Re: [Fwd: Patch for Recovery from Bad Multibyte Characters]

[Jason Merrill <jason at cygnus dot com>]

>>>>> Jeffrey A Law <law@cygnus.com> writes:

 >   In message <3752D581.C2ED469B@cygnus.com>you write:
 >> This patch correct a problem in the C and C++ compilers in which the
 >> compiler sometimes misses the end of a string or character literal
 >> while attempting to complete a multibyte character. It replaces all
 >> but the first character back into the input stream. In order to do
 >> this, I had to add the capability to put back more than one character
 >> to both the C and C++ front ends. I was careful to make sure that the
 >> overhead to sub_getch is still the same as before for the case where
 >> no character is in the buffer.

 > Well, you know more about this code than anyone else.  If you're happy
 > with it, I see no reason to object.

 > I don't think SJIS bugs are critical for the gcc-2.95 release, so
 > installing it into just the mainline sources ought to be OK.

Contrary to what Dave says, this problem occurs with any code that uses
high-ASCII values if gcc is configured with --enable-c-mbchar.  It doesn't
depend on the value of LANG at all.

Jason