build and test against scala 3.7.1 #2227
Conversation
bjaglin
changed the title
bjaglin
changed the title
bjaglin
force-pushed
the
scala371
branch
2 times, most recently
from
a8674bd to
aa48bde
Compare
| val d = | ||
| if (true) Day.Weekday | ||
| else Day.Weekend | ||
| day(d) | ||
| val b = | ||
| if (true) True | ||
| else False |
There was a problem hiding this comment.
The test infrastructure is not great to represent the intent of that change, but this was copied in order to skip scala3-next.
That's because the 3.7.1-RC1 PC is trying to be smarter when shortening types but gets fooled by the import Bool. somehow , and we end up with that code that does not compile
| val d = | |
| if (true) Day.Weekday | |
| else Day.Weekend | |
| day(d) | |
| val b = | |
| if (true) True | |
| else False | |
| val d: Value = | |
| if (true) Day.Weekday | |
| else Day.Weekend | |
| day(d) | |
| val b: Value = | |
| if (true) True | |
| else False |
[error] -- [E007] Type Mismatch Error: /Users/brice.jaglin/git/scala/scalafix/scalafix-tests/output/src/main/scala-3next/test/explicitResultTypes/EnumerationValue.scala:15:18
[error] 15 | if (true) Day.Weekday
[error] | ^^^^^^^^^^^
[error] | Found: (test.explicitResultTypes.EnumerationValue.Day.Weekday :
[error] | test.explicitResultTypes.EnumerationValue.Day.Value)
[error] | Required: test.explicitResultTypes.EnumerationValue.Bool.Value
[error] |
[error] | longer explanation available when compiling with `-explain`
[error] -- [E007] Type Mismatch Error: /Users/brice.jaglin/git/scala/scalafix/scalafix-tests/output/src/main/scala-3next/test/explicitResultTypes/EnumerationValue.scala:16:13
[error] 16 | else Day.Weekend
[error] | ^^^^^^^^^^^
[error] | Found: (test.explicitResultTypes.EnumerationValue.Day.Weekend :
[error] | test.explicitResultTypes.EnumerationValue.Day.Value)
[error] | Required: test.explicitResultTypes.EnumerationValue.Bool.Value
[error] |
[error] | longer explanation available when compiling with `-explain`
[error] -- [E007] Type Mismatch Error: /Users/brice.jaglin/git/scala/scalafix/scalafix-tests/output/src/main/scala-3next/test/explicitResultTypes/EnumerationValue.scala:17:6
[error] 17 | day(d)
[error] | ^
[error] | Found: (test.explicitResultTypes.EnumerationValue.d :
[error] | test.explicitResultTypes.EnumerationValue.Bool.Value)
[error] | Required: test.explicitResultTypes.EnumerationValue.Day.Value
[error] |
[error] | longer explanation available when compiling with `-explain`
There was a problem hiding this comment.
I just went through the recent changes of the PC and scala/scala3#22898 might be a candidate for the regression.
@tgodzik since you are the author, maybe you can confirm/infirm? I will end up filing an issue upstream eventually, but my knowledge of the dotty repo is limited so it will take me a while to write a unit test :)
There was a problem hiding this comment.
It might be some side effect of that PR yeah. It was a tricky part of the compiler. If you could just put the example in a issue I will take a look.
There was a problem hiding this comment.
Ach, I think I know what is going on. Those are path dependent types and it's not handled at all properly :/
There was a problem hiding this comment.
I've got something working, but I broke another test:
https://github.com/scala/scala3/compare/main...tgodzik:scala3:fix-path-dep?expand=1
I am not sure how to differentiate types that are actually in Scope from those that are not if they actually have the same symbol.
There was a problem hiding this comment.
Ok, let's try get scala/scala3#23124 merged, took me longer than expected
There was a problem hiding this comment.
Thanks! Taking note to simplify the test once 3.7.2-RC1 is tagged
There was a problem hiding this comment.
bjaglin
force-pushed
the
scala371
branch
2 times, most recently
from
3423c8f to
221ddc0
Compare
bjaglin
changed the title
bjaglin
changed the title
2 participants
No description provided.