perf(parser): speed up simple lookaheads by introducing Lexer::peek_token

[ulrichstark]

This is my proof of concept function to implement @overlookmotel's tip:

We could also introduce a more slimmed-down (cheaper) version of lookahead when all you need is to get the next 1 or 2 tokens from lexer and then reset its pointer and errors_pos back to where they were before. Current lookahead does quite a lot of work which isn't required in these cases.

source: #11334 (comment)

[codspeed-hq]

CodSpeed Instrumentation Performance Report

Merging #11358 will not alter performance

_{Comparing ulrichstark:attempt-at-speeding-up-simple-lookaheads-in-parser (7495fb4) with main (9736a7f)}

Summary

✅ 38 untouched benchmarks

[leaysgur]

oxc/crates/oxc_parser/src/js/statement.rs

Lines 347 to 354 in adbb3a3

	&& self.lookahead(|p| {
	p.bump_any();
	if !p.at(Kind::Using) || p.cur_token().is_on_new_line() {
	return false;
	}
	p.bump_any();
	!p.cur_token().is_on_new_line()
	})

FYI: Not often, but in a few places this happens.

[Boshen]

This looks very promising! Let's wait until we fix all the code around parser peek before going into this.

[ulrichstark]

Just rebased to include #11356 that was causing conflicts.

[ulrichstark]

• Rebased again
• Changed api from Lexer::lookahead_token taking predicate to Lexer::peek_token returning peeked token
• Found three more cases to use Lexer::peek_token

[Boshen]

I'll take a look at the remaining token peeks before this PR.

[Boshen]

Can't resist the 2%!