Conversation
a7a9bcf to
3078531
Compare
| interface I { | ||
| struct S { S s; } | ||
| function f(E storage e) { | ||
| error E; | ||
| emit E(); | ||
| ++c; | ||
| uint calldata c = 123.4; | ||
| } | ||
| } |
There was a problem hiding this comment.
What's the point of this test?
I believe error E; as a statement in a function should now cause a parser error.
If this test is supposed to get past the parsing stage then we need to remove this statement.
There was a problem hiding this comment.
What's the point of this test?
I think we discussed it in chat, but yes, the point is to test the behaviour of the parser without the subsequent analysis step (that would absolutely fail all of these).
If this test is supposed to get past the parsing stage then we need to remove this statement.
Yup, that's exactly right.
This is an updated version of the original commit from 2021.
3078531 to
b7e1db6
Compare
|
I think it would be nice for us to have a protocol in place when opening PRs the target the breaking branch; e.g. prefix PRs with |
That's a good idea. Sorry, I got distracted and did not label this PR as breaking. |
error from directive to keyword.error from directive to keyword.
This is an updated version of the original PR #11218.
More context:
Plans for keywords in 0.9: #14770
Discussions specifically about
erroras keyword: #11743, #11859, #14770 (comment).