fix(parser): parse let?.x

[Copilot]

Fixes #16839.

Make parser parse let?.x successfully.

There's a failure in codegen idempotency tests, but that's an unrelated and pre-existing bug, so am leaving that to another PR.

[overlookmotel]

@copilot See below comment and action it.

[codspeed-hq]

CodSpeed Performance Report

Merging #16840 will not alter performance

_{Comparing copilot/fix-parser-let-x (c7c05ba) with main (b16fe64)}

Summary

✅ 42 untouched
⏩ 3 skipped¹

Footnotes

1. 3 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[Copilot]

Pull request overview

This PR fixes the parser to correctly handle optional chaining syntax with the let keyword (e.g., let?.x). Previously, the parser would fail to parse such expressions. The fix recognizes that in script mode (non-strict), let can be used as an identifier, and when followed by optional chaining (?.), it should be parsed as a member expression rather than a variable declaration.

Key Changes:

• Updated parser logic to recognize Kind::QuestionDot after let keyword
• Added test cases for both valid (script mode) and invalid (module mode) usage
• Fixed the dot-notation ESLint conformance test that was previously failing

Reviewed changes

Copilot reviewed 2 out of 9 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
crates/oxc_parser/src/js/declaration.rs	Added Kind::QuestionDot to the pattern match that handles let as an identifier with property access
tasks/coverage/misc/pass/let-optional-chaining.cjs	Test file demonstrating valid usage of let?.x in script mode (CommonJS)
tasks/coverage/misc/fail/let-optional-chaining.js	Test file demonstrating invalid usage of let?.x in module mode (should produce errors)
tasks/coverage/snapshots/parser_misc.snap	Updated snapshot showing 4 new error cases for the fail test and incremented test counts
tasks/coverage/snapshots/semantic_misc.snap	Updated snapshot with incremented test counts
tasks/coverage/snapshots/transformer_misc.snap	Updated snapshot with incremented test counts
tasks/coverage/snapshots/formatter_misc.snap	Updated snapshot with incremented test counts
tasks/coverage/snapshots/codegen_misc.snap	Updated snapshot showing one expected failure (pre-existing bug per PR description)
apps/oxlint/conformance/snapshot.md	Updated conformance metrics showing dot-notation rule now fully passing (69/69 tests)

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[charliecreates]

The parser tweak is minimal and likely correct, but it changes the declaration-vs-expression routing based on a broader lookahead set and deserves extra regression coverage in statement-sensitive positions. The new module-mode failure fixture produces repetitive diagnostics, making snapshots noisier than necessary. codegen_misc.snap indicates the new fixture lowers codegen positive pass rate, which risks institutionalizing a red/unstable coverage metric even if the underlying idempotency issue is “pre-existing.”

Additional notes (2)

• Maintainability | crates/oxc_parser/src/js/declaration.rs:24-31
  The change to include Kind::QuestionDot fixes the immediate parse for let?.x, but this branch now routes into parse_expr() based solely on lookahead. That can widen the set of token sequences that stop being treated as let-declaration candidates and instead become expression statements.

Even if the current tests pass, consider adding a focused regression case where let is followed by ?. inside contexts that are sensitive to statement vs declaration parsing (e.g. for (let?.x;;) {} / if (cond) let?.x / label: let?.x) to ensure this early decision doesn’t create new ambiguities.

• Maintainability | tasks/coverage/misc/fail/let-optional-chaining.js:1-5
  The module-mode failure fixture is currently expecting a reserved-word error at each let?.… line. That’s good for coverage, but it’s somewhat redundant and makes the snapshot noisy (multiple near-identical diagnostics).

Consider reducing to a single representative let?.x; in the fail fixture unless you specifically need to assert multiple optional-chaining forms fail at parse time in module mode.

Summary of changes

What changed

Parser behavior

• Updated crates/oxc_parser/src/js/declaration.rs to treat let?.… as an expression-statement continuation by including Kind::QuestionDot in the lookahead set (Dot | QuestionDot | LParen).

Test coverage

• Added script-mode pass fixture: tasks/coverage/misc/pass/let-optional-chaining.cjs.
• Added module-mode fail fixture: tasks/coverage/misc/fail/let-optional-chaining.js.

Snapshots / conformance

• Updated various snapshots to reflect the additional fixture and its expected pass/fail outcomes:
  • tasks/coverage/snapshots/{parser_misc,formatter_misc,codegen_misc,semantic_misc,transformer_misc}.snap
• Updated apps/oxlint/conformance/snapshot.md totals and moved dot-notation into the fully passing set.