Canonicalization constant folding and handling zeros #19095
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Typically each math operator is parsed as a separate word and that's because it's also typically surrounded by spaces. In case of `/`, this was parsed as a separator instead of a word including the whitespace (if any). The use case we had was to parse `theme(colors.red.500/20%)` where we don't want `colors.red.500/20%` to be a single word, but we want 3 parts instead. This normalizes this a little bit such that all math operators (+, -, *, /) are just the `word` nodes and any surrounding whitespace will be a separator node instead.
This is essentially the same as `walk`, but depth first so we handle leaf nodes first. This helps in situations like constant folding where we can work our way up.
We will use those in the constant folding logic. We could use `ValueParser.parse` but we already know we want `word(…)` values so no need to do the extra work.
This doesn't operate on a declaration, but a declaration value right now. Could extend this in the future by also passing in the declaration property if we want more smart fold logic.
This will rely on 2 things: 1. That we properly constant fold length units into `0` 2. That we prefer the "positive" version (`mt-0`) over its negative version (`-mt-0`)
Given they have the same signature.
RobinMalfait
commented
Oct 9, 2025
packages/@tailwindcss-upgrade/src/codemods/template/migrate.test.ts
Outdated
Show resolved
Hide resolved
RobinMalfait
commented
Oct 9, 2025
Comment on lines
+60
to
+61
| // | ||
| // TODO: Ensure it's safe to do so based on the data types? |
There was a problem hiding this comment.
I think this is fairly safe, and feels like something we only have to add support for when running into actual issues 🤔
There was a problem hiding this comment.
Let me see if I can break this
There was a problem hiding this comment.
Probaaaably, but once you do, the questions is: how common / real is the actual issue.
There was a problem hiding this comment.
I think normalizing things with units to without could cause problems (like rem, em, etc… should probably normalize to 0px instead) but then how would you know margin-left: 0 normalizes to 0px without encoding data type info and that'd be annoying.
All that to say… I think this is fine for now. We can improve this later if need be.
RobinMalfait
commented
Oct 9, 2025
Co-authored-by: Jordan Pittman <jordan@cryptica.me>
thecrypticace
approved these changes
Oct 9, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The main goal of this PR was to support canonicalization of zero like values. We essentially want to canonicalize
-mt-0asmt-0, but alsomt-[0px],mt-[0rem], and other length-like units to justmt-0.To do this, we had to handle 2 things:
0pxand0remfold to0. We only do this for length units. We also normalize-0,+0,-0.0and so on to0.mt-0over-mt-0if both result in the same signature.Moved some of the constant folding logic into its own function and added a bunch of separate tests for it.
Test plan
Added more unit tests where we normalize different zero-like values to
0.Running the canonicalization logic: