refactor(format): use specialised tokens

[ematipico]

Summary

This PR refactors how the formatter formats trivias. Up until now, all sorts of trivia were formatted using generic functions such as: format_replaced, format_removed, format_skipped_token_trivia, etc.

These functions, however, have a shared logic that can't be overridden by specialised formatters.

In this PR, we introduce a trait called FormatToken, which exposes functions such as format_replaced and more. Those functions, now, must be implemented within the language formatters, which means that biome_formatter no longer knows how to do that.

This refactor has some small repercussions on the separated lists, where now we need to pass generic functions on_skipped and on_removed, because the generic formatter doesn't know how to print their trivia anymore.

With this change, we can implement a specialised version of fmt_skipped that doesn't rely on the nestle logic: #5356

Test Plan

The current CI should stay all green

Docs

[changeset-bot]

⚠️ No Changeset found

Latest commit: 7676748

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[codspeed-hq]

CodSpeed Performance Report

Merging #7052 will not alter performance

_{Comparing refactor/format-with-specialised-tokens (a94c114) with main (b091ddf)¹}

Summary

✅ 115 untouched benchmarks

Footnotes

1. No successful run was found on main (c8f7838) during the generation of this report, so b091ddf was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[dyc3]

While this is good, I think the massive diff could be avoidable. See below.

[dyc3]

LGTM! The contributing docs for the formatter will likely need to be updated too. Not going to block on that for this PR.