Transmit a X-Chrome-UMA-Enabled bit to Google domains from clients that have UMA enabled.

BUG=123609
TEST=With UMA enabled (not Chromium), visit www.google.com and ensure that the request header includes X-Chrome-UMA-Enabled with value "1". Also, ensure that unit_tests GoogleUtilTests all pass.

Review URL: http://codereview.chromium.org/10054029

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@132479 0039d316-1c4b-4281-b951-d872f2087c98

8 files changed, 106 insertions, 17 deletions

diff --git a/chrome/browser/google/google_util.cc b/chrome/browser/google/google_util.cc
index d95b116..22b17fc 100644
--- a/chrome/browser/google/google_util.cc
+++ b/chrome/browser/google/google_util.cc
@@ -44,13 +44,6 @@ bool HasQueryParameter(const std::string& str) {
   return false;
 }
 
-// True if |url| is an HTTP[S] request with host "[www.]google.<TLD>" and no
-// explicit port.
-bool IsGoogleDomainUrl(const GURL& url) {
-  return url.is_valid() && (url.SchemeIs("http") || url.SchemeIs("https")) &&
-      url.port().empty() && google_util::IsGoogleHostname(url.host());
-}
-
 }  // anonymous namespace
 
 namespace google_util {
@@ -142,21 +135,32 @@ bool GetReactivationBrand(std::string* brand) {
 
 #endif
 
-bool IsGoogleHostname(const std::string& host) {
+bool IsGoogleDomainUrl(const std::string& url, SubdomainPermission permission) {
+  GURL original_url(url);
+  return original_url.is_valid() && original_url.port().empty() &&
+      (original_url.SchemeIs("http") || original_url.SchemeIs("https")) &&
+      google_util::IsGoogleHostname(original_url.host(), permission);
+}
+
+bool IsGoogleHostname(const std::string& host,
+                      SubdomainPermission permission) {
   size_t tld_length =
       net::RegistryControlledDomainService::GetRegistryLength(host, false);
   if ((tld_length == 0) || (tld_length == std::string::npos))
     return false;
   std::string host_minus_tld(host, 0, host.length() - tld_length);
-  return LowerCaseEqualsASCII(host_minus_tld, "www.google.") ||
-      LowerCaseEqualsASCII(host_minus_tld, "google.");
+  if (LowerCaseEqualsASCII(host_minus_tld, "google."))
+    return true;
+  if (permission == ALLOW_SUBDOMAIN)
+    return EndsWith(host_minus_tld, ".google.", false);
+  return LowerCaseEqualsASCII(host_minus_tld, "www.google.");
 }
 
 bool IsGoogleHomePageUrl(const std::string& url) {
   GURL original_url(url);
 
   // First check to see if this has a Google domain.
-  if (!IsGoogleDomainUrl(original_url))
+  if (!IsGoogleDomainUrl(url, DISALLOW_SUBDOMAIN))
     return false;
 
   // Make sure the path is a known home page path.
@@ -173,7 +177,7 @@ bool IsGoogleSearchUrl(const std::string& url) {
   GURL original_url(url);
 
   // First check to see if this has a Google domain.
-  if (!IsGoogleDomainUrl(original_url))
+  if (!IsGoogleDomainUrl(url, DISALLOW_SUBDOMAIN))
     return false;
 
   // Make sure the path is a known search path.
diff --git a/chrome/browser/google/google_util.h b/chrome/browser/google/google_util.h
index 95ae4b9..d44afdc 100644
--- a/chrome/browser/google/google_util.h
+++ b/chrome/browser/google/google_util.h
@@ -46,8 +46,20 @@ bool GetReactivationBrand(std::string* brand);
 // need to restrict some behavior to only happen on Google's officially-owned
 // domains, use TransportSecurityState::IsGooglePinnedProperty() instead.
 
-// True if |host| is "[www.]google.<TLD>" with a valid TLD.
-bool IsGoogleHostname(const std::string& host);
+// Designate whether or not a URL checking function also checks for specific
+// subdomains, or only "www" and empty subdomains.
+enum SubdomainPermission {
+  ALLOW_SUBDOMAIN,
+  DISALLOW_SUBDOMAIN,
+};
+
+// True if |url| is an HTTP[S] request with host "[www.]google.<TLD>" and no
+// explicit port. If |allow_subdomain| is true, we check against host
+// "*.google.<TLD>" instead.
+bool IsGoogleDomainUrl(const std::string& url, SubdomainPermission permission);
+// True if |host| is "[www.]google.<TLD>" with a valid TLD. If
+// |allow_subdomain| is true, we check against host "*.google.<TLD>" instead.
+bool IsGoogleHostname(const std::string& host, SubdomainPermission permission);
 // True if |url| represents a valid Google home page URL.
 bool IsGoogleHomePageUrl(const std::string& url);
 // True if |url| represents a valid Google search URL.
diff --git a/chrome/browser/google/google_util_unittest.cc b/chrome/browser/google/google_util_unittest.cc
index f51460c..2c6e746 100644
--- a/chrome/browser/google/google_util_unittest.cc
+++ b/chrome/browser/google/google_util_unittest.cc
@@ -6,6 +6,7 @@
 #include "chrome/browser/google/google_util.h"
 #include "testing/gtest/include/gtest/gtest.h"
 
+using google_util::IsGoogleDomainUrl;
 using google_util::IsGoogleHomePageUrl;
 using google_util::IsGoogleSearchUrl;
 
@@ -241,3 +242,37 @@ TEST(GoogleUtilTest, BadSearches) {
   EXPECT_FALSE(IsGoogleSearchUrl(
       "http://www.google.com/WEBHP#q=something"));
 }
+
+TEST(GoogleUtilTest, GoogleDomains) {
+  // Test some good Google domains (valid TLDs).
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.com",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://google.com",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.ca",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.biz.tj",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.com/search?q=something",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.com/webhp",
+                                google_util::ALLOW_SUBDOMAIN));
+
+  // Test some bad Google domains (invalid TLDs).
+  EXPECT_FALSE(IsGoogleDomainUrl("http://www.google.notrealtld",
+                                 google_util::ALLOW_SUBDOMAIN));
+  EXPECT_FALSE(IsGoogleDomainUrl("http://www.google.faketld/search?q=something",
+                                 google_util::ALLOW_SUBDOMAIN));
+  EXPECT_FALSE(IsGoogleDomainUrl("http://www.yahoo.com",
+                                 google_util::ALLOW_SUBDOMAIN));
+
+  // Test subdomain checks.
+  EXPECT_TRUE(IsGoogleDomainUrl("http://images.google.com",
+                                google_util::ALLOW_SUBDOMAIN));
+  EXPECT_FALSE(IsGoogleDomainUrl("http://images.google.com",
+                                 google_util::DISALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://google.com",
+                                google_util::DISALLOW_SUBDOMAIN));
+  EXPECT_TRUE(IsGoogleDomainUrl("http://www.google.com",
+                                google_util::DISALLOW_SUBDOMAIN));
+}
diff --git a/chrome/browser/metrics/metrics_service.h b/chrome/browser/metrics/metrics_service.h
index ff4df36..510b697 100644
--- a/chrome/browser/metrics/metrics_service.h
+++ b/chrome/browser/metrics/metrics_service.h
@@ -32,6 +32,7 @@
 
 class BookmarkModel;
 class BookmarkNode;
+class ChromeNetworkDelegate;
 class MetricsReportingScheduler;
 class PrefService;
 class Profile;
@@ -437,6 +438,7 @@ class MetricsServiceHelper {
   friend class InstantFieldTrial;
   friend bool prerender::IsOmniboxEnabled(Profile* profile);
   friend class extensions::ExtensionDownloader;
+  friend class ChromeNetworkDelegate;
 
   // Returns true if prefs::kMetricsReportingEnabled is set.
   static bool IsMetricsReportingEnabled();
diff --git a/chrome/browser/net/chrome_network_delegate.cc b/chrome/browser/net/chrome_network_delegate.cc
index 545d127..e107661 100644
--- a/chrome/browser/net/chrome_network_delegate.cc
+++ b/chrome/browser/net/chrome_network_delegate.cc
@@ -14,6 +14,8 @@
 #include "chrome/browser/extensions/extension_event_router_forwarder.h"
 #include "chrome/browser/extensions/extension_info_map.h"
 #include "chrome/browser/extensions/extension_process_manager.h"
+#include "chrome/browser/google/google_util.h"
+#include "chrome/browser/metrics/metrics_service.h"
 #include "chrome/browser/prefs/pref_member.h"
 #include "chrome/browser/profiles/profile_manager.h"
 #include "chrome/browser/task_manager/task_manager.h"
@@ -165,6 +167,10 @@ int ChromeNetworkDelegate::OnBeforeSendHeaders(
     net::URLRequest* request,
     const net::CompletionCallback& callback,
     net::HttpRequestHeaders* headers) {
+  // Attach any applicable Chrome metrics headers. This is a best-effort attempt
+  // where failure will not cause OnBeforeSendHeaders to back out.
+  AddChromeMetricsStateHeader(request, headers);
+
   return ExtensionWebRequestEventRouter::GetInstance()->OnBeforeSendHeaders(
       profile_, extension_info_map_.get(), request, callback, headers);
 }
@@ -304,3 +310,23 @@ bool ChromeNetworkDelegate::CanSetCookie(
 
   return allow;
 }
+
+void ChromeNetworkDelegate::AddChromeMetricsStateHeader(
+    net::URLRequest* request,
+    net::HttpRequestHeaders* headers) {
+  // Note our criteria for attaching Chrome experiment headers:
+  // 1. We only transmit to *.google.<TLD> domains. NOTE that this use of
+  //    google_util helpers to check this does not guarantee that the URL is
+  //    Google-owned, only that it is of the form *.google.<TLD>. In the future
+  //    we may choose to reinforce this check.
+  // 2. We must verify that the transmitting profile is not off the record.
+  // 3. For the X-Chrome-UMA-Enabled bit, we only set it if UMA is in fact
+  //    enabled for this install of Chrome.
+  Profile* profile_instance = reinterpret_cast<Profile*>(profile_);
+  if (google_util::IsGoogleDomainUrl(request->url().spec(),
+                                     google_util::ALLOW_SUBDOMAIN) &&
+      profile_instance && !profile_instance->IsOffTheRecord() &&
+      MetricsServiceHelper::IsMetricsReportingEnabled()) {
+    headers->SetHeader("X-Chrome-UMA-Enabled", "1");
+  }
+}
diff --git a/chrome/browser/net/chrome_network_delegate.h b/chrome/browser/net/chrome_network_delegate.h
index a582c38..121256a 100644
--- a/chrome/browser/net/chrome_network_delegate.h
+++ b/chrome/browser/net/chrome_network_delegate.h
@@ -83,6 +83,11 @@ class ChromeNetworkDelegate : public net::NetworkDelegate {
                             const std::string& cookie_line,
                             net::CookieOptions* options) OVERRIDE;
 
+  // Adds Chrome experiment and metrics state as custom headers to |headers|
+  // based on local state and |request|.
+  void AddChromeMetricsStateHeader(net::URLRequest* request,
+                                   net::HttpRequestHeaders* headers);
+
   scoped_refptr<ExtensionEventRouterForwarder> event_router_;
   void* profile_;
   scoped_refptr<CookieSettings> cookie_settings_;
diff --git a/chrome/browser/protector/protector_service.cc b/chrome/browser/protector/protector_service.cc
index edc24cc..27f9485 100644
--- a/chrome/browser/protector/protector_service.cc
+++ b/chrome/browser/protector/protector_service.cc
@@ -26,8 +26,11 @@ namespace {
 bool CanMerge(const GURL& url1, const GURL& url2) {
   VLOG(1) << "Checking if can merge " << url1.spec() << " with " << url2.spec();
   // All Google URLs are considered the same one.
-  if (google_util::IsGoogleHostname(url1.host()))
-    return google_util::IsGoogleHostname(url2.host());
+  if (google_util::IsGoogleHostname(url1.host(),
+                                    google_util::DISALLOW_SUBDOMAIN)) {
+    return google_util::IsGoogleHostname(url2.host(),
+                                         google_util::DISALLOW_SUBDOMAIN);
+  }
   // Otherwise URLs must have the same domain.
   return net::RegistryControlledDomainService::SameDomainOrHost(url1, url2);
 }
diff --git a/chrome/browser/search_engines/template_url_prepopulate_data.cc b/chrome/browser/search_engines/template_url_prepopulate_data.cc
index 5f785ce..c8cdcc9 100644
--- a/chrome/browser/search_engines/template_url_prepopulate_data.cc
+++ b/chrome/browser/search_engines/template_url_prepopulate_data.cc
@@ -3281,8 +3281,10 @@ static const PrepopulatedEngine* GetEngineForURL(const std::string& url) {
   // First special-case Google, because the prepopulate URL for it will not
   // convert to a GURL and thus won't have an origin.  Instead see if the
   // incoming URL's host is "[*.]google.<TLD>".
-  if (google_util::IsGoogleHostname(as_gurl.host()))
+  if (google_util::IsGoogleHostname(as_gurl.host(),
+                                    google_util::DISALLOW_SUBDOMAIN)) {
     return &google;
+  }
 
   // Now check the rest of the prepopulate data.
   GURL origin(as_gurl.GetOrigin());