Make IsStringUTF8 reject (U+FDD0 .. U+FDEF) ...

base/string_util_unittest.cc

 // Copyright (c) 2010 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include <math.h>
 #include <stdarg.h>
 
 #include <limits>
 #include <sstream>
 
@@ skipping 207 matching lines @@
   // BOMs in UTF-16(BE|LE) and UTF-32(BE|LE)
   EXPECT_FALSE(IsStringUTF8("\xfe\xff"));
   EXPECT_FALSE(IsStringUTF8("\xff\xfe"));
   EXPECT_FALSE(IsStringUTF8(std::string("\x00\x00\xfe\xff", 4)));
   EXPECT_FALSE(IsStringUTF8("\xff\xfe\x00\x00"));
 
   // Non-characters : U+xxFFF[EF] where xx is 0x00 through 0x10 and <FDD0,FDEF>
   EXPECT_FALSE(IsStringUTF8("\xef\xbf\xbe"));  // U+FFFE)
   EXPECT_FALSE(IsStringUTF8("\xf0\x8f\xbf\xbe"));  // U+1FFFE
   EXPECT_FALSE(IsStringUTF8("\xf3\xbf\xbf\xbf"));  // U+10FFFF
-
-  // This should also be false, but currently we pass them through.
-  // Disable them for now.
-#if 0
   EXPECT_FALSE(IsStringUTF8("\xef\xb7\x90"));  // U+FDD0
   EXPECT_FALSE(IsStringUTF8("\xef\xb7\xaf"));  // U+FDEF
-#endif
-
   // Strings in legacy encodings. We can certainly make up strings
   // in a legacy encoding that are valid in UTF-8, but in real data,
   // most of them are invalid as UTF-8.
   EXPECT_FALSE(IsStringUTF8("caf\xe9"));  // cafe with U+00E9 in ISO-8859-1
   EXPECT_FALSE(IsStringUTF8("\xb0\xa1\xb0\xa2"));  // U+AC00, U+AC001 in EUC-KR
   EXPECT_FALSE(IsStringUTF8("\xa7\x41\xa6\x6e"));  // U+4F60 U+597D in Big5
   // "abc" with U+201[CD] in windows-125[0-8]
   EXPECT_FALSE(IsStringUTF8("\x93" "abc\x94"));
   // U+0639 U+064E U+0644 U+064E in ISO-8859-6
   EXPECT_FALSE(IsStringUTF8("\xd9\xee\xe4\xee"));
@@ skipping 1245 matching lines @@
   EXPECT_FALSE(ContainsOnlyChars("Hello", ""));
 
   EXPECT_TRUE(ContainsOnlyChars("", "1234"));
   EXPECT_TRUE(ContainsOnlyChars("1", "1234"));
   EXPECT_TRUE(ContainsOnlyChars("1", "4321"));
   EXPECT_TRUE(ContainsOnlyChars("123", "4321"));
   EXPECT_FALSE(ContainsOnlyChars("123a", "4321"));
 }
 
 }  // namaspace base