All host names with nonascii characters (cyrillic) + escapable characters (,)...

googleurl/src/url_canon_unittest.cc

 // Copyright 2007, Google Inc.
 // All rights reserved.
 //
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 // notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
@@ skipping 81 matching lines @@
   const char* port;
   const char* path;
   const char* query;
   const char* ref;
   const char* expected;
 };
 
 // Wrapper around a UConverter object that managers creation and destruction.
 class UConvScoper {
  public:
-  UConvScoper(const char* charset_name) {
+  explicit UConvScoper(const char* charset_name) {
     UErrorCode err = U_ZERO_ERROR;
     converter_ = ucnv_open(charset_name, &err);
   }
 
   ~UConvScoper() {
     if (converter_)
       ucnv_close(converter_);
   }
 
   // Returns the converter object, may be NULL.
@@ skipping 118 matching lines @@
       // Non-BMP UTF-8.
     {L"!\xd800\xdf00!", "utf-8", "!\xf0\x90\x8c\x80!"},
       // Big5
     {L"\x4f60\x597d", "big5", "\xa7\x41\xa6\x6e"},
       // Unrepresentable character in the destination set.
     {L"hello\x4f60\x06de\x597dworld", "big5", "hello\xa7\x41%26%231758%3B\xa6\x6eworld"},
   };
 
   for (size_t i = 0; i < ARRAYSIZE(icu_cases); i++) {
     UConvScoper conv(icu_cases[i].encoding);
-    ASSERT_TRUE(conv.converter());
+    ASSERT_TRUE(conv.converter() != NULL);
     url_canon::ICUCharsetConverter converter(conv.converter());
 
     std::string str;
     url_canon::StdStringCanonOutput output(&str);
 
     string16 input_str(WStringToUTF16(icu_cases[i].input));
     int input_len = static_cast<int>(input_str.length());
     converter.ConvertFromUTF16(input_str.c_str(), input_len, &output);
     output.Complete();
 
@@ skipping 139 matching lines @@
     {"%3g%78%63%30%2e%30%32%35%30%2E.01", L"%3g%78%63%30%2e%30%32%35%30%2E.01", "%253gxc0.0250..01", url_parse::Component(0, 17), CanonHostInfo::BROKEN, -1},
       // Something that isn't exactly an IP should get treated as a host and
       // spaces escaped.
     {"192.168.0.1 hello", L"192.168.0.1 hello", "192.168.0.1%20hello", url_parse::Component(0, 19), CanonHostInfo::NEUTRAL, -1},
       // Fullwidth and escaped UTF-8 fullwidth should still be treated as IP.
       // These are "0Xc0.0250.01" in fullwidth.
     {"\xef\xbc\x90%Ef%bc\xb8%ef%Bd%83\xef\xbc\x90%EF%BC%8E\xef\xbc\x90\xef\xbc\x92\xef\xbc\x95\xef\xbc\x90\xef\xbc%8E\xef\xbc\x90\xef\xbc\x91", L"\xff10\xff38\xff43\xff10\xff0e\xff10\xff12\xff15\xff10\xff0e\xff10\xff11", "192.168.0.1", url_parse::Component(0, 11), CanonHostInfo::IPV4, 3},
       // Broken IP addresses get marked as such.
     {"192.168.0.257", L"192.168.0.257", "192.168.0.257", url_parse::Component(0, 13), CanonHostInfo::BROKEN, -1},
     {"[google.com]", L"[google.com]", "[google.com]", url_parse::Component(0, 12), CanonHostInfo::BROKEN, -1},
+      // Cyrillic letter followed buy ( should return punicode for ( escaped before punicode string was created. I.e.
+      // if ( is escaped after punicode is created we would get xn--%28-8tb (incorrect).
+    {"\xd1\x82(", L"\x0442(", "xn--%28-7ed", url_parse::Component(0, 11), CanonHostInfo::NEUTRAL, -1},
   };
 
   // CanonicalizeHost() non-verbose.
   std::string out_str;
   for (size_t i = 0; i < arraysize(host_cases); i++) {
     // Narrow version.
     if (host_cases[i].input8) {
       int host_len = static_cast<int>(strlen(host_cases[i].input8));
       url_parse::Component in_comp(0, host_len);
       url_parse::Component out_comp;
@@ skipping 1499 matching lines @@
   url_canon::StdStringCanonOutput repl_output(&repl_str);
   url_canon::ReplaceFileURL(src, parsed, repl, NULL, &repl_output, &repl_parsed);
   repl_output.Complete();
 
   // Generate the expected string and check.
   std::string expected("file:///foo?");
   for (size_t i = 0; i < new_query.length(); i++)
     expected.push_back('a');
   EXPECT_TRUE(expected == repl_str);
 }