fix(js_parser): desambigaute JSX and const type param of arrow function

[Conaclos]

Summary

Fix #846.

This fixes a parsing error that rejected the following valid code when TypeScript and JSX are both enabled:

<const T,>() => {};

See these two comments for some context:

• ☂️ Prettier Divergence collection from the Prettier challenge #739 (comment)
• ☂️ Prettier Divergence collection from the Prettier challenge #739 (comment)

Test Plan

I added tests and updated snapshots.

[github-actions]

Parser conformance results on

js/262

Test result	main count	This PR count	Difference
Total	49701	49701	0
Passed	48721	48721	0
Failed	980	980	0
Panics	0	0	0
Coverage	98.03%	98.03%	0.00%

jsx/babel

Test result	main count	This PR count	Difference
Total	40	40	0
Passed	37	37	0
Failed	3	3	0
Panics	0	0	0
Coverage	92.50%	92.50%	0.00%

symbols/microsoft

Test result	main count	This PR count	Difference
Total	6322	6322	0
Passed	2036	2036	0
Failed	4286	4286	0
Panics	0	0	0
Coverage	32.20%	32.20%	0.00%

ts/babel

Test result	main count	This PR count	Difference
Total	662	662	0
Passed	592	592	0
Failed	70	70	0
Panics	0	0	0
Coverage	89.43%	89.43%	0.00%

ts/microsoft

Test result	main count	This PR count	Difference
Total	17646	17646	0
Passed	13454	13453	❌ ⏬ -1
Failed	4190	4191	❌ ⏫ +1
Panics	2	2	0
Coverage	76.24%	76.24%	-0.01%

🔥 Regression (1):

compiler/parseJsxExtends2.ts

[denbezrukov]

nit:
You can use p.nth_at here.

[Conaclos]

Thanks for the suggestion. By the way, why do we have this helper? Using nth(<n>) = <kind> looks cleaner?

[denbezrukov]

Certainly, we can discuss this. I'm inclined to use one specific approach. However, I was under the impression that using nth_at would be preferable. I'm open to hearing your thoughts on this.

p.nth(T![abstract]) && p.nth(T![class])
p.cur() == T![abstract] && p.cur() == T![class]

p.nth_at(n, T![abstract]) && p.nth_at(n + 1, T![class])
p.nth(n) == T![abstract] && p.nth(n + 1) == T![class]

matches!(p.nth(n), T![abstract] | T![const]) && matches!(p.nth(n + 1), T![class] | T![const)
p.nth_at_ts(n, FIRST_SET) && p.nth_at_ts(n + 1, SECOND_SET)

[Conaclos]

Yeah for consistency I updated the code. However, I think we should remove this helper: we actually end with more characters and a layer of indirection (in a future PR)...

[ematipico]

@Conaclos Is the regression expected?

[Conaclos]

Is the regression expected?

No. I updated the code to avoid the regression. However, the post seems not updated yet?

[Conaclos]

@ematipico Taking a closer look to the regression it is expected. In fact TypeScript expects an error because the tag T is not declared. We previously emitted an error because we were expecting an arrow function. Now we correctly parse it to a JSX element.