Feat(rome_js_parser): typeof private property name

[IWANABETHATGUY]

Summary

part of #2400
typeof on private fields: Announcement

Test Plan

I copy the test case from official pull request https://github.com/microsoft/TypeScript/pull/47696/files#diff-c7d8064feb0c4b5ba2cb75b7987eefc92a880fded5f86a9564f5d30ffef48bff, and pass it.

[IWANABETHATGUY]

!bench_parser

[github-actions]

Parser Benchmark Results

group                                 main                                   pr
-----                                 ----                                   --
parser/checker.ts                     1.03    123.7±4.55ms    21.0 MB/sec    1.00    119.7±1.46ms    21.7 MB/sec
parser/compiler.js                    1.00     72.5±1.44ms    14.4 MB/sec    1.01     73.1±1.44ms    14.3 MB/sec
parser/d3.min.js                      1.00     46.6±0.44ms     5.6 MB/sec    1.00     46.4±0.28ms     5.7 MB/sec
parser/dojo.js                        1.00      4.4±0.00ms    15.8 MB/sec    1.00      4.4±0.00ms    15.7 MB/sec
parser/ios.d.ts                       1.01    105.1±1.61ms    17.8 MB/sec    1.00    104.1±0.86ms    17.9 MB/sec
parser/jquery.min.js                  1.00     12.7±0.02ms     6.5 MB/sec    1.00     12.8±0.03ms     6.5 MB/sec
parser/math.js                        1.02     88.8±1.46ms     7.3 MB/sec    1.00     87.0±1.07ms     7.4 MB/sec
parser/parser.ts                      1.00      3.0±0.00ms    16.0 MB/sec    1.00      3.0±0.00ms    16.0 MB/sec
parser/pixi.min.js                    1.01     55.3±0.72ms     7.9 MB/sec    1.00     54.6±0.58ms     8.0 MB/sec
parser/react-dom.production.min.js    1.00     17.5±0.03ms     6.6 MB/sec    1.00     17.6±0.03ms     6.6 MB/sec
parser/react.production.min.js        1.00    921.0±2.11µs     6.7 MB/sec    1.00    920.2±1.98µs     6.7 MB/sec
parser/router.ts                      1.00      2.5±0.01ms    24.1 MB/sec    1.00      2.5±0.00ms    24.1 MB/sec
parser/tex-chtml-full.js              1.01    116.1±1.11ms     7.8 MB/sec    1.00    114.4±1.26ms     8.0 MB/sec
parser/three.min.js                   1.02     62.9±0.89ms     9.3 MB/sec    1.00     61.7±0.39ms     9.5 MB/sec
parser/typescript.js                  1.02    491.8±7.43ms    19.3 MB/sec    1.00    480.6±8.04ms    19.8 MB/sec
parser/vue.global.prod.js             1.00     21.3±0.15ms     5.7 MB/sec    1.00     21.4±0.09ms     5.6 MB/sec

[IWANABETHATGUY]

@MichaReiser could you help with reviewing my pull request about the new typescript 4.7 syntax?

[MichaReiser]

I'm having concerns about adding JsAnyName to the right-hand side. This will not only allow private names in typeof expressions but anywhere else where TsQualifiedName is used (and none of these allow private names).

Can you go through all types that have a field with the type TsQualifiedName and check if TS now allows the usage of private names for these use cases? For example, TsImportTypeQualifier, TsNameWithTypeArguments, TsReferenceType, TsAnyModuleReference

Does TS support typeof this.#a.b?

[IWANABETHATGUY]

p

Ts support typeof this.#a.b , https://www.typescriptlang.org/play?#code/FAYwNghgzlAEAqsDexZtgYgrAvMiAXLAOTEC+q6ALgKZRUAUAlMpetQJ4AONA9gGawqACwCWUAHRYJEANxtYFMkA, but I don't have enough knowledge about Ts specification, so I can't write code snippet include the ast node you mentioned above.

[MichaReiser]

We'll need to understand the exact change of typescript to model the AST correctly.

Are you able to find the related PR in the TypeScript repository?

[IWANABETHATGUY]

microsoft/TypeScript#47696, this one

[MichaReiser]

Reading the code of this PR should help you understand which rules allow private names and which do not (allowPrivateIdentifiers parameter).

The next step is then to go to TypeScript's playground and try to come up with examples for the cases where private names are/aren't allowed.

[jakebailey]

We're reverting the linked change in TS, so you probably shouldn't merge this (we may not even allow this syntax).

[MichaReiser]

We're reverting the linked change in TS, so you probably shouldn't merge this (we may not even allow this syntax).

Thanks @jakebailey for letting us know!