aprantl added a comment.
> I didn't realize that the string int [] is produced by ValueObject itself; I
> think this makes this option more palatable to me. I'll give it a try.
So, it turns out it isn't. The only way to get the length into the typename is
to hack ClangASTContext::GetTypeName to replace the training "[]" of what clang
returns as the typename with a different string. The main problem with this is
that this will only work for outermost types.
Something that has been requested in the past is to support C structs with
trailing array members, such as
struct variable_size {
unsigned length;
int __attribute__((size=.length)) elements[]; // I just made up this
attribute, but you get the basic idea.
};
in a similar fashion. When printing such a struct, there's no way of safely
injecting the size into array type string any more.
If we dynamically created the correctly-sized array type instead, this would
work just fine.
I haven't yet understood the motivation behind why overriding
GetNumChildren/GetTypeName/GetChildAtIndex is preferable over creating a
dynamic type in the language runtime. Is there something that I need to know?
https://reviews.llvm.org/D53530
_______________________________________________
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
