src/chrome/browser/history/top_sites_cache_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 "chrome/browser/history/top_sites_cache.h"
   6
   7 #include <set>
   8
   9 #include "base/basictypes.h"
  10 #include "base/logging.h"
  11 #include "base/strings/string16.h"
  12 #include "base/strings/string_number_conversions.h"
  13 #include "base/strings/utf_string_conversions.h"
  14 #include "testing/gtest/include/gtest/gtest.h"
  15
  16 namespace history {
  17
  18 namespace {
  19
  20 class TopSitesCacheTest : public testing::Test {
  21  public:
  22   TopSitesCacheTest() {
  23   }
  24
  25  protected:
  26   // Initializes |top_sites_| and |cache_| based on |spec|, which is a list of
  27   // URL strings with optional indents: indentated URLs redirect to the last
  28   // non-indented URL. Titles are assigned as "Title 1", "Title 2", etc., in the
  29   // order of appearance. See |kTopSitesSpecBasic| for an example.
  30   void InitTopSiteCache(const char** spec, size_t size);
  31
  32   MostVisitedURLList top_sites_;
  33   TopSitesCache cache_;
  34
  35  private:
  36   DISALLOW_COPY_AND_ASSIGN(TopSitesCacheTest);
  37 };
  38
  39 void TopSitesCacheTest::InitTopSiteCache(const char** spec, size_t size) {
  40   std::set<std::string> urls_seen;
  41   for (size_t i = 0; i < size; ++i) {
  42     const char* spec_item = spec[i];
  43     while (*spec_item && *spec_item == ' ')  // Eat indent.
  44       ++spec_item;
  45     if (urls_seen.find(spec_item) != urls_seen.end())
  46       NOTREACHED() << "Duplicate URL found: " << spec_item;
  47     urls_seen.insert(spec_item);
  48     if (spec_item == spec[i]) {  // No indent: add new MostVisitedURL.
  49       string16 title(ASCIIToUTF16("Title ") +
  50                      base::Uint64ToString16(top_sites_.size() + 1));
  51       top_sites_.push_back(MostVisitedURL(GURL(spec_item), title));
  52     }
  53     ASSERT_TRUE(!top_sites_.empty());
  54     // Set up redirect to canonical URL. Canonical URL redirects to itself, too.
  55     top_sites_.back().redirects.push_back(GURL(spec_item));
  56   }
  57   cache_.SetTopSites(top_sites_);
  58 }
  59
  60 const char* kTopSitesSpecBasic[] = {
  61   "http://www.google.com",
  62   "  http://www.gogle.com",  // Redirects.
  63   "  http://www.gooogle.com",  // Redirects.
  64   "http://www.youtube.com/a/b",
  65   "  http://www.youtube.com/a/b?test=1",  // Redirects.
  66   "https://www.google.com/",
  67   "  https://www.gogle.com",  // Redirects.
  68   "http://www.example.com:3141/",
  69 };
  70
  71 TEST_F(TopSitesCacheTest, GetCanonicalURL) {
  72   InitTopSiteCache(kTopSitesSpecBasic, arraysize(kTopSitesSpecBasic));
  73   struct {
  74     const char* expected;
  75     const char* query;
  76   } test_cases[] = {
  77     // Already is canonical: redirects.
  78     {"http://www.google.com/", "http://www.google.com"},
  79     // Exact match with stored URL: redirects.
  80     {"http://www.google.com/", "http://www.gooogle.com"},
  81     // Recognizes despite trailing "/": redirects
  82     {"http://www.google.com/", "http://www.gooogle.com/"},
  83     // Exact match with URL with query: redirects.
  84     {"http://www.youtube.com/a/b", "http://www.youtube.com/a/b?test=1"},
  85     // No match with URL with query: as-is.
  86     {"http://www.youtube.com/a/b?test", "http://www.youtube.com/a/b?test"},
  87     // Never-seen-before URL: as-is.
  88     {"http://maps.google.com/", "http://maps.google.com/"},
  89     // Changing port number, does not match: as-is.
  90     {"http://www.example.com:1234/", "http://www.example.com:1234"},
  91     // Smart enough to know that port 80 is HTTP: redirects.
  92     {"http://www.google.com/", "http://www.gooogle.com:80"},
  93     // Prefix should not work: as-is.
  94     {"http://www.youtube.com/a", "http://www.youtube.com/a"},
  95   };
  96   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(test_cases); ++i) {
  97     std::string expected(test_cases[i].expected);
  98     std::string query(test_cases[i].query);
  99     EXPECT_EQ(expected, cache_.GetCanonicalURL(GURL(query)).spec())
 100       << " for test_case[" << i << "]";
 101   }
 102 }
 103
 104 TEST_F(TopSitesCacheTest, IsKnownUrl) {
 105   InitTopSiteCache(kTopSitesSpecBasic, arraysize(kTopSitesSpecBasic));
 106   // Matches.
 107   EXPECT_TRUE(cache_.IsKnownURL(GURL("http://www.google.com")));
 108   EXPECT_TRUE(cache_.IsKnownURL(GURL("http://www.gooogle.com")));
 109   EXPECT_TRUE(cache_.IsKnownURL(GURL("http://www.google.com/")));
 110
 111   // Non-matches.
 112   EXPECT_FALSE(cache_.IsKnownURL(GURL("http://www.google.com?")));
 113   EXPECT_FALSE(cache_.IsKnownURL(GURL("http://www.google.net")));
 114   EXPECT_FALSE(cache_.IsKnownURL(GURL("http://www.google.com/stuff")));
 115   EXPECT_FALSE(cache_.IsKnownURL(GURL("https://www.gooogle.com")));
 116   EXPECT_FALSE(cache_.IsKnownURL(GURL("http://www.youtube.com/a")));
 117 }
 118
 119 const char* kTopSitesSpecPrefix[] = {
 120   "http://www.google.com/",
 121   "  http://www.google.com/test?q=3",  // Redirects.
 122   "  http://www.google.com/test/y?d",  // Redirects.
 123   "  http://www.chromium.org/a/b",  // Redirects.
 124   "http://www.google.com/2",
 125   "  http://www.google.com/test/q",  // Redirects.
 126   "  http://www.google.com/test/y?b",  // Redirects.
 127   "http://www.google.com/3",
 128   "  http://www.google.com/testing",  // Redirects.
 129   "http://www.google.com/test-hyphen",
 130   "http://www.google.com/sh",
 131   "  http://www.google.com/sh/1/2/3",  // Redirects.
 132   "http://www.google.com/sh/1",
 133 };
 134
 135 TEST_F(TopSitesCacheTest, GetCanonicalURLExactMatch) {
 136   InitTopSiteCache(kTopSitesSpecPrefix, arraysize(kTopSitesSpecPrefix));
 137   for (size_t i = 0; i < arraysize(kTopSitesSpecPrefix); ++i) {
 138     // Go through each entry in kTopSitesSpecPrefix, trimming space.
 139     const char* s = kTopSitesSpecPrefix[i];
 140     while (*s && *s == ' ')
 141       ++s;
 142     // Get the answer from direct lookup.
 143     GURL stored_url(s);
 144     GURL expected(cache_.GetCanonicalURL(stored_url));
 145     // Test specialization.
 146     GURL result1(cache_.GetSpecializedCanonicalURL(stored_url));
 147     EXPECT_EQ(expected, result1) << " for kTopSitesSpecPrefix[" << i << "]";
 148     // Test generalization.
 149     GURL result2(cache_.GetGeneralizedCanonicalURL(stored_url));
 150     EXPECT_EQ(expected, result2) << " for kTopSitesSpecPrefix[" << i << "]";
 151   }
 152 }
 153
 154 TEST_F(TopSitesCacheTest, GetSpecializedCanonicalURL) {
 155   InitTopSiteCache(kTopSitesSpecPrefix, arraysize(kTopSitesSpecPrefix));
 156   struct {
 157     const char* expected;
 158     const char* query;
 159   } test_cases[] = {
 160     // Exact match after trimming "?query": redirects.
 161     {"http://www.google.com/", "http://www.google.com/test"},
 162     // Specialized match: redirects.
 163     {"http://www.google.com/sh", "http://www.google.com/sh/1/2"},
 164     // Specialized match with trailing "/": redirects.
 165     {"http://www.google.com/sh", "http://www.google.com/sh/1/2/"},
 166     // Unique specialization match: redirects.
 167     {"http://www.google.com/", "http://www.chromium.org/a"},
 168     // Multiple exact matches after trimming: redirects to first.
 169     {"http://www.google.com/2", "http://www.google.com/test/y"},
 170     // Multiple specialized matches: redirects to least specialized.
 171     {"http://www.google.com/2", "http://www.google.com/test/q"},
 172     // No specialized match: fails.
 173     {"", "http://www.google.com/no-match"},
 174     // String prefix match but not URL-prefix match: fails.
 175     {"", "http://www.google.com/t"},
 176     // Different protocol: fails.
 177     {"", "https://www.google.com/test"},
 178     // Smart enough to know that port 80 is HTTP: redirects.
 179     {"http://www.google.com/", "http://www.google.com:80/test"},
 180     // Generalization match only: fails.
 181     {"", "http://www.google.com/sh/1/2/3/4"},
 182   };
 183   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(test_cases); ++i) {
 184     std::string expected(test_cases[i].expected);
 185     std::string query(test_cases[i].query);
 186     GURL result(cache_.GetSpecializedCanonicalURL(GURL(query)));
 187     EXPECT_EQ(expected, result.spec()) << " for test_case[" << i << "]";
 188   }
 189 }
 190
 191 TEST_F(TopSitesCacheTest, GetGeneralizedCanonicalURL) {
 192   InitTopSiteCache(kTopSitesSpecPrefix, arraysize(kTopSitesSpecPrefix));
 193   struct {
 194     const char* expected;
 195     const char* query;
 196   } test_cases[] = {
 197     // Exact match after trimming "?query": redirects.
 198     {"http://www.google.com/", "http://www.google.com/test"},
 199     // Same, but different code path: redirects.
 200     {"http://www.google.com/", "http://www.google.com/test/y?e"},
 201     {"http://www.google.com/", "http://www.google.com/test/y?c"},
 202     // Same, but code path leads to different result: redirects.
 203     {"http://www.google.com/2", "http://www.google.com/test/y?a"},
 204     // Generalized match: redirects.
 205     {"http://www.google.com/3", "http://www.google.com/3/1/4/1/5/9"},
 206     // Generalized match with trailing "/": redirects.
 207     {"http://www.google.com/3", "http://www.google.com/3/1/4/1/5/9/"},
 208     // Unique generalization match: redirects.
 209     {"http://www.google.com/", "http://www.chromium.org/a/b/c"},
 210     // Multiple exact matches after trimming: redirects to first.
 211     {"http://www.google.com/2", "http://www.google.com/test/y"},
 212     // Multiple generalized matches: redirects to least general.
 213     {"http://www.google.com/sh", "http://www.google.com/sh/1/2/3/4/"},
 214     // Multiple generalized matches: redirects to least general.
 215     {"http://www.google.com/sh", "http://www.google.com/sh/1/2/3/4/"},
 216     // Competing generalized match: take the most specilized.
 217     {"http://www.google.com/2", "http://www.google.com/test/q"},
 218     // No generalized match, early element: fails.
 219     {"", "http://www.a.com/"},
 220     // No generalized match, intermediate element: fails.
 221     {"", "http://www.e-is-between-chromium-and-google.com/"},
 222     // No generalized match, late element: fails.
 223     {"", "http://www.zzzzzzz.com/"},
 224     // String prefix match but not URL-prefix match: fails.
 225     {"", "http://www.chromium.org/a/beeswax"},
 226     // String prefix match and URL-prefix match: redirects.
 227     {"http://www.google.com/", "http://www.google.com/shhhhhh"},
 228     // Different protocol: fails.
 229     {"", "https://www.google.com/test"},
 230     // Smart enough to know that port 80 is HTTP: redirects.
 231     {"http://www.google.com/", "http://www.google.com:80/test"},
 232     // Specialized match only: fails.
 233     {"", "http://www.chromium.org/a"},
 234   };
 235   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(test_cases); ++i) {
 236     std::string expected(test_cases[i].expected);
 237     std::string query(test_cases[i].query);
 238     GURL result(cache_.GetGeneralizedCanonicalURL(GURL(query)));
 239     EXPECT_EQ(expected, result.spec()) << " for test_case[" << i << "]";
 240   }
 241 }
 242
 243 // This tests a special case where there are 2 specialized and generalized
 244 // matches, and both should be checked to find the correct match.
 245 TEST_F(TopSitesCacheTest, GetPrefixCanonicalURLDiffByQuery) {
 246   const char* top_sites_spec[] = {
 247     "http://www.dest.com/1",
 248     "  http://www.source.com/a?m=5",  // Redirects.
 249     "http://www.dest.com/2",
 250     "  http://www.source.com/a/t?q=3",  // Redirects.
 251   };
 252   InitTopSiteCache(top_sites_spec, arraysize(top_sites_spec));
 253
 254   // Shared by GetSpecializedCanonicalURL() and GetGeneralizedCanonicalURL
 255   struct {
 256     const char* expected;
 257     const char* query;
 258   } test_cases[] = {
 259     // Slightly before "http://www.source.com/a?m=5".
 260     {"http://www.dest.com/1", "http://www.source.com/a?l=5"},
 261     // Slightly after "http://www.source.com/a?m=5".
 262     {"http://www.dest.com/1", "http://www.source.com/a?n=5"},
 263     // Slightly before "http://www.source.com/a/t?q=3".
 264     {"http://www.dest.com/2", "http://www.source.com/a/t?q=2"},
 265     // Slightly after "http://www.source.com/a/t?q=3".
 266     {"http://www.dest.com/2", "http://www.source.com/a/t?q=4"},
 267   };
 268
 269   for (size_t i = 0; i < ARRAYSIZE_UNSAFE(test_cases); ++i) {
 270     std::string expected(test_cases[i].expected);
 271     std::string query(test_cases[i].query);
 272     GURL result1(cache_.GetSpecializedCanonicalURL(GURL(query)));
 273     EXPECT_EQ(expected, result1.spec()) << " for test_case[" << i << "]";
 274     GURL result2(cache_.GetGeneralizedCanonicalURL(GURL(query)));
 275     EXPECT_EQ(expected, result2.spec()) << " for test_case[" << i << "]";
 276   }
 277 }
 278
 279 }  // namespace
 280
 281 }  // namespace history