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

components/url_formatter/url_formatter.cc

Index: components/url_formatter/url_formatter.cc
diff --git a/components/url_formatter/url_formatter.cc b/components/url_formatter/url_formatter.cc
index 298b1c18ff42bfae9102433a16200dd4a65d642b..4413eb0392c59733a0d13d42443ffedb0e97c277 100644
--- a/components/url_formatter/url_formatter.cc
+++ b/components/url_formatter/url_formatter.cc
@@ -17,6 +17,7 @@
 #include "base/strings/utf_string_conversions.h"
 #include "base/threading/thread_local_storage.h"
 #include "components/url_formatter/idn_spoof_checker.h"
+#include "net/base/registry_controlled_domains/registry_controlled_domain.h"
 #include "third_party/icu/source/common/unicode/uidna.h"
 #include "third_party/icu/source/common/unicode/utypes.h"
 #include "url/gurl.h"
@@ -578,9 +579,30 @@ base::string16 StripWWW(const base::string16& text) {
       ? text.substr(www.length()) : text;
 }
 
-base::string16 StripWWWFromHost(const GURL& url) {
+base::StringPiece StripSubdomains(const GURL& url,
+                                  StripSubdomainTypes subdomains) {
   DCHECK(url.is_valid());
-  return StripWWW(base::ASCIIToUTF16(url.host_piece()));
+
+  base::StringPiece host_piece = url.host_piece();
+  std::string domain_and_registry =
+      net::registry_controlled_domains::GetDomainAndRegistry(
+          host_piece,
+          net::registry_controlled_domains::INCLUDE_PRIVATE_REGISTRIES);
+
+  if (host_piece == domain_and_registry)
+    return host_piece;
+
+  std::string subdomain_piece =
+      host_piece.substr(0, host_piece.length() - domain_and_registry.length())
+          .as_string();
+
+  if ((subdomains & kStripWWW) != 0)

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

Nit: Personally I find omitting the "!= 0" more readable, since these bitmasks
function more like bools than ints. (2 places)

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

Done. Cool. Me too. I was just following precedent elsewhere.

+    base::ReplaceSubstringsAfterOffset(&subdomain_piece, 0, "www.", "");
+
+  if ((subdomains & kStripM) != 0)
+    base::ReplaceSubstringsAfterOffset(&subdomain_piece, 0, "m.", "");
+
+  return subdomain_piece + domain_and_registry;
 }
 
 }  // namespace url_formatter