akr	2007-12-09 12:50:11 +0900 (Sun, 09 Dec 2007)

  New Revision: 14154

  Modified files:
    trunk/ChangeLog
    trunk/re.c
    trunk/test/ruby/test_m17n.rb

  Log:
    * re.c (append_utf8): check unicode range.


  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/ChangeLog?r1=14154&r2=14153
  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/re.c?r1=14154&r2=14153
  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/test/ruby/test_m17n.rb?r1=14154&r2=14153

Index: re.c
===================================================================
--- re.c	(revision 14153)
+++ re.c	(revision 14154)
@@ -1425,9 +1425,22 @@
 }
 
 static int
+check_unicode_range(unsigned long code, onig_errmsg_buffer err)
+{
+    if ((0xd800 <= code && code <= 0xdfff) || /* Surrogates */
+        0x10ffff < code) {
+        strcpy(err, "invalid Unicode range");
+        return -1;
+    }
+    return 0;
+}
+
+static int
 append_utf8(unsigned long uv,
         VALUE buf, rb_encoding **encp, onig_errmsg_buffer err)
 {
+    if (check_unicode_range(uv, err) != 0)
+        return -1;
     if (uv < 0x80) {
         char escbuf[5];
         snprintf(escbuf, sizeof(escbuf), "\\x%02x", (int)uv);
@@ -1468,10 +1481,6 @@
             strcpy(err, "invalid Unicode range");
             return -1;
         }
-        if (0x10ffff < code) {
-            strcpy(err, "invalid Unicode range");
-            return -1;
-        }
         p += len;
         if (append_utf8(code, buf, encp, err) != 0)
             return -1;
Index: ChangeLog
===================================================================
--- ChangeLog	(revision 14153)
+++ ChangeLog	(revision 14154)
@@ -1,3 +1,7 @@
+Sun Dec  9 12:49:34 2007  Tanaka Akira  <akr@f...>
+
+	* re.c (append_utf8): check unicode range.
+
 Sun Dec  9 12:39:01 2007  Nobuyoshi Nakada  <nobu@r...>
 
 	* lib/cgi.rb (read_multipart): exclude blanks from header values.
Index: test/ruby/test_m17n.rb
===================================================================
--- test/ruby/test_m17n.rb	(revision 14153)
+++ test/ruby/test_m17n.rb	(revision 14154)
@@ -446,6 +446,16 @@
     #assert_raise(SyntaxError) { s1, s2 = u('\xc2'), u('\xa1'); /#{s1}#{s2}/ }
   end
 
+  def test_regexp_unicode
+    assert_nothing_raised { eval '/\u{0}/' }
+    assert_nothing_raised { eval '/\u{D7FF}/' }
+    assert_raise(SyntaxError) { eval '/\u{D800}/' }
+    assert_raise(SyntaxError) { eval '/\u{DFFF}/' }
+    assert_nothing_raised { eval '/\u{E000}/' }
+    assert_nothing_raised { eval '/\u{10FFFF}/' }
+    assert_raise(SyntaxError) { eval '/\u{110000}/' }
+  end
+
   def test_regexp_mixed_unicode
     assert_raise(SyntaxError) { eval(a(%{/\xc2\xa0\\u{6666}/})) }
     assert_raise(SyntaxError) { eval(e(%{/\xc2\xa0\\u{6666}/})) }

--
ML: ruby-changes@q...
Info: http://www.atdot.net/~ko1/quickml