Ken Dyck wrote:
On Wednesday, November 25, 2009 4:09 PM, John McCall wrote:
Ken Dyck wrote:
Okay. I'll take a stab at implementing this.
As I understand it, these are the changes that need to happen:
[snipped]
Is this what you had in mind?
I was thinking more along the lines of a type that represented a
size/offset in bytes; call it bytes_t for the sake of argument.
[snipped]
Okay. As a nested class in ASTContext? Or as a stand-alone class?
Standalone and in its own header file.
Since we're making changes to the API anyways, what do you think of the
units being bytes? The getTypeSizeInBytes() method returns the size of
the type in characters, so it would be more accurate to call the units
chars_t. And if we're doing that, we might as well change the method
names to 'getTypeSizeInChars()' and 'getCharWidth()', too.
I agree that 'chars' would be way better than 'bytes'. For consistency
with LLVM, the type name should probably be in CamelCase; whether that
name is 'Chars' or 'CharCount' or 'CharUnits' or 'CharWidth', I don't
really care; use your judgment, and hopefully this won't get
bike-shedded. :)
John.
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
