[Bug preprocessor/11439] New: CPP insists on UTF-8 input files

[zack at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11439

           Summary: CPP insists on UTF-8 input files
           Product: gcc
           Version: 3.4
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P1
         Component: preprocessor
        AssignedTo: zack at gcc dot gnu dot org
        ReportedBy: zack at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: n/a
  GCC host triplet: *-*-*
GCC target triplet: n/a

My patch to implement selectable execution character sets has the unfortunate
side effect of forcing users to write source code in either plain ASCII or
UTF-8.  There are a lot of source files out there in other encodings.  GCC needs
to support runtime, per-file selection of the input character set.

This problem affects the following test cases already in GCC:
* gcc/testsuite/gcc.c-torture/execute/wchar_t-1.c
* libstdc++-v3/testsuite/22_locale/collate/compare/wchar_t/2.cc
* libstdc++-v3/testsuite/22_locale/collate/compare/wchar_t/wrapped_env.cc
* libstdc++-v3/testsuite/22_locale/collate/compare/wchar_t/wrapped_locale.cc
* libstdc++-v3/testsuite/22_locale/collate/hash/wchar_t/2.cc
* libstdc++-v3/testsuite/22_locale/collate/hash/wchar_t/wrapped_env.cc
* libstdc++-v3/testsuite/22_locale/collate/hash/wchar_t/wrapped_locale.cc
* libstdc++-v3/testsuite/22_locale/collate/transform/wchar_t/2.cc
* libstdc++-v3/testsuite/22_locale/collate/transform/wchar_t/wrapped_env.cc
* libstdc++-v3/testsuite/22_locale/collate/transform/wchar_t/wrapped_locale.cc

which have therefore been marked XFAIL with the patch.  All contain characters
encoded in ISO 8859.1.