this PR changes the
build step to emit an IIFE (so all vars defined are scroped to the
function closure instead of the global namespace).
## Test plan
- Ensure UI tests still work
- <img width="533" alt="Screenshot 2025-01-28 at 16 49 27"
src="https://github.com/user-attachments/assets/1e027451-f58b-4252-bf97-c016a90eb78b"
/>
This PR adds a `main` and `browser` field for the `@tailwindcss/browser`
package.
In the package, we do have the `exports` field setup, which is an
alternative to the `main` field according to the docs:
> The "exports" provides a modern alternative to "main" …
>
> —
https://docs.npmjs.com/cli/v10/configuring-npm/package-json?v=true#exports
However, if you look at the unpkg link:
https://unpkg.com/@tailwindcss/browser, it tries to load the `index.js`
file. This is probably a bug in the unpkg resolver.
That said, if you look at other CDNs such as esm.sh, it does resolve
correctly: https://esm.sh/@tailwindcss/browser
According to the npm docs:
> If `main` is not set, it defaults to `index.js` in the package's root
folder.
>
> —
https://docs.npmjs.com/cli/v10/configuring-npm/package-json?v=true#main
This explains why unpkg is trying to load the `index.js` file.
Additionally, the npm docs also mention the `browser` field:
> If your module is meant to be used client-side the browser field
should be used instead of the main field. This is helpful to hint users
that it might rely on primitives that aren't available in Node.js
modules. (e.g. window)
>
> —
https://docs.npmjs.com/cli/v10/configuring-npm/package-json?v=true#browser
So this PR also adds that field just to be sure.
