Support floating point literals with exponent but no point

GCC allows literals of the form "1e-16f" which clang currently rejects. The following patch is proposed:

Index: Lex/LiteralSupport.cpp

GCC allows literals of the form "1e-16f" which clang currently
rejects. The following patch is proposed:

Looks good to me, please commit. Also, please verify that the constants come out right in the -emit-llvm or ast-dump output, thanks Christopher!

-Chris