OLD | NEW |
| (Empty) |
1 // Copyright 2017 The Chromium Authors. All rights reserved. | |
2 // Use of this source code is governed by a BSD-style license that can be | |
3 // found in the LICENSE file. | |
4 | |
5 #include "chrome/browser/page_load_metrics/ads_detection.h" | |
6 | |
7 #include <memory> | |
8 #include <string> | |
9 #include <utility> | |
10 | |
11 #include "base/logging.h" | |
12 #include "base/macros.h" | |
13 #include "base/memory/ptr_util.h" | |
14 #include "base/strings/string_util.h" | |
15 #include "base/supports_user_data.h" | |
16 #include "content/public/browser/navigation_handle.h" | |
17 #include "content/public/browser/render_frame_host.h" | |
18 #include "content/public/browser/web_contents.h" | |
19 #include "url/gurl.h" | |
20 | |
21 namespace page_load_metrics { | |
22 | |
23 namespace { | |
24 | |
25 bool IsGoogleAd(content::NavigationHandle* navigation_handle) { | |
26 // Because sub-resource filtering isn't always enabled, and doesn't work | |
27 // well in monitoring mode (no CSS enforcement), it's difficult to identify | |
28 // ads. Google ads are prevalent and easy to track, so we'll start by | |
29 // tracking those. Note that the frame name can be very large, so be careful | |
30 // to avoid full string searches if possible. | |
31 // TODO(jkarlin): Track other ad networks that are easy to identify. | |
32 | |
33 // In case the navigation aborted, look up the RFH by the Frame Tree Node | |
34 // ID. It returns the committed frame host or the initial frame host for the | |
35 // frame if no committed host exists. Using a previous host is fine because | |
36 // once a frame has an ad we always consider it to have an ad. | |
37 // We use the unsafe method of FindFrameByFrameTreeNodeId because we're not | |
38 // concerned with which process the frame lives on (we're just measuring | |
39 // bytes and not granting security privileges). | |
40 content::RenderFrameHost* current_frame_host = | |
41 navigation_handle->GetWebContents()->UnsafeFindFrameByFrameTreeNodeId( | |
42 navigation_handle->GetFrameTreeNodeId()); | |
43 if (current_frame_host) { | |
44 const std::string& frame_name = current_frame_host->GetFrameName(); | |
45 if (base::StartsWith(frame_name, "google_ads_iframe", | |
46 base::CompareCase::SENSITIVE) || | |
47 base::StartsWith(frame_name, "google_ads_frame", | |
48 base::CompareCase::SENSITIVE)) { | |
49 return true; | |
50 } | |
51 } | |
52 | |
53 const GURL& frame_url = navigation_handle->GetURL(); | |
54 if (frame_url.host_piece() == "tpc.googlesyndication.com" && | |
55 base::StartsWith(frame_url.path_piece(), "/safeframe", | |
56 base::CompareCase::SENSITIVE)) { | |
57 return true; | |
58 } | |
59 | |
60 return false; | |
61 } | |
62 | |
63 // Associates the detected AdTypes for a navigation with its NavigationHandle. | |
64 class NavigationHandleAdsData : public base::SupportsUserData::Data { | |
65 public: | |
66 static NavigationHandleAdsData* GetOrCreate( | |
67 content::NavigationHandle* navigation_handle) { | |
68 DCHECK(navigation_handle); | |
69 NavigationHandleAdsData* ads_data = static_cast<NavigationHandleAdsData*>( | |
70 navigation_handle->GetUserData(kUserDataKey)); | |
71 if (!ads_data) { | |
72 std::unique_ptr<NavigationHandleAdsData> new_ads_data = | |
73 base::MakeUnique<NavigationHandleAdsData>(); | |
74 | |
75 // It is safe to retain |ads_data| raw pointer, despite passing an | |
76 // ownership of |new_ads_data| to SetUserData, because |navigation_handle| | |
77 // will keep the NavigationHandleAdsData instance alive until the | |
78 // |navigation_handle| is destroyed. | |
79 ads_data = new_ads_data.get(); | |
80 | |
81 navigation_handle->SetUserData(kUserDataKey, std::move(new_ads_data)); | |
82 } | |
83 | |
84 return ads_data; | |
85 } | |
86 | |
87 NavigationHandleAdsData() = default; | |
88 ~NavigationHandleAdsData() override {} | |
89 | |
90 AdTypes& ad_types() { return ad_types_; } | |
91 const AdTypes& ad_types() const { return ad_types_; } | |
92 | |
93 void CalculateGoogleAdsIfNecessary( | |
94 content::NavigationHandle* navigation_handle) { | |
95 if (has_calculated_google_ads_) | |
96 return; | |
97 has_calculated_google_ads_ = true; | |
98 | |
99 if (IsGoogleAd(navigation_handle)) | |
100 ad_types_.set(AD_TYPE_GOOGLE); | |
101 } | |
102 | |
103 private: | |
104 AdTypes ad_types_; | |
105 bool has_calculated_google_ads_ = false; | |
106 | |
107 static const char kUserDataKey[]; | |
108 | |
109 DISALLOW_COPY_AND_ASSIGN(NavigationHandleAdsData); | |
110 }; | |
111 | |
112 const char NavigationHandleAdsData::kUserDataKey[] = "AdsData"; | |
113 | |
114 } // namespace | |
115 | |
116 const AdTypes& GetDetectedAdTypes( | |
117 content::NavigationHandle* navigation_handle) { | |
118 DCHECK(navigation_handle); | |
119 | |
120 // TODO(lukasza): https://crbug.com/746638: Add a DCHECK that verifies that | |
121 // GetDetectedAdTypes is called not earlier than when the |navigation_handle| | |
122 // is ready to commit. | |
123 | |
124 NavigationHandleAdsData* ads_data = | |
125 NavigationHandleAdsData::GetOrCreate(navigation_handle); | |
126 ads_data->CalculateGoogleAdsIfNecessary(navigation_handle); | |
127 return ads_data->ad_types(); | |
128 } | |
129 | |
130 void SetDetectedAdType(content::NavigationHandle* navigation_handle, | |
131 AdType type) { | |
132 DCHECK(navigation_handle); | |
133 NavigationHandleAdsData::GetOrCreate(navigation_handle)->ad_types().set(type); | |
134 } | |
135 | |
136 } // namespace page_load_metrics | |
OLD | NEW |