perf(lexer): optimize Source

[overlookmotel]

SourcePosition has a couple of invariants:

• Any SourcePosition must be in bounds of Source.
• Distance between any 2 SourcePositions cannot exceed u32::MAX, because that's the upper limit on size of source text.

However, prior to this PR, all compiler could see is pointers which could be anything. It wasn't able to understand the invariants, or utilize them for efficient codegen.

This PR makes compiler aware of these invariants in 2 ways:

1. Replace comparing pointers with >, < etc, and calculating offsets via ptr as usize calculations with std::ptr::offset_from. offset_from has safety invariants which match Source's.
2. Calculate offsets between SourcePositions with u32::try_from(distance).unwrap_unchecked(). This informs compiler offset is definitely non-negative and <= u32::MAX.

[overlookmotel]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• perf(lexer): optimize Source #8295 👈 (View in Graphite)
• fix(lexer): Source is not Clone #8294
• refactor(lexer): tighten safety of lexer by always including lifetime on SourcePosition #8293
• test(lexer): assert size of Token in 32-bit WASM #8292
• refactor(lexer): make handle_byte a method of Lexer #8291
• style(lexer): reorder imports #8290
• refactor(lexer): replace #[allow] with #[expect] #8289
• refactor(lexer): rename function param #8288
• main

---

How to use the Graphite Merge Queue

Add either label to this PR to merge it via the merge queue:

• 0-merge - adds this PR to the back of the merge queue
• hotfix - for urgent hot fixes, skip the queue and merge this PR next

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

An organization admin has enabled the Graphite Merge Queue in this repository.

Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.

This stack of pull requests is managed by Graphite. Learn more about stacking.

[codspeed-hq]

CodSpeed Performance Report

Merging #8295 will not alter performance

_{Comparing 01-01-perf_lexer_optimize_source_ (703d983) with main (e1f8ea4)}

Summary

✅ 29 untouched benchmarks

[overlookmotel]

Ha! It actually seems to reduce perf slightly, according to CodSpeed. Probably it's the usize to u32 conversion in offsets. I'll try to remove that overhead.