#2976. Add type inference tests. Part 2

[sgrekhov]

No description provided.

[eernstg]

Looks good, couple of comments.

One thing needs to be reconsidered: It's the context type schema of .id.a.b.c which is used to compute the static namespace where id is looked up, not the context type schema of .id.

[eernstg]

OK. I'm not sure dynamic as a context type schema is equivalent to _. But we should have the error anyway, and it's a delicate question, so let's have it.

[sgrekhov]

Proposal states that a context type schema in this case is _.

https://github.com/dart-lang/language/blob/1b0e339c52e73acc875bd68451b46cef9391f028/working/3616%20-%20enum%20value%20shorthand/proposal-simple-lrhn.md?plain=1#L306

[eernstg]

Interesting... I can't see that this has been specified anywhere, but the choice of context type schema in every given situation is highly underspecified anyway.

One source of confusion is that the implemented type inference algorithm chooses dynamic in a number of cases where the specification in inference.md says that the inferred result is the greatest closure. I created dart-lang/language#4197 to shed some light on that.

If we assume that this is fixed then I'd expect the following behavior:

X getContextType<X>(Object? o) {
  print(X);
  return o as X;
}

void main() {
  getContextType(1)..arglebargle; // Compile-time error.
  dynamic d = getContextType(1)..arglebargle; // No compile-time error.
}

This should illustrate that the context type schema _ and the context type dynamic is not the same thing. So we still need some clarifications, but I'd maintain that it isn't safe to assume that they are the same thing.

[sgrekhov]

Thank you. I misunderstood the spec at first. Now updated, PTAL

[sgrekhov]

Proposal states that a context type schema in this case is _.

https://github.com/dart-lang/language/blob/1b0e339c52e73acc875bd68451b46cef9391f028/working/3616%20-%20enum%20value%20shorthand/proposal-simple-lrhn.md?plain=1#L306

[eernstg]

LGTM!