Optimize mb_strcut for text encodings with mblen_table

For legacy text encodings where mb_strcut is implemented using an
mblen_table (such as the various SJIS variants), mb_strcut is now
~30% faster on small strings (about 10 bytes). This is because we
are now avoiding an extra, unnecessary copy operation on the
output string.

When used on large strings, the difference in performance is
negligible, as almost the entire runtime is spent stepping through
the string to find the starting and ending cut points.

ext/mbstring/mbstring.c

@@ -2455,6 +2455,29 @@ PHP_FUNCTION(mb_strcut)
 		RETURN_STR(zend_string_init_fast((const char*)(string.val + from), len & -char_len));
 	}
 
+	if (enc->mblen_table) {
+		const unsigned char *mbtab = enc->mblen_table;
+		const unsigned char *p, *q, *end;
+		int m = 0;
+		/* Search for start position */
+		for (p = (const unsigned char*)string.val, q = p + from; p < q; p += (m = mbtab[*p]));
+		if (p > q) {
+			p -= m;
+		}
+		const unsigned char *start = p;
+		/* Search for end position */
+		if (len >= string.len - (start - (const unsigned char*)string.val)) {
+			end = (const unsigned char*)(string.val + string.len);
+		} else {
+			for (q = p + len; p < q; p += (m = mbtab[*p]));
+			if (p > q) {
+				p -= m;
+			}
+			end = p;
+		}
+		RETURN_STR(zend_string_init_fast((const char*)start, end - start));
+	}
+
 	ret = mbfl_strcut(&string, &result, from, len);
 	ZEND_ASSERT(ret != NULL);
 	RETVAL_STRINGL((char *)ret->val, ret->len); /* the string is already strdup()'ed */