Intrinsify StringContext.f #9929
Conversation
|
This will unblock #9925 |
compiler/src/dotty/tools/dotc/transform/localopt/StringContextChecker.scala
Outdated
Show resolved
Hide resolved
| import dotty.tools.dotc.core.Types._ | ||
|
|
||
| // Ported from old dotty.internal.StringContextMacro | ||
| // TODO: port Scala 2 logic? (see https://github.com/scala/scala/blob/2.13.x/src/compiler/scala/tools/reflect/FormatInterpolator.scala#L74) |
There was a problem hiding this comment.
What does the Scala 2 logic do that is different?
There was a problem hiding this comment.
It is a completely different implementation. Not sure which one is better. Someone will have to check that at some point.
There was a problem hiding this comment.
scala 2 f goes overboard to verify the format, but is it worthwhile maintaining code for a slow string mechanism.
There was a problem hiding this comment.
As far as I remember this implementation covered all the Scala 2 f interpolator test cases. They diverged slightly on the second error after some kinds of error in the format part of the string.
There was a problem hiding this comment.
I thought I saw checks that weren't implemented (way in the past); if I have a chance, I'll compare again.
There was a problem hiding this comment.
We might have missed something. It would be ok for me to use the Scala 2 implementation but it would take a more effort to port. I could have a look in the future or let someone else try it. The current priority is to unblock #9925.
Ported the current macro implementation
nicolasstucki
force-pushed
the
intrinsify-StringContext.f
branch
from
14ae862 to
ef19bcd
Compare
compiler/src/dotty/tools/dotc/transform/localopt/StringContextChecker.scala
Outdated
Show resolved
Hide resolved
Ported the current macro implementation