> Surly the only space that needs concern is the "ascii"/U+0020 space, > as that's used as a separator, not all the other unicode "space" > characters. ?
rc also treats tab and newline specially, so ascii space is not the
only whitespace character that needs quoting.
; grep -n '\\t' *
lex.c:11: return !strchr("\n \t#;&|^$=`'{}()<>", c) && c!=EOF;
lex.c:118: if(c==' ' || c=='\t')
simple.c:246: while(*s==' ' || *s=='\t' || *s=='\n') s++;
