Querying the history service to get the redirect information for urls.

Bug=60831
TEST=unit_tests
Review URL: http://codereview.chromium.org/6710004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@86401 0039d316-1c4b-4281-b951-d872f2087c98

5 files changed, 329 insertions, 3 deletions

diff --git a/chrome/browser/safe_browsing/malware_details.cc b/chrome/browser/safe_browsing/malware_details.cc
index 63d3f3a..bc35cec 100644
--- a/chrome/browser/safe_browsing/malware_details.cc
+++ b/chrome/browser/safe_browsing/malware_details.cc
@@ -11,6 +11,7 @@
 #include "chrome/browser/net/chrome_url_request_context.h"
 #include "chrome/browser/profiles/profile.h"
 #include "chrome/browser/safe_browsing/malware_details_cache.h"
+#include "chrome/browser/safe_browsing/malware_details_history.h"
 #include "chrome/browser/safe_browsing/report.pb.h"
 #include "chrome/browser/safe_browsing/safe_browsing_service.h"
 #include "chrome/common/safe_browsing/safebrowsing_messages.h"
@@ -76,7 +77,8 @@ MalwareDetails::MalwareDetails(
       request_context_getter_(tab_contents->profile()->GetRequestContext()),
       sb_service_(sb_service),
       resource_(resource),
-      cache_collector_(new MalwareDetailsCacheCollector) {
+      cache_collector_(new MalwareDetailsCacheCollector),
+      redirects_collector_(new MalwareDetailsRedirectsCollector) {
   StartCollection();
 }
 
@@ -224,6 +226,11 @@ void MalwareDetails::AddDOMDetails(
   DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
   DVLOG(1) << "Nodes from the DOM: " << params.size();
 
+  // If we have already started getting redirects from history service,
+  // don't modify state, otherwise will invalidate the iterators.
+  if (redirects_collector_->HasStarted())
+    return;
+
   // If we have already started collecting data from the HTTP cache, don't
   // modify our state.
   if (cache_collector_->HasStarted())
@@ -246,6 +253,25 @@ void MalwareDetails::AddDOMDetails(
 void MalwareDetails::FinishCollection() {
   DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
 
+  std::vector<GURL> urls;
+  for (safe_browsing::ResourceMap::const_iterator it = resources_.begin();
+       it != resources_.end(); it++) {
+    urls.push_back(GURL(it->first));
+  }
+  redirects_collector_->StartHistoryCollection(
+      urls, tab_contents(),
+      NewRunnableMethod(this, &MalwareDetails::OnRedirectionCollectionReady));
+}
+
+void MalwareDetails::OnRedirectionCollectionReady() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  const std::vector<safe_browsing::RedirectChain>& redirects =
+      redirects_collector_->GetCollectedUrls();
+
+  for (size_t i = 0; i < redirects.size(); ++i)
+    AddRedirectUrlList(redirects[i]);
+
+  // Call the cache collector
   cache_collector_->StartCacheCollection(
       request_context_getter_,
       &resources_,
@@ -253,6 +279,13 @@ void MalwareDetails::FinishCollection() {
       NewRunnableMethod(this, &MalwareDetails::OnCacheCollectionReady));
 }
 
+void MalwareDetails::AddRedirectUrlList(const std::vector<GURL>& urls) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  for (size_t i = 0; i < urls.size()-1; ++i) {
+    AddUrl(urls[i], urls[i+1], "", NULL);
+  }
+}
+
 void MalwareDetails::OnCacheCollectionReady() {
   DVLOG(1) << "OnCacheCollectionReady.";
   // Add all the urls in our |resources_| maps to the |report_| protocol buffer.
diff --git a/chrome/browser/safe_browsing/malware_details.h b/chrome/browser/safe_browsing/malware_details.h
index 494837e..473eab8 100644
--- a/chrome/browser/safe_browsing/malware_details.h
+++ b/chrome/browser/safe_browsing/malware_details.h
@@ -21,6 +21,7 @@
 #include "base/memory/scoped_ptr.h"
 #include "chrome/browser/safe_browsing/report.pb.h"
 #include "chrome/browser/safe_browsing/safe_browsing_service.h"
+#include "content/browser/browser_thread.h"
 #include "content/browser/tab_contents/tab_contents_observer.h"
 #include "net/base/completion_callback.h"
 
@@ -28,6 +29,7 @@ class TabContents;
 struct SafeBrowsingHostMsg_MalwareDOMDetails_Node;
 
 class MalwareDetailsCacheCollector;
+class MalwareDetailsRedirectsCollector;
 class MalwareDetailsFactory;
 
 namespace safe_browsing {
@@ -54,12 +56,15 @@ class MalwareDetails : public base::RefCountedThreadSafe<MalwareDetails>,
 
   // The SafeBrowsingBlockingPage calls this from the IO thread when
   // the user is leaving the blocking page and has opted-in to sending
-  // the report. We start the cache collection, and when we are done,
-  // we send the report.
+  // the report. We start the redirection urls collection from history service
+  // in UI thread; then do cache collection back in IO thread.
+  // When we are done, we send the report.
   void FinishCollection();
 
   void OnCacheCollectionReady();
 
+  void OnRedirectionCollectionReady();
+
   // TabContentsObserver implementation.
   virtual bool OnMessageReceived(const IPC::Message& message);
 
@@ -108,6 +113,8 @@ class MalwareDetails : public base::RefCountedThreadSafe<MalwareDetails>,
   void OnReceivedMalwareDOMDetails(
       const std::vector<SafeBrowsingHostMsg_MalwareDOMDetails_Node>& params);
 
+  void AddRedirectUrlList(const std::vector<GURL>& urls);
+
   scoped_refptr<SafeBrowsingService> sb_service_;
 
   const SafeBrowsingService::UnsafeResource resource_;
@@ -127,9 +134,13 @@ class MalwareDetails : public base::RefCountedThreadSafe<MalwareDetails>,
   // Used to collect details from the HTTP Cache.
   scoped_refptr<MalwareDetailsCacheCollector> cache_collector_;
 
+  // Used to collect redirect urls from the history service
+  scoped_refptr<MalwareDetailsRedirectsCollector> redirects_collector_;
+
   FRIEND_TEST_ALL_PREFIXES(MalwareDetailsTest, MalwareDOMDetails);
   FRIEND_TEST_ALL_PREFIXES(MalwareDetailsTest, HTTPCache);
   FRIEND_TEST_ALL_PREFIXES(MalwareDetailsTest, HTTPCacheNoEntries);
+  FRIEND_TEST_ALL_PREFIXES(MalwareDetailsTest, HistoryServiceUrls);
 
   DISALLOW_COPY_AND_ASSIGN(MalwareDetails);
 };
diff --git a/chrome/browser/safe_browsing/malware_details_history.cc b/chrome/browser/safe_browsing/malware_details_history.cc
new file mode 100644
index 0000000..e28c729
--- /dev/null
+++ b/chrome/browser/safe_browsing/malware_details_history.cc
@@ -0,0 +1,121 @@
+// Copyright (c) 2011 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.
+//
+// Implementation of the MalwareDetailsRedirectsCollector class.
+
+#include "chrome/browser/safe_browsing/malware_details_history.h"
+
+#include "chrome/browser/profiles/profile.h"
+#include "chrome/browser/safe_browsing/malware_details.h"
+#include "content/browser/browser_thread.h"
+#include "content/browser/renderer_host/render_view_host.h"
+#include "content/browser/tab_contents/navigation_entry.h"
+#include "content/browser/tab_contents/tab_contents.h"
+
+
+MalwareDetailsRedirectsCollector::MalwareDetailsRedirectsCollector()
+    : history_(NULL),
+      has_started_(false) {
+}
+
+MalwareDetailsRedirectsCollector::~MalwareDetailsRedirectsCollector() {
+}
+
+void MalwareDetailsRedirectsCollector::StartHistoryCollection(
+    const std::vector<GURL>& urls,
+    TabContents* tab_contents,
+    Task* callback) {
+  DVLOG(1) << "Num of urls to check in history service: " << urls.size();
+  tab_contents_ = tab_contents;
+  has_started_ = true;
+  callback_ = callback;
+
+  if (urls.size() == 0) {
+    AllDone();
+    return;
+  }
+
+  BrowserThread::PostTask(
+      BrowserThread::UI, FROM_HERE,
+      NewRunnableMethod(
+          this, &MalwareDetailsRedirectsCollector::StartGetRedirects, urls));
+}
+
+void MalwareDetailsRedirectsCollector::StartGetRedirects(
+        const std::vector<GURL>& urls) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
+
+  // History access from profile needs to happen in UI thread
+  if (history_ == NULL) {
+    DVLOG(1) << "Getting access to history service";
+    history_ = tab_contents_->profile()->GetHistoryService(
+        Profile::EXPLICIT_ACCESS);
+  }
+
+  if (history_ == NULL) {
+    DVLOG(1) << "Could not access history service.";
+    AllDone();
+    return;
+  }
+
+  for (size_t i = 0; i < urls.size(); ++i)
+    urls_.push_back(urls[i]);
+
+  urls_it_ = urls_.begin();
+  GetRedirects(*urls_it_);
+}
+
+void MalwareDetailsRedirectsCollector::GetRedirects(const GURL& url) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
+
+  history_->QueryRedirectsTo(
+      url,
+      &request_consumer_,
+      NewCallback(this,
+                  &MalwareDetailsRedirectsCollector::OnGotQueryRedirectsTo));
+}
+
+void MalwareDetailsRedirectsCollector::OnGotQueryRedirectsTo(
+    HistoryService::Handle handle,
+    GURL url,
+    bool success,
+    history::RedirectList* redirect_list) {
+
+  if (success && redirect_list->size() > 0) {
+    std::vector<GURL> urllist;
+    urllist.push_back(url);
+    for (size_t i = 0; i < redirect_list->size(); i++) {
+      urllist.push_back(redirect_list->at(i));
+    }
+    redirects_urls_.push_back(urllist);
+  }
+
+  // Proceed to next url
+  ++urls_it_;
+
+  if (urls_it_ == urls_.end()) {
+    AllDone();
+    return;
+  }
+
+  GetRedirects(*urls_it_);
+}
+
+bool MalwareDetailsRedirectsCollector::HasStarted() const {
+  return has_started_;
+}
+
+void MalwareDetailsRedirectsCollector::AllDone() {
+  DVLOG(1) << "AllDone";
+  BrowserThread::PostTask(BrowserThread::IO, FROM_HERE, callback_);
+}
+
+const std::vector<safe_browsing::RedirectChain>&
+MalwareDetailsRedirectsCollector::GetCollectedUrls() const {
+  return redirects_urls_;
+}
+
+void MalwareDetailsRedirectsCollector::SetHistory(HistoryService* history) {
+  history_ = history;
+}
diff --git a/chrome/browser/safe_browsing/malware_details_history.h b/chrome/browser/safe_browsing/malware_details_history.h
new file mode 100644
index 0000000..23a2571
--- /dev/null
+++ b/chrome/browser/safe_browsing/malware_details_history.h
@@ -0,0 +1,82 @@
+// Copyright (c) 2011 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.
+
+#ifndef CHROME_BROWSER_SAFE_BROWSING_MALWARE_DETAILS_HISTORY_H_
+#define CHROME_BROWSER_SAFE_BROWSING_MALWARE_DETAILS_HISTORY_H_
+#pragma once
+
+// This class gets redirect chain for urls from the history service.
+
+#include <string>
+#include <vector>
+
+#include "base/hash_tables.h"
+#include "base/memory/linked_ptr.h"
+#include "base/memory/ref_counted.h"
+#include "chrome/browser/history/history.h"
+#include "net/base/completion_callback.h"
+
+class TabContents;
+
+namespace safe_browsing {
+typedef std::vector<GURL> RedirectChain;
+}
+
+class MalwareDetailsRedirectsCollector
+    : public base::RefCountedThreadSafe<MalwareDetailsRedirectsCollector> {
+
+ public:
+  MalwareDetailsRedirectsCollector();
+  virtual ~MalwareDetailsRedirectsCollector();
+
+  // Collects urls' redirects chain information from the history service.
+  // We get access to history service via tab_contents in UI thread.
+  // Notice the callback will be posted to the IO thread.
+  void StartHistoryCollection(const std::vector<GURL>& urls,
+                              TabContents* tab_contents,
+                              Task* callback);
+
+  // Returns whether or not StartCacheCollection has been called.
+  bool HasStarted() const;
+
+  // Returns the redirect urls we get from history service
+  const std::vector<safe_browsing::RedirectChain>& GetCollectedUrls() const;
+
+  // Sets the pointer to the history service. Use it in unittest.
+  void SetHistory(HistoryService* history);
+
+ private:
+  TabContents* tab_contents_;
+  HistoryService* history_;
+  CancelableRequestConsumer request_consumer_;
+
+  // Method we call when we are done. The caller must be alive for the
+  // whole time, we are modifying its state (see above).
+  Task* callback_;
+
+  // Sets to true once StartHistoryCollection is called
+  bool has_started_;
+
+  // The urls we need to get redirects for
+  std::vector<GURL> urls_;
+  // The iterator goes over urls_
+  std::vector<GURL>::iterator urls_it_;
+  // The collected directs from history service
+  std::vector<safe_browsing::RedirectChain> redirects_urls_;
+
+  void StartGetRedirects(const std::vector<GURL>& urls);
+  void GetRedirects(const GURL& url);
+  void OnGotQueryRedirectsTo(HistoryService::Handle handle,
+                             GURL url,
+                             bool success,
+                             history::RedirectList* redirect_list);
+
+  // Posts the callback method back to IO thread when redirects collecting
+  // is all done.
+  void AllDone();
+
+  DISALLOW_COPY_AND_ASSIGN(MalwareDetailsRedirectsCollector);
+};
+
+#endif  // CHROME_BROWSER_SAFE_BROWSING_MALWARE_DETAILS_HISTORY_H_
diff --git a/chrome/browser/safe_browsing/malware_details_unittest.cc b/chrome/browser/safe_browsing/malware_details_unittest.cc
index 73ccfdd..60a65f8 100644
--- a/chrome/browser/safe_browsing/malware_details_unittest.cc
+++ b/chrome/browser/safe_browsing/malware_details_unittest.cc
@@ -6,8 +6,11 @@
 
 #include "base/pickle.h"
 #include "base/time.h"
+#include "chrome/browser/history/history.h"
+#include "chrome/browser/history/history_backend.h"
 #include "chrome/browser/profiles/profile.h"
 #include "chrome/browser/safe_browsing/malware_details.h"
+#include "chrome/browser/safe_browsing/malware_details_history.h"
 #include "chrome/browser/safe_browsing/report.pb.h"
 #include "chrome/common/render_messages.h"
 #include "chrome/common/safe_browsing/safebrowsing_messages.h"
@@ -176,10 +179,13 @@ class MalwareDetailsTest : public RenderViewHostTestHarness {
     // The URLFetcher checks that the messageloop type is IO.
     ASSERT_TRUE(io_thread_.StartWithOptions(
         base::Thread::Options(MessageLoop::TYPE_IO, 0)));
+
+    profile_->CreateHistoryService(true /* delete_file */, false /* no_db */);
   }
 
   virtual void TearDown() {
     io_thread_.Stop();
+    profile_->DestroyHistoryService();
     RenderViewHostTestHarness::TearDown();
   }
 
@@ -201,6 +207,10 @@ class MalwareDetailsTest : public RenderViewHostTestHarness {
     return sb_service_->GetSerialized();
   }
 
+  HistoryService* history_service() {
+    return profile_->GetHistoryService(Profile::EXPLICIT_ACCESS);
+  }
+
  protected:
   void InitResource(SafeBrowsingService::UnsafeResource* resource,
                     ResourceType::Type resource_type,
@@ -283,6 +293,18 @@ class MalwareDetailsTest : public RenderViewHostTestHarness {
               resource->response().remote_ip());
   }
 
+  // Adds a page to history.
+  // The redirects is the redirect url chain leading to the url.
+  void AddPageToHistory(const GURL& url,
+                        history::RedirectList* redirects) {
+    // The last item of the redirect chain has to be the final url when adding
+    // to history backend.
+    redirects->push_back(url);
+    history_service()->AddPage(
+        url, static_cast<void*>(this), 0, GURL(), PageTransition::TYPED,
+        *redirects, history::SOURCE_BROWSED, false);
+  }
+
   BrowserThread ui_thread_;
   BrowserThread io_thread_;
   scoped_refptr<MockSafeBrowsingService> sb_service_;
@@ -603,3 +625,60 @@ TEST_F(MalwareDetailsTest, HTTPCacheNoEntries) {
 
   VerifyResults(actual, expected);
 }
+
+// Test getting redirects from history service.
+TEST_F(MalwareDetailsTest, HistoryServiceUrls) {
+  // Add content to history service.
+  // There are two redirect urls before reacing malware url:
+  // kFirstRedirectURL -> kSecondRedirectURL -> kMalwareURL
+  GURL baseurl(kMalwareURL);
+  history::RedirectList redirects;
+  redirects.push_back(GURL(kFirstRedirectURL));
+  redirects.push_back(GURL(kSecondRedirectURL));
+  AddPageToHistory(baseurl, &redirects);
+  // Wait for history service operation finished.
+  profile_->BlockUntilHistoryProcessesPendingRequests();
+
+  controller().LoadURL(GURL(kLandingURL), GURL(), PageTransition::TYPED);
+
+  SafeBrowsingService::UnsafeResource resource;
+  InitResource(&resource, ResourceType::SUB_RESOURCE, GURL(kMalwareURL));
+  scoped_refptr<MalwareDetailsWrap> report = new MalwareDetailsWrap(
+      sb_service_.get(), contents(), resource, NULL);
+
+  // Reset the history service pointer.
+  report->redirects_collector_->SetHistory(history_service());
+
+  // The redirects collection starts after the IPC from the DOM is fired.
+  std::vector<SafeBrowsingHostMsg_MalwareDOMDetails_Node> params;
+  report->OnReceivedMalwareDOMDetails(params);
+
+  // Let the redirects callbacks complete.
+  MessageLoop::current()->RunAllPending();
+
+  std::string serialized = WaitForSerializedReport(report);
+  ClientMalwareReportRequest actual;
+  actual.ParseFromString(serialized);
+
+  ClientMalwareReportRequest expected;
+  expected.set_malware_url(kMalwareURL);
+  expected.set_page_url(kLandingURL);
+  expected.set_referrer_url("");
+
+  ClientMalwareReportRequest::Resource* pb_resource = expected.add_resources();
+  pb_resource->set_id(0);
+  pb_resource->set_url(kLandingURL);
+  pb_resource = expected.add_resources();
+  pb_resource->set_id(1);
+  pb_resource->set_parent_id(2);
+  pb_resource->set_url(kMalwareURL);
+  pb_resource = expected.add_resources();
+  pb_resource->set_id(2);
+  pb_resource->set_parent_id(3);
+  pb_resource->set_url(kSecondRedirectURL);
+  pb_resource = expected.add_resources();
+  pb_resource->set_id(3);
+  pb_resource->set_url(kFirstRedirectURL);
+
+  VerifyResults(actual, expected);
+}