I think, but am unsure, that this commit isn't necessary. Without it,
the parse pointer can point to an interior byte of a multi-byte
character. Recent work has sensitized me to trying to avoid this. So
this commit advances the pointer a full character width at a time.
RExC_parse++;
if (*RExC_parse == 'c') {
/* Skip the \cX notation for control characters */
- RExC_parse++;
+ RExC_parse += UTF ? UTF8SKIP(RExC_parse) : 1;
}
break;
case '[':