arphaman added a comment.
In https://reviews.llvm.org/D25520#568448, @akyrtzi wrote:
> What if the user just wants to invoke the block, this is as common or more,
> The assign literal completion is useful but it should be an additional entry
> (with maybe lower priority) not replace the property completion.
> BTW, it would be great if we had completion as if it was a function call, so
> 2 entries, 1 for block invocation and 1 for assigning literal. Absence the
> block invocation call it should be 1 for normal property completion and 1 for
> assigning literal.
You're right, I forgot about the most common use of block properties. I will
update the patch today.
It makes sense to complete block invocation as well, but that has to be done in
a separate patch. I suppose that completion for block invocations should be
provided for normal sub expressions as well then.
cfe-commits mailing list