url/url_util_unittest.cc

   1 // Copyright 2013 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4
   5 #include "testing/gtest/include/gtest/gtest.h"
   6 #include "url/url_canon.h"
   7 #include "url/url_canon_stdstring.h"
   8 #include "url/url_parse.h"
   9 #include "url/url_test_utils.h"
  10 #include "url/url_util.h"
  11
  12 namespace url {
  13
  14 TEST(URLUtilTest, FindAndCompareScheme) {
  15   Component found_scheme;
  16
  17   // Simple case where the scheme is found and matches.
  18   const char kStr1[] = "http://www.com/";
  19   EXPECT_TRUE(FindAndCompareScheme(
  20       kStr1, static_cast<int>(strlen(kStr1)), "http", NULL));
  21   EXPECT_TRUE(FindAndCompareScheme(
  22       kStr1, static_cast<int>(strlen(kStr1)), "http", &found_scheme));
  23   EXPECT_TRUE(found_scheme == Component(0, 4));
  24
  25   // A case where the scheme is found and doesn't match.
  26   EXPECT_FALSE(FindAndCompareScheme(
  27       kStr1, static_cast<int>(strlen(kStr1)), "https", &found_scheme));
  28   EXPECT_TRUE(found_scheme == Component(0, 4));
  29
  30   // A case where there is no scheme.
  31   const char kStr2[] = "httpfoobar";
  32   EXPECT_FALSE(FindAndCompareScheme(
  33       kStr2, static_cast<int>(strlen(kStr2)), "http", &found_scheme));
  34   EXPECT_TRUE(found_scheme == Component());
  35
  36   // When there is an empty scheme, it should match the empty scheme.
  37   const char kStr3[] = ":foo.com/";
  38   EXPECT_TRUE(FindAndCompareScheme(
  39       kStr3, static_cast<int>(strlen(kStr3)), "", &found_scheme));
  40   EXPECT_TRUE(found_scheme == Component(0, 0));
  41
  42   // But when there is no scheme, it should fail.
  43   EXPECT_FALSE(FindAndCompareScheme("", 0, "", &found_scheme));
  44   EXPECT_TRUE(found_scheme == Component());
  45
  46   // When there is a whitespace char in scheme, it should canonicalize the url
  47   // before comparison.
  48   const char whtspc_str[] = " \r\n\tjav\ra\nscri\tpt:alert(1)";
  49   EXPECT_TRUE(FindAndCompareScheme(whtspc_str,
  50                                    static_cast<int>(strlen(whtspc_str)),
  51                                    "javascript", &found_scheme));
  52   EXPECT_TRUE(found_scheme == Component(1, 10));
  53
  54   // Control characters should be stripped out on the ends, and kept in the
  55   // middle.
  56   const char ctrl_str[] = "\02jav\02scr\03ipt:alert(1)";
  57   EXPECT_FALSE(FindAndCompareScheme(ctrl_str,
  58                                     static_cast<int>(strlen(ctrl_str)),
  59                                     "javascript", &found_scheme));
  60   EXPECT_TRUE(found_scheme == Component(1, 11));
  61 }
  62
  63 TEST(URLUtilTest, ReplaceComponents) {
  64   Parsed parsed;
  65   RawCanonOutputT<char> output;
  66   Parsed new_parsed;
  67
  68   // Check that the following calls do not cause crash
  69   Replacements<char> replacements;
  70   replacements.SetRef("test", Component(0, 4));
  71   ReplaceComponents(NULL, 0, parsed, replacements, NULL, &output, &new_parsed);
  72   ReplaceComponents("", 0, parsed, replacements, NULL, &output, &new_parsed);
  73   replacements.ClearRef();
  74   replacements.SetHost("test", Component(0, 4));
  75   ReplaceComponents(NULL, 0, parsed, replacements, NULL, &output, &new_parsed);
  76   ReplaceComponents("", 0, parsed, replacements, NULL, &output, &new_parsed);
  77
  78   replacements.ClearHost();
  79   ReplaceComponents(NULL, 0, parsed, replacements, NULL, &output, &new_parsed);
  80   ReplaceComponents("", 0, parsed, replacements, NULL, &output, &new_parsed);
  81   ReplaceComponents(NULL, 0, parsed, replacements, NULL, &output, &new_parsed);
  82   ReplaceComponents("", 0, parsed, replacements, NULL, &output, &new_parsed);
  83 }
  84
  85 static std::string CheckReplaceScheme(const char* base_url,
  86                                       const char* scheme) {
  87   // Make sure the input is canonicalized.
  88   RawCanonOutput<32> original;
  89   Parsed original_parsed;
  90   Canonicalize(base_url, strlen(base_url), true, NULL, &original,
  91                &original_parsed);
  92
  93   Replacements<char> replacements;
  94   replacements.SetScheme(scheme, Component(0, strlen(scheme)));
  95
  96   std::string output_string;
  97   StdStringCanonOutput output(&output_string);
  98   Parsed output_parsed;
  99   ReplaceComponents(original.data(), original.length(), original_parsed,
 100                     replacements, NULL, &output, &output_parsed);
 101
 102   output.Complete();
 103   return output_string;
 104 }
 105
 106 TEST(URLUtilTest, ReplaceScheme) {
 107   EXPECT_EQ("https://google.com/",
 108             CheckReplaceScheme("http://google.com/", "https"));
 109   EXPECT_EQ("file://google.com/",
 110             CheckReplaceScheme("http://google.com/", "file"));
 111   EXPECT_EQ("http://home/Build",
 112             CheckReplaceScheme("file:///Home/Build", "http"));
 113   EXPECT_EQ("javascript:foo",
 114             CheckReplaceScheme("about:foo", "javascript"));
 115   EXPECT_EQ("://google.com/",
 116             CheckReplaceScheme("http://google.com/", ""));
 117   EXPECT_EQ("http://google.com/",
 118             CheckReplaceScheme("about:google.com", "http"));
 119   EXPECT_EQ("http:", CheckReplaceScheme("", "http"));
 120
 121 #ifdef WIN32
 122   // Magic Windows drive letter behavior when converting to a file URL.
 123   EXPECT_EQ("file:///E:/foo/",
 124             CheckReplaceScheme("http://localhost/e:foo/", "file"));
 125 #endif
 126
 127   // This will probably change to "about://google.com/" when we fix
 128   // http://crbug.com/160 which should also be an acceptable result.
 129   EXPECT_EQ("about://google.com/",
 130             CheckReplaceScheme("http://google.com/", "about"));
 131
 132   EXPECT_EQ("http://example.com/%20hello%20# world",
 133             CheckReplaceScheme("myscheme:example.com/ hello # world ", "http"));
 134 }
 135
 136 TEST(URLUtilTest, DecodeURLEscapeSequences) {
 137   struct DecodeCase {
 138     const char* input;
 139     const char* output;
 140   } decode_cases[] = {
 141     {"hello, world", "hello, world"},
 142     {"%01%02%03%04%05%06%07%08%09%0a%0B%0C%0D%0e%0f/",
 143      "\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0B\x0C\x0D\x0e\x0f/"},
 144     {"%10%11%12%13%14%15%16%17%18%19%1a%1B%1C%1D%1e%1f/",
 145      "\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1a\x1B\x1C\x1D\x1e\x1f/"},
 146     {"%20%21%22%23%24%25%26%27%28%29%2a%2B%2C%2D%2e%2f/",
 147      " !\"#$%&'()*+,-.//"},
 148     {"%30%31%32%33%34%35%36%37%38%39%3a%3B%3C%3D%3e%3f/",
 149      "0123456789:;<=>?/"},
 150     {"%40%41%42%43%44%45%46%47%48%49%4a%4B%4C%4D%4e%4f/",
 151      "@ABCDEFGHIJKLMNO/"},
 152     {"%50%51%52%53%54%55%56%57%58%59%5a%5B%5C%5D%5e%5f/",
 153      "PQRSTUVWXYZ[\\]^_/"},
 154     {"%60%61%62%63%64%65%66%67%68%69%6a%6B%6C%6D%6e%6f/",
 155      "`abcdefghijklmno/"},
 156     {"%70%71%72%73%74%75%76%77%78%79%7a%7B%7C%7D%7e%7f/",
 157      "pqrstuvwxyz{|}~\x7f/"},
 158     // Test un-UTF-8-ization.
 159     {"%e4%bd%a0%e5%a5%bd", "\xe4\xbd\xa0\xe5\xa5\xbd"},
 160   };
 161
 162   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(decode_cases); i++) {
 163     const char* input = decode_cases[i].input;
 164     RawCanonOutputT<base::char16> output;
 165     DecodeURLEscapeSequences(input, strlen(input), &output);
 166     EXPECT_EQ(decode_cases[i].output,
 167               test_utils::ConvertUTF16ToUTF8(base::string16(output.data(),
 168                                                             output.length())));
 169   }
 170
 171   // Our decode should decode %00
 172   const char zero_input[] = "%00";
 173   RawCanonOutputT<base::char16> zero_output;
 174   DecodeURLEscapeSequences(zero_input, strlen(zero_input), &zero_output);
 175   EXPECT_NE("%00", test_utils::ConvertUTF16ToUTF8(
 176       base::string16(zero_output.data(), zero_output.length())));
 177
 178   // Test the error behavior for invalid UTF-8.
 179   const char invalid_input[] = "%e4%a0%e5%a5%bd";
 180   const base::char16 invalid_expected[4] = {0x00e4, 0x00a0, 0x597d, 0};
 181   RawCanonOutputT<base::char16> invalid_output;
 182   DecodeURLEscapeSequences(invalid_input, strlen(invalid_input),
 183                            &invalid_output);
 184   EXPECT_EQ(base::string16(invalid_expected),
 185             base::string16(invalid_output.data(), invalid_output.length()));
 186 }
 187
 188 TEST(URLUtilTest, TestEncodeURIComponent) {
 189   struct EncodeCase {
 190     const char* input;
 191     const char* output;
 192   } encode_cases[] = {
 193     {"hello, world", "hello%2C%20world"},
 194     {"\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0A\x0B\x0C\x0D\x0E\x0F",
 195      "%01%02%03%04%05%06%07%08%09%0A%0B%0C%0D%0E%0F"},
 196     {"\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C\x1D\x1E\x1F",
 197      "%10%11%12%13%14%15%16%17%18%19%1A%1B%1C%1D%1E%1F"},
 198     {" !\"#$%&'()*+,-./",
 199      "%20!%22%23%24%25%26%27()*%2B%2C-.%2F"},
 200     {"0123456789:;<=>?",
 201      "0123456789%3A%3B%3C%3D%3E%3F"},
 202     {"@ABCDEFGHIJKLMNO",
 203      "%40ABCDEFGHIJKLMNO"},
 204     {"PQRSTUVWXYZ[\\]^_",
 205      "PQRSTUVWXYZ%5B%5C%5D%5E_"},
 206     {"`abcdefghijklmno",
 207      "%60abcdefghijklmno"},
 208     {"pqrstuvwxyz{|}~\x7f",
 209      "pqrstuvwxyz%7B%7C%7D~%7F"},
 210   };
 211
 212   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(encode_cases); i++) {
 213     const char* input = encode_cases[i].input;
 214     RawCanonOutputT<char> buffer;
 215     EncodeURIComponent(input, strlen(input), &buffer);
 216     std::string output(buffer.data(), buffer.length());
 217     EXPECT_EQ(encode_cases[i].output, output);
 218   }
 219 }
 220
 221 TEST(URLUtilTest, TestResolveRelativeWithNonStandardBase) {
 222   // This tests non-standard (in the sense that GIsStandard() == false)
 223   // hierarchical schemes.
 224   struct ResolveRelativeCase {
 225     const char* base;
 226     const char* rel;
 227     bool is_valid;
 228     const char* out;
 229   } resolve_non_standard_cases[] = {
 230       // Resolving a relative path against a non-hierarchical URL should fail.
 231     {"scheme:opaque_data", "/path", false, ""},
 232       // Resolving a relative path against a non-standard authority-based base
 233       // URL doesn't alter the authority section.
 234     {"scheme://Authority/", "../path", true, "scheme://Authority/path"},
 235       // A non-standard hierarchical base is resolved with path URL
 236       // canoncialization rules.
 237     {"data:/Blah:Blah/", "file.html", true, "data:/Blah:Blah/file.html"},
 238     {"data:/Path/../part/part2", "file.html", true,
 239       "data:/Path/../part/file.html"},
 240       // Path URL canonicalization rules also apply to non-standard authority-
 241       // based URLs.
 242     {"custom://Authority/", "file.html", true,
 243       "custom://Authority/file.html"},
 244     {"custom://Authority/", "other://Auth/", true, "other://Auth/"},
 245     {"custom://Authority/", "../../file.html", true,
 246       "custom://Authority/file.html"},
 247     {"custom://Authority/path/", "file.html", true,
 248       "custom://Authority/path/file.html"},
 249     {"custom://Authority:NoCanon/path/", "file.html", true,
 250       "custom://Authority:NoCanon/path/file.html"},
 251       // It's still possible to get an invalid path URL.
 252     {"custom://Invalid:!#Auth/", "file.html", false, ""},
 253       // A path with an authority section gets canonicalized under standard URL
 254       // rules, even though the base was non-standard.
 255     {"content://content.Provider/", "//other.Provider", true,
 256       "content://other.provider/"},
 257       // Resolving an absolute URL doesn't cause canonicalization of the
 258       // result.
 259     {"about:blank", "custom://Authority", true, "custom://Authority"},
 260       // Fragment URLs can be resolved against a non-standard base.
 261     {"scheme://Authority/path", "#fragment", true,
 262       "scheme://Authority/path#fragment"},
 263     {"scheme://Authority/", "#fragment", true, "scheme://Authority/#fragment"},
 264       // Resolving should fail if the base URL is authority-based but is
 265       // missing a path component (the '/' at the end).
 266     {"scheme://Authority", "path", false, ""},
 267       // Test resolving a fragment (only) against any kind of base-URL.
 268     {"about:blank", "#id42", true, "about:blank#id42" },
 269     {"about:blank", " #id42", true, "about:blank#id42" },
 270     {"about:blank#oldfrag", "#newfrag", true, "about:blank#newfrag" },
 271       // A surprising side effect of allowing fragments to resolve against
 272       // any URL scheme is we might break javascript: URLs by doing so...
 273     {"javascript:alert('foo#bar')", "#badfrag", true,
 274       "javascript:alert('foo#badfrag" },
 275   };
 276
 277   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(resolve_non_standard_cases); i++) {
 278     const ResolveRelativeCase& test_data = resolve_non_standard_cases[i];
 279     Parsed base_parsed;
 280     ParsePathURL(test_data.base, strlen(test_data.base), false, &base_parsed);
 281
 282     std::string resolved;
 283     StdStringCanonOutput output(&resolved);
 284     Parsed resolved_parsed;
 285     bool valid = ResolveRelative(test_data.base, strlen(test_data.base),
 286                                  base_parsed, test_data.rel,
 287                                  strlen(test_data.rel), NULL, &output,
 288                                  &resolved_parsed);
 289     output.Complete();
 290
 291     EXPECT_EQ(test_data.is_valid, valid) << i;
 292     if (test_data.is_valid && valid)
 293       EXPECT_EQ(test_data.out, resolved) << i;
 294   }
 295 }
 296
 297 }  // namespace url