Conversation
|
I've moved this to the 2.13.1 milestone since it isn't really a release blocker, but it would certainly be very good to have this fixed. |
|
|
af2347c to
22e8876
Compare
|
Very tough to reason about a stew of side effects. I had even forgotten who tells global what is the current run. (Answer is the current run.) This fix is to remove |
22e8876 to
ef48850
Compare
Emit the summary after successful parse in case "there were deprecations". To reduce repetitive parser warnings on multiline input, only reset the reporter after a final parse and also after final compilation result (a result that is not incomplete, but may represent either a compile error or success). The underlying global compiler will also reset the reporter before compilation of repl artifacts.
ef48850 to
0e6426d
Compare
dwijnand
left a comment
There was a problem hiding this comment.
The change sounds good and the diff looks reasonable, and all the tests are passing. I think it would be good to capture this as a test no? I don't exactly see the behaviour you describe in the transcript in the PR description in any of the changed tests.
|
The updated check files demonstrate parser-based deprecations warning is shown now. I'll see if a repl test will show the multiline case not warning on every line, see the comment from last Valentine's. |
|
With deprecation enabled: versus |
|
Note that |
Emit the summary "there were deprecations" after
successful parse.
Tweaks #7482
Previously, fatal warning trumped incomplete input, so switch that around. Previously:
Wait for complete input to issue summary.
Summary,
-deprecation,-Xfatal-warnings:Fixes scala/bug#11402