From: nobu.nakada@...
Date: 2004-03-06T11:05:38+09:00
Subject: [ruby-dev:23150] multidigit hex char in leteral

なかだです。

Onigurumaもimportされたことだし、リテラル中の\x{}を有効にしては
どうかと思うのですが。

  $ ./ruby -v -e "p Regexp.new('\x{33}')"
  ruby 1.9.0 (2004-03-06) [i686-linux]
  /\x{33}/

  $ ./ruby -v -e "p /\x{33}/"
  ruby 1.9.0 (2004-03-06) [i686-linux]
  -e:1: warning: ambiguous first argument; put parentheses or even spaces
  -e:1: Invalid escape character syntax
  p /\x{33}/
       ^
  -e:1: unterminated string meets end of file
  -e:1: parse error
  p /\x{33}/
       ^
  -e:1: warning: useless use of a literal in void context



* parse.y (read_escape, tokadd_escape, tokadd_string): allow multiple
  hexadecimal escape.


Index: parse.y
===================================================================
RCS file: /cvs/ruby/src/ruby/parse.y,v
retrieving revision 1.317
diff -u -2 -p -d -r1.317 parse.y
--- parse.y	3 Mar 2004 04:55:31 -0000	1.317
+++ parse.y	6 Mar 2004 02:00:37 -0000
@@ -2720,4 +2720,15 @@ newtok()
 
 static void
+tokspace(n)
+    int n;
+{
+    tokidx += n;
+    if (tokidx >= toksiz) {
+	do {toksiz *= 2;} while (toksiz < tokidx);
+	REALLOC_N(tokenbuf, char, toksiz);
+    }
+}
+
+static void
 tokadd(c)
     char c;
@@ -2731,4 +2742,61 @@ tokadd(c)
 
 static int
+tokadd_hex()
+{
+    int num, n, i, c;
+    char *s;
+
+    if (!peek('{')) {
+	c = scan_hex(lex_p, 2, &num);
+	if (num) tokadd(c);
+	return num;
+    }
+
+    s = lex_p;
+    do {
+	if (++s >= lex_pend) return 0;
+    } while (ISXDIGIT(*s));
+    if (*s != '}') return 0;
+    num = s - ++lex_p;
+    s = lex_p;
+    lex_p += num + 1;		/* skip {...} */
+    if (!num) return 0;
+    i = tokidx;
+    tokspace((num + 1) >> 1);
+    n = num;
+    if (n & 1) {
+	tokenbuf[i++] = scan_hex(s++, 1, &c);
+	if (!--n) return num;
+    }
+    do {
+	tokenbuf[i++] = scan_hex(s, 2, &c);
+    } while (s += 2, (n -= 2) > 0);
+    return num;
+}
+
+static int
+tok_hex(numlen)
+    int *numlen;
+{
+    int c, n;
+    char *s, *p;
+
+    if (!peek('{')) return scan_hex(lex_p, 2, numlen);
+    s = lex_p + 1;
+    p = memchr(s, '}', lex_pend - s);
+    if (!p || (n = p - s) == 0) {
+	*numlen = 0;
+	return 0;
+    }
+    c = scan_hex(s, n, numlen);
+    if (*numlen != n) {
+	*numlen = 0;
+	return 0;
+    }
+    *numlen += 2;
+    return c;
+}
+
+static int
 read_escape()
 {
@@ -2775,5 +2843,5 @@ read_escape()
 	    int numlen;
 
-	    c = scan_hex(lex_p, 2, &numlen);
+	    c = tok_hex(&numlen);
 	    if (numlen == 0) {
 		yyerror("Invalid escape character syntax");
@@ -2864,5 +2932,5 @@ tokadd_escape(term)
 	    tokadd('\\');
 	    tokadd(c);
-	    scan_hex(lex_p, 2, &numlen);
+	    tok_hex(&numlen);
 	    if (numlen == 0) {
 		yyerror("Invalid escape character syntax");
@@ -3034,4 +3102,8 @@ tokadd_string(func, term, paren, nest)
 		}
 		else if (func & STR_FUNC_EXPAND) {
+		    if (c == 'x') {
+			if (!tokadd_hex()) return -1;
+			continue;
+		    }
 		    pushback(c);
 		    if (func & STR_FUNC_ESCAPE) tokadd('\\');


-- 
--- 僕の前にBugはない。
--- 僕の後ろにBugはできる。
    中田 伸悦