URL Formatter: Add StripSubdomain method that preserves eTLD + 1.

components/url_formatter/url_formatter_unittest.cc

 // Copyright 2015 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 "components/url_formatter/url_formatter.h"
 
 #include <stddef.h>
 #include <string.h>
 
 #include <vector>
@@ skipping 1205 matching lines @@
                        omit_http_start_with_ftp_offsets);
 
   const size_t omit_all_offsets[] = {
     0, kNpos, kNpos, kNpos, kNpos, kNpos, kNpos, 0, kNpos, kNpos, kNpos, kNpos,
     0, 1, 2, 3, 4, 5, 6, 7
   };
   CheckAdjustedOffsets("http://user@foo.com/", kFormatUrlOmitAll,
                        net::UnescapeRule::NORMAL, omit_all_offsets);
 }
 
+TEST(UrlFormatterTest, StripSubdomains) {
+  struct TestCase {
+    std::string expected;
+    std::string input_host;
+    StripSubdomainTypes subdomains_to_strip;
+  } cases[] = {
+      // Test successful strippings of subdomains.
+      {"google.com", "www.google.com", kStripWWW},
+      {"google.com", "m.google.com", kStripM},
+      {"google.com", "www.google.com", kStripWWW | kStripM},
+      {"google.com", "m.google.com", kStripWWW | kStripM},
+      {"google.com", "www.m.google.com", kStripWWW | kStripM},
+      {"google.com", "m.www.google.com", kStripWWW | kStripM},
+
+      // Test partial strippings.
+      {"m.google.com", "www.m.google.com", kStripWWW},
+      {"www.google.com", "m.www.google.com", kStripM},
+      {"en.wikipedia.org", "en.m.wikipedia.org", kStripWWW | kStripM},
+      {"foo.google.com", "www.foo.google.com", kStripWWW | kStripM},
+      {"foo.google.com", "foo.www.google.com", kStripWWW | kStripM},
+
+      // Leave subdomains alone if the flags specify it as such.
+      {"www.google.com", "www.google.com", kStripM},
+      {"m.google.com", "m.google.com", kStripWWW},
+
+      // Sanity check that we respect other registries.

[Peter Kasting, 2017/06/19 23:42:32]

Are these tests just extensions of the block below?  Maybe put them into the end
of it?

[tommycli, 2017/06/19 23:48:51]

Done.

+      {"www.co.uk", "www.co.uk", kStripWWW},
+      {"www.garden", "www.garden", kStripWWW},
+
+      // Leave eTLD+1s (registry and domain) alone.
+      {"google.com", "google.com", kStripWWW},
+      {"www.com", "www.com", kStripWWW | kStripM},
+      {"m.com", "m.com", kStripWWW | kStripM},
+      {"mail.google.com", "mail.google.com", kStripWWW | kStripM},
+  };
+
+  for (TestCase& test_case : cases) {
+    EXPECT_EQ(test_case.expected, url_formatter::StripSubdomains(
+                                      GURL("http://" + test_case.input_host),
+                                      test_case.subdomains_to_strip))
+        << "  input_host = " << test_case.input_host << std::endl
+        << "  subdomains_to_strip = " << test_case.subdomains_to_strip;
+  }
+}
+
 }  // namespace
 
 }  // namespace url_formatter