refactor(semantic)!: Do not expose CompactStr from Reference

[DonIsaac]

Changes Reference::name() to return a &str instead of a &CompactStr. As such, it is technically a breaking change. I scanned through Rolldown to see how they use References, and I do not think this PR will break anything.

Doing this will give us flexibility to change the representation of a reference name in the future. I'm thinking of making this an Rc<CompactStr> or something similar to reduce allocations within semantic.

[graphite-app]

Your org has enabled the Graphite merge queue for merging into main

Add the label “merge” to the PR and Graphite will automatically add it to the merge queue when it’s ready to merge. Or use the label “hotfix” to add to the merge queue as a hot fix.

You must have a Graphite account and log in to Graphite in order to use the merge queue. Sign up using this link.

[codspeed-hq]

CodSpeed Performance Report

Merging #3985 will not alter performance

_{Comparing don/semantic/refactor/name-api (cf7913f) with main (b257d53)}

Summary

✅ 28 untouched benchmarks

[Dunqing]

It seems to have some impact on performance

[Boshen]

Handling out CompactStr is performant here because majority of the symbols are inlined short identifiers.

[overlookmotel]

@DonIsaac Just to say, in my view we shouldn't be using CompactString / CompactStr at all in Semantic. We could just use Atom and avoid conversion costs to/from CompactStr.

Please see oxc-project/backlog#32 which is a work-in-progress shopping list of various optimizations we could make to semantic. But... my intent is to review all the code which uses Semantic, try to think about its design "holistically", and then write up a RFC proposal to revamp it. I hope to do that later this month, or more likely next month.

There's plenty of room for improvement, but in my view, at this point it'll be more fruitful to tackle it as a whole with a defined roadmap, rather than making lots of incremental changes.

However... if you're keen to get busy on this, I'm pretty sure replacing CompactStr with Atom in semantic will be a good idea. If you'd like to do that, please open an issue for it and make sure it's agreed that's a good way forwards (I think it is, but others may disagree with me) before spending your time doing a PR.

[overlookmotel]

Oh I just saw #3972 and #3974 (am catching up on issues in reverse order). Damn! Clearly this needs more thought...