Index: chrome/browser/page_load_metrics/observers/ads_page_load_metrics_observer.cc |
diff --git a/chrome/browser/page_load_metrics/observers/ads_page_load_metrics_observer.cc b/chrome/browser/page_load_metrics/observers/ads_page_load_metrics_observer.cc |
new file mode 100644 |
index 0000000000000000000000000000000000000000..959b0ad9cbc9e82c9f80046a8b0376d51d051e7b |
--- /dev/null |
+++ b/chrome/browser/page_load_metrics/observers/ads_page_load_metrics_observer.cc |
@@ -0,0 +1,223 @@ |
+// Copyright 2017 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 "chrome/browser/page_load_metrics/observers/ads_page_load_metrics_observer.h" |
+ |
+#include <string> |
+ |
+#include "base/feature_list.h" |
+#include "base/strings/string_util.h" |
+#include "chrome/browser/page_load_metrics/page_load_metrics_util.h" |
+#include "content/public/browser/navigation_handle.h" |
+#include "content/public/browser/render_frame_host.h" |
+#include "content/public/browser/web_contents.h" |
+#include "url/gurl.h" |
+ |
+namespace { |
+ |
+const base::Feature kAdsFeature{"AdsMetrics", base::FEATURE_ENABLED_BY_DEFAULT}; |
+ |
+bool FrameIsAd(content::NavigationHandle* navigation_handle) { |
+ int frame_tree_node_id = navigation_handle->GetFrameTreeNodeId(); |
+ content::RenderFrameHost* current_frame_host = |
+ navigation_handle->GetWebContents()->FindFrameByFrameTreeNodeId( |
+ frame_tree_node_id); |
+ const std::string& name = current_frame_host->GetFrameName(); |
+ const GURL& url = navigation_handle->GetURL(); |
+ |
+ // Because sub-resource filtering isn't always enabled, and doesn't work |
+ // well in monitoring mode (no CSS enforcement), it's difficult to identify |
+ // ads. Google ads are prevalent and easy to track, so we'll start by |
+ // tracking those. Note that the frame name can be very large, so be careful |
+ // to avoid full string searches if possible. |
+ // TODO(jkarlin): Track other ad networks that are easy to identify. |
+ return base::StartsWith(name, "google_ads_iframe", |
+ base::CompareCase::SENSITIVE) || |
+ base::StartsWith(name, "google_ads_frame", |
+ base::CompareCase::SENSITIVE) || |
+ (url.host_piece() == "tpc.googlesyndication.com" && |
+ base::StartsWith(url.path_piece(), "/safeframe", |
+ base::CompareCase::SENSITIVE)); |
+} |
+ |
+} // namespace |
+ |
+AdsPageLoadMetricsObserver::AdsPageLoadMetricsObserver() = default; |
+AdsPageLoadMetricsObserver::~AdsPageLoadMetricsObserver() = default; |
+ |
+page_load_metrics::PageLoadMetricsObserver::ObservePolicy |
+AdsPageLoadMetricsObserver::OnCommit( |
+ content::NavigationHandle* navigation_handle) { |
+ DCHECK(ad_frames_data_.empty()); |
+ |
+ if (!base::FeatureList::IsEnabled(kAdsFeature)) |
+ return STOP_OBSERVING; |
Charlie Harrison
2017/04/11 17:30:06
Can you do this feature detection at construction
jkarlin
2017/04/13 17:25:30
Done.
|
+ |
+ // The main frame is never considered an ad. |
+ ad_frames_data_[navigation_handle->GetFrameTreeNodeId()] = nullptr; |
+ ProcessDelayedResources(navigation_handle->GetFrameTreeNodeId()); |
+ return CONTINUE_OBSERVING; |
+} |
+ |
+page_load_metrics::PageLoadMetricsObserver::ObservePolicy |
+AdsPageLoadMetricsObserver::OnCommitSubFrame( |
+ content::NavigationHandle* navigation_handle) { |
+ DCHECK(!base::ContainsKey(ad_frames_data_, |
+ navigation_handle->GetFrameTreeNodeId())); |
+ |
+ FrameTreeNodeId frame_tree_node_id = navigation_handle->GetFrameTreeNodeId(); |
+ AdFrameData* ancestor_data = FindAdAncestor(navigation_handle); |
+ |
+ if (ancestor_data) { |
+ // This frame is nested within a known ad frame. |
+ ad_frames_data_[frame_tree_node_id] = ancestor_data; |
+ } else if (FrameIsAd(navigation_handle)) { |
+ // This is the first ad frame in the frame's ancestry. |
+ ad_frames_data_storage_.push_back(AdFrameData()); |
+ ad_frames_data_[frame_tree_node_id] = &ad_frames_data_storage_.back(); |
+ } else { |
+ // No ads in the frame's ancestry, including itself. |
+ ad_frames_data_[frame_tree_node_id] = nullptr; |
+ } |
+ |
+ ProcessDelayedResources(frame_tree_node_id); |
+ return CONTINUE_OBSERVING; |
+} |
+ |
+page_load_metrics::PageLoadMetricsObserver::ObservePolicy |
+AdsPageLoadMetricsObserver::FlushMetricsOnAppEnterBackground( |
+ const page_load_metrics::PageLoadTiming& timing, |
+ const page_load_metrics::PageLoadExtraInfo& extra_info) { |
+ // The browser may come back, but there is no guarantee. To be safe, record |
+ // what we have now and ignore future changes to this navigation. |
+ if (extra_info.did_commit) |
+ RecordHistograms(); |
+ |
+ return STOP_OBSERVING; |
+} |
+ |
+void AdsPageLoadMetricsObserver::OnLoadedResource( |
+ const page_load_metrics::ExtraRequestInfo& extra_request_info) { |
+ const auto& id_and_data = |
+ ad_frames_data_.find(extra_request_info.frame_tree_node_id); |
+ if (id_and_data == ad_frames_data_.end()) { |
+ // This resouce is for a frame that hasn't yet committed. It must be the |
+ // main document for the frame. Hold onto it and once it commits we'll run |
+ // it in ProcessDelayedResources. |
+ auto it_and_success = delayed_resources_.insert(std::make_pair( |
+ extra_request_info.frame_tree_node_id, extra_request_info)); |
+ DCHECK(it_and_success.second); |
+ return; |
+ } |
+ |
+ page_bytes_ += extra_request_info.raw_body_bytes; |
+ if (!extra_request_info.was_cached) |
+ uncached_page_bytes_ += extra_request_info.raw_body_bytes; |
+ |
+ // Determine if the frame (or its ancestor) is an ad, if so attribute the |
+ // bytes to the highest ad ancestor. |
+ AdFrameData* ancestor_data = id_and_data->second; |
+ |
+ if (ancestor_data) { |
+ ancestor_data->frame_bytes += extra_request_info.raw_body_bytes; |
+ if (!extra_request_info.was_cached) { |
+ ancestor_data->frame_bytes_uncached += extra_request_info.raw_body_bytes; |
+ } |
+ } |
+} |
+ |
+void AdsPageLoadMetricsObserver::OnComplete( |
+ const page_load_metrics::PageLoadTiming& timing, |
+ const page_load_metrics::PageLoadExtraInfo& info) { |
+ RecordHistograms(); |
+} |
+ |
+void AdsPageLoadMetricsObserver::RecordHistograms() { |
+ if (page_bytes_ == 0) |
+ return; |
+ |
+ size_t total_ad_frame_bytes = 0; |
+ size_t uncached_ad_frame_bytes = 0; |
+ int ad_frames = 0; |
+ |
+ for (const AdFrameData& ad_frame_data : ad_frames_data_storage_) { |
+ total_ad_frame_bytes += ad_frame_data.frame_bytes; |
+ uncached_ad_frame_bytes += ad_frame_data.frame_bytes_uncached; |
+ |
+ if (ad_frame_data.frame_bytes > 0) { |
+ ad_frames += 1; |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.AdFrame", |
+ ad_frame_data.frame_bytes); |
+ PAGE_BYTES_HISTOGRAM( |
+ "PageLoad.Clients.Ads.Google.Bytes.AdFrameFromNetwork", |
+ ad_frame_data.frame_bytes_uncached); |
+ UMA_HISTOGRAM_PERCENTAGE( |
+ "PageLoad.Clients.Ads.Google.BytesPercent.AdFrameFromNetwork", |
+ ad_frame_data.frame_bytes_uncached * 100 / ad_frame_data.frame_bytes); |
+ } |
+ } |
+ |
+ // Don't post UMA for pages that don't have ads or content. |
+ if (total_ad_frame_bytes == 0) { |
+ UMA_HISTOGRAM_COUNTS("PageLoad.Clients.Ads.Google.PageHasNoAds", 1); |
+ return; |
+ } |
+ |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.AdFrames", |
+ total_ad_frame_bytes); |
+ |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.PageSansAdFrames", |
+ page_bytes_ - total_ad_frame_bytes); |
+ |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.Page", page_bytes_); |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.PageFromNetwork", |
+ uncached_page_bytes_); |
+ |
+ UMA_HISTOGRAM_PERCENTAGE("PageLoad.Clients.Ads.Google.BytesPercent.AdFrames", |
+ total_ad_frame_bytes * 100 / page_bytes_); |
+ |
+ UMA_HISTOGRAM_COUNTS_1000("PageLoad.Clients.Ads.Google.AdFrameCount", |
+ ad_frames); |
+ |
+ PAGE_BYTES_HISTOGRAM("PageLoad.Clients.Ads.Google.Bytes.AdFramesFromNetwork", |
+ uncached_ad_frame_bytes); |
+ |
+ UMA_HISTOGRAM_PERCENTAGE( |
+ "PageLoad.Clients.Ads.Google.BytesPercent." |
+ "AdFramesFromNetworkOfAdFramesTotal", |
+ uncached_ad_frame_bytes * 100 / total_ad_frame_bytes); |
+ |
+ int percent_bytes_from_uncached_ads = |
+ uncached_page_bytes_ == 0 |
+ ? 0 |
+ : uncached_ad_frame_bytes * 100 / uncached_page_bytes_; |
+ UMA_HISTOGRAM_PERCENTAGE( |
+ "PageLoad.Clients.Ads.Google.Bytes.Percent." |
+ "AdFramesFromNetworkOfPageFromNetwork", |
+ percent_bytes_from_uncached_ads); |
+} |
+ |
+AdsPageLoadMetricsObserver::AdFrameData* |
+AdsPageLoadMetricsObserver::FindAdAncestor( |
+ content::NavigationHandle* navigation_handle) { |
+ // We haven't seen a load from this frame before. We should have seen its |
+ // parent though. Use the ad ancestor of its parent. |
+ content::RenderFrameHost* parent_frame_host = |
+ navigation_handle->GetRenderFrameHost()->GetParent(); |
+ DCHECK(parent_frame_host); // Since this isn't a main frame. |
+ const auto& frame_id_and_data = |
+ ad_frames_data_.find(parent_frame_host->GetFrameTreeNodeId()); |
+ DCHECK(frame_id_and_data != ad_frames_data_.end()); |
+ return frame_id_and_data->second; |
+} |
+ |
+void AdsPageLoadMetricsObserver::ProcessDelayedResources( |
+ FrameTreeNodeId frame_tree_node_id) { |
+ const auto& frame_id_and_request = |
+ delayed_resources_.find(frame_tree_node_id); |
+ if (frame_id_and_request == delayed_resources_.end()) |
+ return; |
+ OnLoadedResource(frame_id_and_request->second); |
+ delayed_resources_.erase(frame_id_and_request); |
+} |