Support re-aliasing of type alias instantiations

[ahejlsberg]

This PR continues the work in #42149 by supporting re-aliasing of type alias instantiations. For example:

typeThing<T>={value: T};typeArrayThing<U>=Thing<U[]>;typeStringArrayThing=ArrayThing<string>;declarelett1: Thing<string[]>;// Thing<string[]>declarelett2: ArrayThing<string>;// ArrayThing<string> (previously Thing<string[]>)declarelett3: StringArrayThing;// StringArrayThing (previously Thing<string[]>)interfaceItem<T>{value: T}typeArrayItem<U>=Item<U[]>;typeNumberArrayItem=ArrayItem<number>;declarelett4: Item<number[]>;// Item<number[]>declarelett5: ArrayItem<number>;// ArrayItem<number>declarelett6: NumberArrayItem;// NumberArrayItem (previously ArrayItem<number>)functionfoo<T,KextendskeyofT>(obj: T,key: K){letx: Omit<T,K>;// Omit<T, K> (previously Pick<T, Exclude<keyof T, K>>)}

Previously, instantiations of aliased anonymous object, mapped, or conditional types with a new type alias would continue to display the original type alias in diagnostics, quick info, and declaration files. Now, re-aliasing of such types consistently adopts and displays the new type alias.

In addition, instantiations of distributive types (such as homomorphic mapped types and distributive conditional types) now properly preserve the original type alias when the instantiation results in a union type:

typeA={a: string};typeB={b: string};declareletp: Partial<A|B>;// Partial<A | B> (previously Partial<A> | Partial<B>)

[ahejlsberg]

@typescript-bot test this
@typescript-bot user test this
@typescript-bot run dt
@typescript-bot perf test this faster

[typescript-bot]

Heya @ahejlsberg, I've started to run the abridged perf test suite on this PR at 8ac7030. You can monitor the build here.

Update: The results are in!

[typescript-bot]

Heya @ahejlsberg, I've started to run the parallelized community code test suite on this PR at 8ac7030. You can monitor the build here.

[typescript-bot]

Heya @ahejlsberg, I've started to run the extended test suite on this PR at 8ac7030. You can monitor the build here.

[typescript-bot]

Heya @ahejlsberg, I've started to run the parallelized Definitely Typed test suite on this PR at 8ac7030. You can monitor the build here.

[ahejlsberg]

@robpalme This is the continuation PR I mentioned in #42149. Give it a try and let me know how it goes.

[typescript-bot]

@ahejlsberg
The results of the perf run you requested are in!

Here they are:

Comparison Report - master..42284

Metric	master	42284	Delta	Best	Worst
Angular - node (v12.1.0, x64)
Memory used	324,053k (± 0.02%)	324,025k (± 0.02%)	-28k (- 0.01%)	323,871k	324,182k
Parse Time	1.98s (± 0.66%)	1.99s (± 0.84%)	+0.01s (+ 0.50%)	1.95s	2.01s
Bind Time	0.81s (± 0.92%)	0.82s (± 1.35%)	+0.00s (+ 0.49%)	0.81s	0.86s
Check Time	4.88s (± 0.33%)	4.88s (± 0.58%)	+0.01s (+ 0.21%)	4.82s	4.94s
Emit Time	5.51s (± 0.77%)	5.51s (± 0.61%)	-0.00s (- 0.02%)	5.46s	5.59s
Total Time	13.18s (± 0.43%)	13.21s (± 0.50%)	+0.03s (+ 0.21%)	13.06s	13.36s
Compiler-Unions - node (v12.1.0, x64)
Memory used	200,247k (± 0.08%)	200,294k (± 0.07%)	+47k (+ 0.02%)	199,982k	200,507k
Parse Time	0.78s (± 0.77%)	0.77s (± 0.98%)	-0.00s (- 0.00%)	0.76s	0.79s
Bind Time	0.50s (± 0.66%)	0.50s (± 1.04%)	-0.00s (- 0.20%)	0.49s	0.51s
Check Time	9.94s (± 1.06%)	9.92s (± 0.79%)	-0.02s (- 0.17%)	9.75s	10.15s
Emit Time	2.37s (± 0.97%)	2.36s (± 1.29%)	-0.01s (- 0.42%)	2.30s	2.46s
Total Time	13.58s (± 0.74%)	13.55s (± 0.70%)	-0.03s (- 0.23%)	13.34s	13.75s
Monaco - node (v12.1.0, x64)
Memory used	337,446k (± 0.03%)	337,473k (± 0.01%)	+27k (+ 0.01%)	337,335k	337,541k
Parse Time	1.59s (± 0.30%)	1.58s (± 0.56%)	-0.00s (- 0.31%)	1.57s	1.61s
Bind Time	0.70s (± 1.08%)	0.70s (± 0.64%)	-0.01s (- 0.99%)	0.69s	0.71s
Check Time	4.95s (± 0.46%)	4.94s (± 0.43%)	-0.01s (- 0.12%)	4.88s	4.98s
Emit Time	2.88s (± 0.60%)	2.89s (± 0.42%)	+0.01s (+ 0.38%)	2.86s	2.91s
Total Time	10.12s (± 0.24%)	10.11s (± 0.36%)	-0.01s (- 0.06%)	9.99s	10.16s
TFS - node (v12.1.0, x64)
Memory used	292,356k (± 0.02%)	292,396k (± 0.02%)	+41k (+ 0.01%)	292,282k	292,533k
Parse Time	1.25s (± 0.76%)	1.25s (± 0.71%)	-0.01s (- 0.64%)	1.23s	1.26s
Bind Time	0.65s (± 0.73%)	0.65s (± 0.86%)	-0.01s (- 0.77%)	0.64s	0.66s
Check Time	4.54s (± 0.70%)	4.52s (± 0.54%)	-0.02s (- 0.51%)	4.47s	4.58s
Emit Time	2.95s (± 0.89%)	2.94s (± 0.72%)	-0.00s (- 0.14%)	2.90s	2.98s
Total Time	9.40s (± 0.52%)	9.36s (± 0.42%)	-0.04s (- 0.37%)	9.28s	9.45s
material-ui - node (v12.1.0, x64)
Memory used	478,950k (± 0.07%)	473,781k (± 0.01%)	-5,169k (- 1.08%)	473,585k	473,886k
Parse Time	2.08s (± 0.73%)	2.07s (± 0.54%)	-0.02s (- 0.91%)	2.04s	2.09s
Bind Time	0.65s (± 0.73%)	0.64s (± 0.69%)	-0.00s (- 0.77%)	0.63s	0.65s
Check Time	12.49s (± 0.90%)	12.51s (± 0.71%)	+0.02s (+ 0.14%)	12.28s	12.72s
Emit Time	0.00s (± 0.00%)	0.00s (± 0.00%)	0.00s ( NaN%)	0.00s	0.00s
Total Time	15.23s (± 0.65%)	15.22s (± 0.62%)	-0.00s (- 0.03%)	14.98s	15.44s

System

Machine Name	ts-ci-ubuntu
Platform	linux 4.4.0-197-generic
Architecture	x64
Available Memory	16 GB
Available Memory	4 GB
CPUs	4 × Intel(R) Core(TM) i7-4770 CPU @ 3.40GHz

Hosts

• node (v12.1.0, x64)

Scenarios

• Angular - node (v12.1.0, x64)
• Compiler-Unions - node (v12.1.0, x64)
• Monaco - node (v12.1.0, x64)
• TFS - node (v12.1.0, x64)
• material-ui - node (v12.1.0, x64)

Benchmark	Name	Iterations
Current	42284	10
Baseline	master	10

[typescript-bot]

The user suite test run you requested has finished and failed. I've opened a PR with the baseline diff from master.

[ahejlsberg]

Tests all look good, only changes are improved diagnostics. Performance is unaffected.

[DanielRosenwasser]

@typescript-bot pack this

[typescript-bot]

Heya @DanielRosenwasser, I've started to run the tarball bundle task on this PR at 8ac7030. You can monitor the build here.

[typescript-bot]

Hey @DanielRosenwasser, I've packed this into an installable tgz. You can install it for testing by referencing it in your package.json like so:

{
"devDependencies": {
"typescript": "https://typescript.visualstudio.com/cf7ac146-d525-443c-b23c-0d58337efebc/_apis/build/builds/93139/artifacts?artifactName=tgz&fileId=4FE49FBE4115C3F38EC02DE524EF9087A6729DD9744C9B257007FE5D69F77D1202&fileName=/typescript-4.2.0-insiders.20210111.tgz"
}
}

and then running npm install.

---

There is also a playground for this build and an npm module you can use via "typescript": "npm:@typescript-deploys/pr-build@4.2.0-pr-42284-11".;

[DanielRosenwasser]

Why's this happening? What was avoiding the elaboration before?

[DanielRosenwasser]

Sounds like we can't avoid re-elaboration based on type IDs

[robpalme]

@ahejlsberg This is even better than the previous attempt 🎉

As before, there are no inlining/size regressions - I reviewed every changed line and this PR is a pure win. This PR improves upon 2c2d06d by also handling aliases to Record and Pick<obj, union>.

// Previous: 2c2d06
myRecordProp: Record<string,any>;
myObjectProp: Pick<{lotsOfProps,},"red"|"green"|"blue>;

...is replaced by...

// Master: Using PR#42284
myRecordProp: MyRecordType;
myObjectProp: MyColorsOnly<{lotsOfProps,}>;

On the large package I referred to in previousreports, this PR results in a further 10KB decrease vs the previous attempt, i.e the Record & Pick changes explains the 10KB win. Meaning the net effect of this PR is now:

• 50 declaration files out of 394 were affected by this PR
• Net size of all declaration files shrank by 41KB
  • 50 files shrank (total size of these files reduced by 41KB from 585KB to 544KB)
  • 0 files grew