This PR fixes two issues we found when testing the candidate codemodes:
1. Sometimes, core would emit the same candidate twice. This would
result into rewriting a range multiple times, without realizing that
this change might already be applied, causing it to swallow or duplicate
some bytes.
2. The codemods were mutating the `Candidate` object, however since the
`Candidate` parsing is _cached_ in core, it would sometimes return the
same instance. This is an issue especially since we monkey patch the
prefix to `null` when migrating prefixed candidates. This means that a
candidate would be cached that would be _invalid relative to the real
design system_. We fixed this by making sure the mutations would only be
applied to clones of the `Candidate` and I changed the `DesignSystem`
API to return `ReadOnly<T>` versions of these candidates. A better
solution would maybe be to disable the cache at all but this requires
broader changes in Core.
In order to properly migrate your Tailwind CSS v3 project to v4, we need
access to the JavaScript configuration object. This was previously only
required for template migrations, but in this PR we're making it so that
this is also a prerequisite of the CSS migrations. This is because some
migrations, like `@apply`, also need to convert candidates that to the
v4 syntax and we need the full config in order to properly validate
them.
In addition to requiring a JS config, we also now attempt to
automatically find the right configuration file inside the current
working directory. This is now matching the behavior of the Tailwind CSS
v3 CLI where it will find the config automatically if it's in the
current directory and called `tailwind.conf.js`.
---------
Co-authored-by: Robin Malfait <malfait.robin@gmail.com>
This PR extracts all _candidate migrations_ from the existing _template
migrations_ and reuses these in the `@apply` CSS migration. Seems like
this _JustWorks✨_.
---------
Co-authored-by: Adam Wathan <adam.wathan@gmail.com>
One of the breaking changes of v4 is the [inversion of variant order
application](https://github.com/tailwindlabs/tailwindcss/pull/13478). In
v3, variants are applied "inside-out".
For example a candidate like `*:first:underline` would produce the
following CSS in v3:
```css
.\*\:first\:underline:first-child > * {
text-decoration-line: underline;
}
```
To get the same behavior in v4, you would need to invert the candidate
order to `first:*:underline`. This would generate the following CSS in
v4:
```css
:where(.first\:\*\:underline:first-child > *) {
text-decoration-line: underline;
}
```
## The Migration
The most naive approach would be to invert the variants for every
candidate with at least two variants. This, however, runs into one issue
and some unexpected inconsistencies. I have identified the following
areas:
1. Some pseudo class variants _must appear at the end of the selector_.
v3 was patching over this by doing some manual reordering in for these
variants. For example, in v3, both of these variants create the same
output CSS: `hover:before:underline` and `before:hover:underline`. In v4
we simplified this system though and no longer generate the same output
in both cases. Instead, you'd always want to write
`hover:before:underline`, ensuring that these variants will appear at
the end.
For an exact list of which variants these affect, take a look [at this
diff](https://github.com/tailwindlabs/tailwindcss/pull/13478/files#diff-7779a0eebf6b980dd3abd63b39729b3023cf9a31c91594f5a25ea020b066e1c0L228-L246).
2. The `dark` variant and other at-rule variants are usually written
before other variants. This is more of a recommendation to make it
easier to read candidates rather than a difference in behavior as
`@media` queries are hoisted by the engine. For this reason, both of
these variants are _correct_ yet in real applications we prefer the
first one: `lg:hover:underline`, `hover:lg:underline`.
To avoid shuffling these rules across all candidates during the
migration, we bucket `dark` and other at-rule variants into a special
bucket that will not have their order changed (since people wrote stacks
like `sm:max-lg:` before and we want to keep them as-is) and appear
before all other variants.
3. For some variant stacks, the order does not matter. E.g.:
`focus:hover:underline` and `hover:focus:underline` will be the same. We
don't want to needlessly shuffle their order if we have to.
With these considerations, the migration now works as follows:
- If there is less then two variants, we do not need to migrate the
candidate
- If _every_ variant in the stack is an order-independent variant, we do
not need to migrate the candidate
- _Note that this is currently hardcoded to only support `&:hover` and
`&:focus`._
- Otherwise, we loop over the candidates and put them into three
buckets:
- `mediaVariants` hold variants that only contribute `@media` rules
_and_ the `dark` variant.
- `pseudoElementVariants` hold variants that _must appear at the end of
the selector_. This is based on the allow list from v3/early v4.
- `regularVariants` contains the rest.
- We now compute if any of the variants inside `regularVariants` is
order dependent.
- With this list of variants, we now construct the new order of variants
as:
```ts
[
...atRuleVariants,
...(anyRegularVariantOrderDependent ? regularVariants.reverse() :
regularVariants),
...pseudoElementVariants,
]
```
---------
Co-authored-by: Adam Wathan <adam.wathan@gmail.com>
This PR adds a new migration that can migrate Tailwind CSS v3 style
prefixes into Tailwind CSS v4.
The migration is split into three separate pieces of work:
1. Firstly, we need to read the full JavaScript config to get the _old_
prefix option. This is necessary because in v4, we will not allow things
like custom-separators for the prefix. From this option we will then try
and compute a new prefix (in 90% of the cases this is going to just
remove the trailing `-` but it can also work in more complex cases).
2. Then we migrate all Candidates. The important thing here is that we
need to operate on the raw candidate string because by relying on
`parseCandidate` (which we do for all other migrations) would not work,
as the candidates are not valid in v4 syntax. More on that in a bit.
3. Lastly we also make sure to update the CSS config to include the new
prefix. This is done by prepending the prefix option like so:
```css
@import "tailwindcss" prefix(tw);
```
### Migrating candidates
The main difference between v3 prefixes and v4 prefixes is that in v3,
the prefix was _part of the utility_ where as in v4 it is _always in
front of the CSS class.
So, for example, this candidate in v3:
```
hover:-tw-mr-4
```
Would be converted to the following in v4:
```
tw:hover:-mr-4
```
Since the first example _won't parse as a valid Candidate in v4, as the
`tw-mr` utility does not exist, we have to operate on the raw candidate
string first. To do this I created a fork of the `parseCandidate`
function _without any validation of utilities or variants_. This is used
to identify part of the candidate that is the `base` and then ensuring
the `base` starts with the old prefix. We then remove this to create an
"unprefixed" candidate that we validate against a version of the
DesignSystem _with no prefixes configured_. If the variant is valid this
way, we can then print it again with the `DesignSystem` that has the new
prefix to get the migrated version.
Since we set up the `DesignSystem` to include the new prefix, we can
also be certain that migrations that happen afterwards would still
disqualify candidates that aren't valid according to the new prefix
policy. This does mean we need to have the prefix fixup be the first
step in our pipeline.
One interesting bit is that in v3, arbitrary properties did not require
prefixes where as in v4 they do. So the following candidate:
```
[color:red]
```
Will be converted to:
```
tw:[color:red]
```
In v4, we're [removing automatic var
injection](https://github.com/tailwindlabs/tailwindcss/pull/13657)
(please refer to this PR for more detail as to why).
Automatic var injection made it so that if you have a candidate like
`bg-[--my-color]`, v3 would automatically wrap the content of the
arbitrary section with a `var(…)`, resulting in the same as typing
`bg-[var(--my-color)]`.
This PR adds codemods that go over various arbitrary fields and does the
`var(…)` injection for you. To be precise, we will add `var(…)` to:
- Modifiers, e.g.: `bg-red-500/[var(--my-opacity)]`
- Variants, e.g.: `supports-[var(--test)]:flex`
- Arbitrary candidates, e.g.: `[color:var(--my-color)]`
- Arbitrary values for functional candidates, e.g.:
`bg-[var(--my-color)]`
---------
Co-authored-by: Robin Malfait <malfait.robin@gmail.com>
This PR lands a quick interface update for template migration with some
lessons learned form our existing migrations. Specifically, this version
attempts to:
- Allow migrations to access the raw candidate. This way we can migrate
candidates that _would not parse as valid in v4_. This will help us
migrate prefixes in candidates from v3 to v4.
- There is no more awkward "return null" if nothing has changed. The
return `null` was necessary because we relied on mutating the Variant
and since parsing/printing could remove some information, it was not
easy to find out wether a candidate needed to be migrated at all. With a
string though, we can do this cheaply by returning the `rawCandidate`.
- We previously asserted that if `parseCandidate` returns more than one
candidate, we only picked the first one. This behavior is now moved into
the migrations where we have more context. For now though, we still do
not need to worry about this since in all cases, these duplicate
candidates would serialize to the same `Candidate`. It is helpful if you
only want to run a migration on a specific type of candidate (e.g. if
there's a `static` one and a more generic `functional` one).
- We need access to the `DesignSystem` inside migrations now to be able
to `parseCandidate`s.
Opening this up as a separate PR since it can take some time to iron out
the edge cases for the individual codemod PRs and I don't want to be
rebasing all the time.
## Before
```ts
type Migration = (candidate: Candidate) => Candidate | null
```
## After
```ts
type Migration = (designSystem: DesignSystem, rawCandidate: string) => string
```
In v4, we're making room for more gradient primitives with the addition
of [gradient and conic gradient
utilities](https://github.com/tailwindlabs/tailwindcss/pull/14467). To
make this clearer, we are renaming all v3 `bg-gradient-*` utilities to
`bg-linear-*` to make it clear that these refer to linear gradients.
This PR adds a migration that will automatically update the class names
based on this migration.
This PR adds the initial setup and a first codemod for the template
migrations. These are a new set of migrations that operate on files
defined in the Tailwind v3 config as part of the `content` option (so
your HTML, JavaScript, TSX files etc.).
The migration for this is integrated in the new `@tailwindcss/upgrade`
package and will require pointing the migration to an input JavaScript
config file, like this:
```
npx @tailwindcss/upgrade --config tailwind.config.js
```
The idea of template migrations is to apply breaking changes from the v3
to v4 migration within your template files.
## Migrating !important syntax
The first migration that I’m adding with this PR is to ensure we use the
v4 important syntax that has the exclamation mark at the end of the
utility.
For example, this:
```html
<div class="!flex sm:!block"></div>
```
Will now turn into:
```html
<div class="flex! sm:block!"></div>
```
## Architecture considerations
Implementation wise, we make use of Oxide to scan the content files fast
and efficiently. By relying on the same scanner als Tailwind v4, we
guarantee that all candidates that are part of the v4 output will have
gone through a migration.
Migrations itself operate on the abstract `Candidate` type, similar to
the type we use in the v4 codebase. It will parse the candidate into its
parts so they can easily be introspected/modified. Migrations are typed
as:
```ts
type TemplateMigration = (candidate: Candidate) => Candidate | null
```
`null` should be returned if the `Candidate` does not need a migration.
We currently use the v4 `parseCandidate` function to get an abstract
definition of the candidate rule that we can operate on. _This will
likely need to change in the future as we need to fork `parseCandidate`
for v3 specific syntax_.
Additionally, we're inlining a `printCandidate` function that can
stringify the abstract `Candidate` type. It is not guaranteed that this
is an identity function since some information can be lost during the
parse step. This is not a problem though, because migrations will only
run selectively and if none of the selectors trigger, the candidates are
not updated. h/t to @RobinMalfait for providing the printer.
So the overall flow of a migration looks like this:
- Scan the config file for `content` files
- Use Oxide to extract a list of candidate and their positions from
these `content` files
- Run a few migrations that operate on the `Candidate` abstract type.
- Print the updated `Candidate` back into the original `content` file.
---------
Co-authored-by: Robin Malfait <malfait.robin@gmail.com>
Co-authored-by: Jordan Pittman <jordan@cryptica.me>
