]> granicus.if.org Git - clang/commit
PR12717: Clang supports hexadecimal floating-point literals in all language
authorRichard Smith <richard-llvm@metafoo.co.uk>
Fri, 15 Jun 2012 05:07:49 +0000 (05:07 +0000)
committerRichard Smith <richard-llvm@metafoo.co.uk>
Fri, 15 Jun 2012 05:07:49 +0000 (05:07 +0000)
commitd2e95d1538ff91fe902464f02f83429f96117af5
tree3cfb21b12e22e2576fdb70b1e8a070891344b563
parent0982205bade2fb4fc984c27b2ab401e683963b10
PR12717: Clang supports hexadecimal floating-point literals in all language
modes. For languages other than C99/C11, this isn't quite a conforming
extension, and for C++11, it breaks some reasonable code containing
user-defined literals.

In languages which don't officially have hexfloats, pare back this extension
to only apply in cases where the token starts 0x and does not contain an
underscore. The extension is still not quite conforming, but it's a lot closer
now.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@158487 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Lex/Lexer.cpp
test/Lexer/c90.c
test/Lexer/hexfloat.cpp