More extension operators declaration work #78457
Conversation
AlekseyTs
added
Area-Compilers
Feature - Extension Everything
ghost
added
the
untriaged
|
@jcouv, @dotnet/roslyn-compiler For the second review |
|
|
||
| this.CheckUnsafeModifier(declarationModifiers, diagnostics); | ||
|
|
||
| if (isCompoundAssignmentOrIncrementAssignment) |
There was a problem hiding this comment.
Is this covered by test? #Closed
There was a problem hiding this comment.
Is this covered by test?
I am not sure what this question is about. There is no change at this position in the latest revision. Consider reviewing the latest revision. I had to do a merge to resolve a conflict. Apparently CodeFlow is now going crazy on merge commits and showing changes from them as undone in previous commits.
| case RefKind.In: | ||
| case RefKind.RefReadOnlyParameter: | ||
| // 'in' and 'ref readonly' receivers are disallowed for anything that is not a concrete struct (class or a type parameter) | ||
| if (extensionParameter.Type is { TypeKind: not TypeKind.TypeParameter, IsValueType: true }) |
| { | ||
| if (extensionParameter.Name == "") | ||
| { | ||
| diagnostics.Add(ErrorCode.ERR_InstanceMemberWithUnnamedExtensionsParameter, _location, Name); |
There was a problem hiding this comment.
Is this covered by test?
This is the line I was trying to point to (producing ERR_InstanceMemberWithUnnamedExtensionsParameter diagnostic) #Closed
| { | ||
| extension(ref S1? s1) | ||
| { | ||
| public void operator {{{op}}}() {} |
There was a problem hiding this comment.
nit: consider adding number comments (// N) to make it easier to review this long test. Same below in CompoundAssignment_001_Declaration #Closed
There was a problem hiding this comment.
consider adding number comments
I am using #line directives, not planning to use comments because they are creating a maintenance issues since some devs confuse them with a form of a base-line
| return true; // An error scenario | ||
| } | ||
|
|
||
| if (checkStrippedType && type.IsNullableType()) |
There was a problem hiding this comment.
We should keep track of this and include in a spec update. Also applies to the proposed validation rules (marked PROTOTYPE in other file) #Resolved
There was a problem hiding this comment.
We should keep track of this and include in a spec update
We certainly can re-confirm the behavior, but it follows the current speclet, which doesn't prevent declaration of operators for Nullable<T>. The other place (marked PROTOTYPE in other file) is adding restrictions that are not mentioned in the speclet
Relates to test plan #76130