[xemacs-hg @ 2005-02-28 23:36:30 by aidan] Add details about losing data to the unicode-to-char docstring. The Mule character encoding model means that extant Unicode coding systems will lose data. Make this clear in the docstring of unicode-to-char, give sample code to implement support for some desired code points should our distributors want to do that.

while (<STDIN>) {
  if (/^\$Revision:\s*([0-9\.]*)/) {
    print "char *version = \"$1\";\n";
    print "static char *id = \"\\n%%% setup-version $1\\n\";\n";
  }
}