[cfe-commits] [patch] Add support for UCNs in character literals

[Nico Weber]

Hi,

the attached patch adds support for UCN notation to character
literals, e.g. L'\u1234'. It splits EncodeUCNEscape() into two pieces
and calls only the first piece for character literals.

Note that overflow (e.g. L'\U0010fffe' with -fshort-wchar) is handled
differently than in gcc, because the gcc behavior makes no sense to
me, and I guess this is implementation-defined anyhow and not used in
practice.

Please review.

Thanks,
Nico
-------------- next part --------------
A non-text attachment was scrubbed...
Name: clang-ucn-char-literals.patch
Type: application/octet-stream
Size: 9745 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20101006/20ecffcf/attachment.obj>