Provide requested type as factory param

[charabaruk]

For RESOLVE_OPEN_GENERICS add a new "__requestedType" parameter when calling a factory method providing the type object for the type being resolved. This allows factory methods for registered open generic types to properly resolve the particular closed generic type requested.

[niemyjski]

Could you please add a test for this

[charabaruk]

I'll see what I can do but it probably won't be for a few days given my work schedule.

[charabaruk]

@niemyjski I've added a test for __requestedType. In commit ee3c4cb I've also turned on RESOLVE_OPEN_GENERICS tests in the test project, however Resolve_RegisteredOpenGenericInParent_CanBeResolvedByChild() which was failing before this work continues to fail so should be ignored in test results.

[niemyjski]

Thanks for the updates. Is there any way we can resolve the failing tests, they were all green before this but I get enabling a flag may break others but we need to keep the build passing.

[charabaruk]

I see two options:

1. Merging Fix child resolve for types with generic interfaces #137 and resolving the merge conflict between it and this PR in this PR's favour
2. Rolling back ee3c4cb, so that the RESOLVE_OPEN_GENERICS tests aren't run

I have no problem with bumping this PR's branch back to b8f7bd2 and providing a separate PR for enabling the RESOLVE_OPEN_GENERICS tests. That other PR would by necessity include #137's commits within it though, due to the conflict that would otherwise occur if both that and this are merged together.

Let me know which way we should take this.

[niemyjski]

I'd like to go with #1, is there any chance you could fork his fork and submit a new pr with the requested changes?

[charabaruk]

I'll give it a shot during my lunch break today.

[charabaruk]

@niemyjski I just created #141 which is the same as #137 with commit ee3c4cb cherry-picked onto it. That takes care of moving RESOLVE_OPEN_GENERICS to the project file, but I can't answer your other concern with #137. After it's merged in, I can rebase this PR's branch and it should pass all tests (including the new one for its functionality).

[niemyjski]

Thanks for the update! I just merged it in, let me know when you rebase and this is good to be merged in.

[charabaruk]

@niemyjski and we are good to go!

[niemyjski]

Thanks for the pr

[khanyuriy]

why this is only for conditional compilation?
if we delegate service location dont we need to tell what specific type was requested ?

[niemyjski]

@khanyuriy Can you please elaborate further