Include the HTTP status code as a feature in the ClientPhishingRequest.

BUG=none
TEST=BrowserFeatureExtractorTest


Review URL: http://codereview.chromium.org/8400050

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@107867 0039d316-1c4b-4281-b951-d872f2087c98

6 files changed, 16 insertions, 2 deletions

diff --git a/chrome/browser/safe_browsing/browser_feature_extractor.cc b/chrome/browser/safe_browsing/browser_feature_extractor.cc
index a5e2e6b..5dd4186 100644
--- a/chrome/browser/safe_browsing/browser_feature_extractor.cc
+++ b/chrome/browser/safe_browsing/browser_feature_extractor.cc
@@ -28,7 +28,7 @@
 
 namespace safe_browsing {
 
-BrowseInfo::BrowseInfo() {}
+BrowseInfo::BrowseInfo() : http_status_code(0) {}
 
 BrowseInfo::~BrowseInfo() {}
 
@@ -232,7 +232,9 @@ void BrowserFeatureExtractor::ExtractBrowseInfoFeatures(
                static_cast<double>(info.unsafe_resource->threat_type),
                request);
   }
-
+  if (info.http_status_code != 0) {
+    AddFeature(features::kHttpStatusCode, info.http_status_code, request);
+  }
 }
 
 void BrowserFeatureExtractor::StartExtractFeatures(
diff --git a/chrome/browser/safe_browsing/browser_feature_extractor.h b/chrome/browser/safe_browsing/browser_feature_extractor.h
index d227407..f7b8dab 100644
--- a/chrome/browser/safe_browsing/browser_feature_extractor.h
+++ b/chrome/browser/safe_browsing/browser_feature_extractor.h
@@ -49,6 +49,9 @@ struct BrowseInfo {
   std::vector<GURL> host_redirects;
   std::vector<GURL> url_redirects;
 
+  // The HTTP status code from this navigation.
+  int http_status_code;
+
   BrowseInfo();
   ~BrowseInfo();
 };
diff --git a/chrome/browser/safe_browsing/browser_feature_extractor_unittest.cc b/chrome/browser/safe_browsing/browser_feature_extractor_unittest.cc
index 8e0309e..28c86a4 100644
--- a/chrome/browser/safe_browsing/browser_feature_extractor_unittest.cc
+++ b/chrome/browser/safe_browsing/browser_feature_extractor_unittest.cc
@@ -272,6 +272,7 @@ TEST_F(BrowserFeatureExtractorTest, BrowseFeatures) {
   redirect_chain.push_back(GURL("http://somerandomwebsite.com/"));
   redirect_chain.push_back(GURL("http://www.foo.com/"));
   SetRedirectChain(redirect_chain, true);
+  browse_info_->http_status_code = 200;
   NavigateAndCommit(GURL("http://www.foo.com/"),
                     GURL("http://google.com/"),
                     content::PageTransitionFromInt(
@@ -299,6 +300,7 @@ TEST_F(BrowserFeatureExtractorTest, BrowseFeatures) {
   EXPECT_EQ(0.0, features[features::kHasSSLReferrer]);
   EXPECT_EQ(2.0, features[features::kPageTransitionType]);
   EXPECT_EQ(1.0, features[features::kIsFirstNavigation]);
+  EXPECT_EQ(200.0, features[features::kHttpStatusCode]);
 
   request.Clear();
   request.set_url("http://www.foo.com/page.html");
@@ -308,6 +310,7 @@ TEST_F(BrowserFeatureExtractorTest, BrowseFeatures) {
   redirect_chain.push_back(GURL("http://www.foo.com/second_redirect"));
   redirect_chain.push_back(GURL("http://www.foo.com/page.html"));
   SetRedirectChain(redirect_chain, false);
+  browse_info_->http_status_code = 404;
   NavigateAndCommit(GURL("http://www.foo.com/page.html"),
                     GURL("http://www.foo.com"),
                     content::PageTransitionFromInt(
@@ -352,6 +355,7 @@ TEST_F(BrowserFeatureExtractorTest, BrowseFeatures) {
             features[StringPrintf("%s%s",
                                   features::kHostPrefix,
                                   features::kIsFirstNavigation)]);
+  EXPECT_EQ(404.0, features[features::kHttpStatusCode]);
 
   request.Clear();
   request.set_url("http://www.bar.com/");
diff --git a/chrome/browser/safe_browsing/browser_features.cc b/chrome/browser/safe_browsing/browser_features.cc
index cff098d..e43cd5e 100644
--- a/chrome/browser/safe_browsing/browser_features.cc
+++ b/chrome/browser/safe_browsing/browser_features.cc
@@ -27,6 +27,7 @@ const char kRedirectUrlMismatch[] = "RedirectUrlMismatch";
 const char kRedirect[] = "Redirect";
 const char kSecureRedirectValue[] = "SecureRedirect";
 const char kBadIpFetch[] = "BadIpFetch=";
+const char kHttpStatusCode[] = "HttpStatusCode";
 const char kSafeBrowsingMaliciousUrl[] = "SafeBrowsingMaliciousUrl=";
 const char kSafeBrowsingOriginalUrl[] = "SafeBrowsingOriginalUrl=";
 const char kSafeBrowsingIsSubresource[] = "SafeBrowsingIsSubresource";
diff --git a/chrome/browser/safe_browsing/browser_features.h b/chrome/browser/safe_browsing/browser_features.h
index c8ad4c2..0f3f033 100644
--- a/chrome/browser/safe_browsing/browser_features.h
+++ b/chrome/browser/safe_browsing/browser_features.h
@@ -71,6 +71,9 @@ extern const char kSecureRedirectValue[];
 // Resource was fetched from a known bad IP address.
 extern const char kBadIpFetch[];
 
+// The HTTP status code for the main document.
+extern const char kHttpStatusCode[];
+
 // SafeBrowsing related featues.  Fields from the UnsafeResource if there is
 // any.
 extern const char kSafeBrowsingMaliciousUrl[];
diff --git a/chrome/browser/safe_browsing/client_side_detection_host.cc b/chrome/browser/safe_browsing/client_side_detection_host.cc
index dfcce12..06c315a7 100644
--- a/chrome/browser/safe_browsing/client_side_detection_host.cc
+++ b/chrome/browser/safe_browsing/client_side_detection_host.cc
@@ -324,6 +324,7 @@ void ClientSideDetectionHost::DidNavigateMainFramePostCommit(
   }
   browse_info_->host_redirects = cur_host_redirects_;
   browse_info_->url_redirects = params.redirects;
+  browse_info_->http_status_code = details.http_status_code;
 
   // Notify the renderer if it should classify this URL.
   classification_request_ = new ShouldClassifyUrlRequest(params,