#5223 added a new variant `JSXElementName::IdentifierReference` for JSX identifiers which are treated as references (e.g. `<Foo>`) as opposed to `JSXElementName::Identifier` which is for lowercase (e.g. `<div>`).
But we were incorrectly categorizing identifiers beginning with `_` or `$` - these should also be treated as references.
(as discussed in https://github.com/oxc-project/oxc/pull/5339#issuecomment-2321037779)
Follow-on after #5223.
JSX identifiers which start with a capital letter are now `JSXElementName::IdentifierReference`s, so no need to check for capitalized `JSXElementName::Identifier`s.
Preparation for upstack PRs. Graphical reporter diagnostics look exactly the same. This does, however, change `to_string()` to only show diagnostic messages, and not error codes + messages.
Fixing code is an important part of linter logic. We want to make sure fixers for each rule, and the code responsible for applying those fixes, are included in benchmarks.
As it currently stands, fixer closures are applied regardless of whether the user wants fixers to be applied. However, this is an implementation detail and is subject to change. I also want to bench the performance of `Fixer`.
Transform checker don't output spans in errors. They're inaccurate because (a) spans refer to the source text pre-transform, and (b) most errors relate to scopes/symbols/references newly created in transformer, which have no span.
Transformer benchmark measure only the transformer itself. Parse, generate `Semantic`, and drop `Semantic` outside of the measured section.
This should:
1. Give us greater visibility of how changes to transformer affect its performance.
2. Reduce variance in transformer benchmarks, since they will no longer include the variance introduced by `SemanticBuilder`.
Not ready to merge yet. We should first add an "end to end" benchmark testing the entire compilation process (parse - semantic - transform - minify - codegen).
This PR depends on https://github.com/Boshen/criterion2.rs/pull/49. This PR currently makes Oxc's dependency on `criterion2` a git dependency on that PR's branch. That can be changed once the upstream PR is merged.
Include dropping `Semantic` in semantic benchmark measure.
If you create a `Semantic`, you have to drop it, so the time it takes to
drop is part of the cost of using this API, and we should be working to
reduce it. Therefore I think it should be included in the benchmark.
It'll be interesting to see what effect a PR like #5232 which removes a
bunch of `Vec`s from `Semantic` has on the drop time.
### Before
```
x Flag u is mentioned twice in regular expression literal
,-[1:20]
1 | const a = /\2(.)/uuxig;
: ^
2 | debugger;
`----
x Unexpected flag x in regular expression literal
,-[1:21]
1 | const a = /\2(.)/uuxig;
: ^
2 | debugger;
`----
```
### After
```
x Flag u is mentioned twice in regular expression literal
,-[1:19]
1 | const a = /\2(.)/uuxig;
: ^
2 | debugger;
`----
x Unexpected flag x in regular expression literal
,-[1:20]
1 | const a = /\2(.)/uuxig;
: ^
2 | debugger;
`----
```
Closes#5177
While making this, I noticed an uncaught parse error for accessors: accessors cannot be optional. I'll add a fix for this in an up-stack PR.
Include "Output mismatch" in transform conformance snapshots when a test also fails due to scopes/symbols mismatches.
At present, many of the tests are failing on scopes/symbols mismatches, so it's hard to see from snapshots which also have an output mismatch. In particular, when working to fix scope mismatches, it's hard to notice if changes you make cause the output to be wrong. These extra messages make that visible.
Snapshots generated by transform conformance were missing a few errors (e.g. `declarations/const-enum/input.ts` at bottom of diff). Make sure all errors fail the test.
This reverts the previous changes to handling labels so that all tests can pass.
This passes all false postivies found in `monitor-oxc` (node_modules/flow-parser/flow_parser.js)
As it turns out this requires less code and produces better diagnostics.
Blocked by #5008
In the SemanticBuilder, we insert all CatchClause parameters in the BlockStatement scope, a child-scope of CatchClause
858f510d59/crates/oxc_semantic/src/builder.rs (L709-L718)
So we should do the same thing in this plugin, but because CatchClause has no parameters, SemanticBuilder doesn't create scope for `CatchClause`. This cause we cannot find the `BlockStatement` scope_id.
This PR has changed to the correct logic. Just to wait #5008 solved
Part of #5008. Make scopes for `ForStatement`, `ForInStatement` and `ForOfStatement` unconditional. i.e. always create a scope, even if there is no lexical binding (e.g. `for (i of a) {}`).
Transform checker check root unresolved references.
The transform checker is now checking pretty much everything it can.
Only fields of `ScopeTree` and `SymbolTable` that it's *not* checking are those which contain `AstNodeId`s, because transformer does not create node IDs at present:
* `ScopeTree::node_ids`
* `SymbolTable::declarations`
* `Reference::node_id`
Checking also only proceeds in "from AST" direction.
i.e. for each `SymbolId` which appears in the AST, we check everything about that symbol. But we *don't* go through all the "rows" in `SymbolTable` and check if there are any extra symbols in the table which aren't in the AST.
Presumably transformer will leave a lot of old symbols lying around in `SymbolTable` (ditto scopes and references). We'd need to add `ScopeFlags::Deleted`, `SymbolFlags::Deleted` and `ReferenceFlags::Deleted` for the transformer to be able to "delete" existing symbols.
