Hooking MHTML generation to the browser.

This CL adds a class that can be used to generate MHTML for the current
page of a tab.
It is not yet surfaced in the UI.

BUG=None
TEST=Run the browser tests.


Review URL: http://codereview.chromium.org/7044095

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@89047 0039d316-1c4b-4281-b951-d872f2087c98

8 files changed, 310 insertions, 2 deletions

diff --git a/chrome/browser/browser_process.h b/chrome/browser/browser_process.h
index 3351d0b..82e7970 100644
--- a/chrome/browser/browser_process.h
+++ b/chrome/browser/browser_process.h
@@ -30,6 +30,7 @@ class IconManager;
 class IntranetRedirectDetector;
 class IOThread;
 class MetricsService;
+class MHTMLGenerationManager;
 class NotificationUIManager;
 class PrefService;
 class ProfileManager;
@@ -239,6 +240,8 @@ class BrowserProcess {
     plugin_data_remover_mime_type_ = mime_type;
   }
 
+  virtual MHTMLGenerationManager* mhtml_generation_manager() = 0;
+
  private:
   // User-data-dir based profiles.
   std::vector<std::wstring> user_data_dir_profiles_;
diff --git a/chrome/browser/browser_process_impl.cc b/chrome/browser/browser_process_impl.cc
index 771f67e..41c846c 100644
--- a/chrome/browser/browser_process_impl.cc
+++ b/chrome/browser/browser_process_impl.cc
@@ -26,6 +26,7 @@
 #include "chrome/browser/debugger/devtools_manager.h"
 #include "chrome/browser/debugger/devtools_protocol_handler.h"
 #include "chrome/browser/download/download_file_manager.h"
+#include "chrome/browser/download/mhtml_generation_manager.h"
 #include "chrome/browser/download/save_file_manager.h"
 #include "chrome/browser/extensions/extension_event_router_forwarder.h"
 #include "chrome/browser/extensions/extension_tab_id_map.h"
@@ -691,6 +692,13 @@ prerender::PrerenderTracker* BrowserProcessImpl::prerender_tracker() {
   return prerender_tracker_.get();
 }
 
+MHTMLGenerationManager* BrowserProcessImpl::mhtml_generation_manager() {
+  if (!mhtml_generation_manager_.get())
+    mhtml_generation_manager_ = new MHTMLGenerationManager();
+
+  return mhtml_generation_manager_.get();
+}
+
 void BrowserProcessImpl::ClearLocalState(const FilePath& profile_path) {
   webkit_database::DatabaseTracker::ClearLocalState(profile_path);
   BrowsingDataRemover::ClearGearsData(profile_path);
diff --git a/chrome/browser/browser_process_impl.h b/chrome/browser/browser_process_impl.h
index 0d32964..adb5a1f 100644
--- a/chrome/browser/browser_process_impl.h
+++ b/chrome/browser/browser_process_impl.h
@@ -116,6 +116,8 @@ class BrowserProcessImpl : public BrowserProcess,
   virtual void SetIPCLoggingEnabled(bool enable);
 #endif
 
+  virtual MHTMLGenerationManager* mhtml_generation_manager();
+
  private:
   void ClearLocalState(const FilePath& profile_path);
   bool ShouldClearLocalState(FilePath* profile_path);
@@ -292,6 +294,8 @@ class BrowserProcessImpl : public BrowserProcess,
   NotificationRegistrar notification_registrar_;
   scoped_refptr<PluginDataRemover> plugin_data_remover_;
 
+  scoped_refptr<MHTMLGenerationManager> mhtml_generation_manager_;
+
   // Monitors the state of the 'DisablePluginFinder' policy.
   BooleanPrefMember plugin_finder_disabled_pref_;
 
diff --git a/chrome/browser/download/mhtml_generation_browsertest.cc b/chrome/browser/download/mhtml_generation_browsertest.cc
new file mode 100644
index 0000000..6077b02
--- /dev/null
+++ b/chrome/browser/download/mhtml_generation_browsertest.cc
@@ -0,0 +1,66 @@
+// 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.
+
+#include "base/file_path.h"
+#include "base/scoped_temp_dir.h"
+#include "chrome/browser/download/mhtml_generation_manager.h"
+#include "chrome/browser/ui/browser.h"
+#include "chrome/browser/ui/tab_contents/tab_contents_wrapper.h"
+#include "chrome/test/in_process_browser_test.h"
+#include "chrome/test/testing_browser_process.h"
+#include "chrome/test/ui_test_utils.h"
+#include "content/browser/tab_contents/tab_contents.h"
+#include "net/test/test_server.h"
+#include "testing/gtest/include/gtest/gtest.h"
+
+namespace {
+
+class MHTMLGenerationTest : public InProcessBrowserTest {
+ public:
+  MHTMLGenerationTest() {}
+
+ protected:
+  virtual void SetUp() {
+    ASSERT_TRUE(temp_dir_.CreateUniqueTempDir());
+    InProcessBrowserTest::SetUp();
+  }
+
+  ScopedTempDir temp_dir_;
+};
+
+// Tests that generating a MHTML does create contents.
+// Note that the actual content of the file is not tested, the purpose of this
+// test is to ensure we were successfull in creating the MHTML data from the
+// renderer.
+IN_PROC_BROWSER_TEST_F(MHTMLGenerationTest, GenerateMHTML) {
+  ASSERT_TRUE(test_server()->Start());
+
+  FilePath path(temp_dir_.path());
+  path = path.Append(FILE_PATH_LITERAL("test.mht"));
+
+  ui_test_utils::NavigateToURL(browser(),
+      test_server()->GetURL("files/google/google.html"));
+
+  TabContents* tab = browser()->GetSelectedTabContentsWrapper()->tab_contents();
+  MHTMLGenerationManager* mhtml_generation_manager =
+      g_browser_process->mhtml_generation_manager();
+
+  Source<RenderViewHost> source(tab->render_view_host());
+  ui_test_utils::WindowedNotificationObserverWithDetails<
+      MHTMLGenerationManager::NotificationDetails> signal(
+          NotificationType::MHTML_GENERATED, source);
+  mhtml_generation_manager->GenerateMHTML(tab, path);
+  signal.Wait();
+
+  MHTMLGenerationManager::NotificationDetails details;
+  ASSERT_TRUE(signal.GetDetailsFor(source.map_key(), &details));
+  ASSERT_TRUE(details.success);
+
+  // Make sure the generated file has some contents.
+  int64 file_size;
+  ASSERT_TRUE(file_util::GetFileSize(path, &file_size));
+  EXPECT_GT(file_size, 100);
+}
+
+}  // namespace
diff --git a/chrome/browser/download/mhtml_generation_manager.cc b/chrome/browser/download/mhtml_generation_manager.cc
new file mode 100644
index 0000000..ea98155
--- /dev/null
+++ b/chrome/browser/download/mhtml_generation_manager.cc
@@ -0,0 +1,136 @@
+// 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.
+
+#include "chrome/browser/download/mhtml_generation_manager.h"
+
+#include "base/platform_file.h"
+#include "chrome/browser/tab_contents/tab_util.h"
+#include "chrome/common/render_messages.h"
+#include "content/browser/renderer_host/render_process_host.h"
+#include "content/browser/renderer_host/render_view_host.h"
+#include "content/browser/tab_contents/tab_contents.h"
+#include "content/common/notification_service.h"
+#include "content/common/page_transition_types.h"
+
+MHTMLGenerationManager::Job::Job()
+    : browser_file(base::kInvalidPlatformFileValue),
+      renderer_file(IPC::InvalidPlatformFileForTransit()),
+      process_id(-1),
+      routing_id(-1) {
+}
+
+MHTMLGenerationManager::MHTMLGenerationManager() {
+}
+
+MHTMLGenerationManager::~MHTMLGenerationManager() {
+}
+
+void MHTMLGenerationManager::GenerateMHTML(TabContents* tab_contents,
+                                           const FilePath& file) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
+  static int id_counter = 0;
+
+  int job_id = id_counter++;
+  Job job;
+  job.file_path = file;
+  job.process_id = tab_contents->GetRenderProcessHost()->id();
+  job.routing_id = tab_contents->render_view_host()->routing_id();
+  id_to_job_[job_id] = job;
+
+  base::ProcessHandle renderer_process =
+      tab_contents->GetRenderProcessHost()->GetHandle();
+  BrowserThread::PostTask(BrowserThread::FILE, FROM_HERE,
+      NewRunnableMethod(this, &MHTMLGenerationManager::CreateFile,
+                        job_id, file, renderer_process));
+}
+
+void MHTMLGenerationManager::MHTMLGenerated(int job_id, bool success) {
+  JobFinished(job_id, success);
+}
+
+void MHTMLGenerationManager::CreateFile(int job_id, const FilePath& file_path,
+    base::ProcessHandle renderer_process) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::FILE));
+  base::PlatformFile browser_file = base::CreatePlatformFile(file_path,
+      base::PLATFORM_FILE_CREATE_ALWAYS | base::PLATFORM_FILE_WRITE,
+      NULL, NULL);
+  if (browser_file == base::kInvalidPlatformFileValue) {
+    LOG(ERROR) << "Failed to create file to save MHTML at: " <<
+        file_path.value();
+  }
+
+  IPC::PlatformFileForTransit renderer_file =
+      IPC::GetFileHandleForProcess(browser_file, renderer_process, false);
+
+  BrowserThread::PostTask(BrowserThread::UI, FROM_HERE,
+      NewRunnableMethod(this, &MHTMLGenerationManager::FileCreated,
+                        job_id, browser_file, renderer_file));
+}
+
+void MHTMLGenerationManager::FileCreated(int job_id,
+    base::PlatformFile browser_file,
+    IPC::PlatformFileForTransit renderer_file) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
+  if (browser_file == base::kInvalidPlatformFileValue) {
+    LOG(ERROR) << "Failed to create file";
+    JobFinished(job_id, false);
+    return;
+  }
+
+  IDToJobMap::iterator iter = id_to_job_.find(job_id);
+  if (iter == id_to_job_.end()) {
+    NOTREACHED();
+    return;
+  }
+
+  Job& job = iter->second;
+  job.browser_file = browser_file;
+  job.renderer_file = renderer_file;
+
+  TabContents* tab_contents =
+      tab_util::GetTabContentsByID(job.process_id, job.routing_id);
+  if (!tab_contents) {
+    // The tab went away.
+    JobFinished(job_id, false);
+    return;
+  }
+
+  RenderViewHost* host = tab_contents->render_view_host();
+  host->Send(new ViewMsg_SavePageAsMHTML(host->routing_id(), job_id,
+                                         renderer_file));
+}
+
+void MHTMLGenerationManager::JobFinished(int job_id, bool success) {
+  IDToJobMap::iterator iter = id_to_job_.find(job_id);
+  if (iter == id_to_job_.end()) {
+    NOTREACHED();
+    return;
+  }
+
+  Job& job = iter->second;
+
+  TabContents* tab_contents =
+      tab_util::GetTabContentsByID(job.process_id, job.routing_id);
+  if (tab_contents) {
+    NotificationDetails details;
+    details.file_path = job.file_path;
+    details.success = success;
+
+    NotificationService::current()->Notify(
+        NotificationType::MHTML_GENERATED,
+        Source<RenderViewHost>(tab_contents->render_view_host()),
+        Details<NotificationDetails>(&details));
+  }
+
+  BrowserThread::PostTask(BrowserThread::FILE, FROM_HERE,
+      NewRunnableMethod(this, &MHTMLGenerationManager::CloseFile,
+                        job.browser_file));
+
+  id_to_job_.erase(job_id);
+}
+
+void MHTMLGenerationManager::CloseFile(base::PlatformFile file) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::FILE));
+  base::ClosePlatformFile(file);
+}
diff --git a/chrome/browser/download/mhtml_generation_manager.h b/chrome/browser/download/mhtml_generation_manager.h
new file mode 100644
index 0000000..8e8557f
--- /dev/null
+++ b/chrome/browser/download/mhtml_generation_manager.h
@@ -0,0 +1,81 @@
+// 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_DOWNLOAD_MHTML_GENERATION_MANAGER_H_
+#define CHROME_BROWSER_DOWNLOAD_MHTML_GENERATION_MANAGER_H_
+
+#include <map>
+
+#include "base/memory/ref_counted.h"
+#include "base/platform_file.h"
+#include "base/process.h"
+#include "content/browser/browser_thread.h"
+#include "ipc/ipc_platform_file.h"
+
+class FilePath;
+class TabContents;
+
+class MHTMLGenerationManager
+    : public base::RefCountedThreadSafe<MHTMLGenerationManager,
+                                        BrowserThread::DeleteOnUIThread> {
+ public:
+  MHTMLGenerationManager();
+  ~MHTMLGenerationManager();
+
+  // Instructs the render view to generate a MHTML representation of the current
+  // page for |tab_contents|.
+  void GenerateMHTML(TabContents* tab_contents, const FilePath& file);
+
+  // Notification from the renderer that the MHTML generation succeeded/failed.
+  void MHTMLGenerated(int job_id, bool success);
+
+  // The details sent along with the MHTML_GENERATED notification.
+  struct NotificationDetails {
+    FilePath file_path;
+    bool success;
+  };
+
+ private:
+  struct Job{
+    Job();
+
+    FilePath file_path;
+
+    // The handles to file the MHTML is saved to, for the browser and renderer
+    // processes.
+    base::PlatformFile browser_file;
+    IPC::PlatformFileForTransit renderer_file;
+
+    // The IDs mapping to a specific tab.
+    int process_id;
+    int routing_id;
+  };
+
+  // Called on the file thread to create |file|.
+  void CreateFile(int job_id,
+                  const FilePath& file,
+                  base::ProcessHandle renderer_process);
+
+  // Called on the UI thread when the file that should hold the MHTML data has
+  // been created.  This returns a handle to that file for the browser process
+  // and one for the renderer process. These handles are
+  // kInvalidPlatformFileValue if the file could not be opened.
+  void FileCreated(int job_id,
+                   base::PlatformFile browser_file,
+                   IPC::PlatformFileForTransit renderer_file);
+
+  // Called on the file thread to close the file the MHTML was saved to.
+  void CloseFile(base::PlatformFile file);
+
+  // Called on the UI thread when a job has been processed (successfully or
+  // not).  Closes the file and removes the job from the job map.
+  void JobFinished(int job_id, bool success);
+
+  typedef std::map<int, Job> IDToJobMap;
+  IDToJobMap id_to_job_;
+
+  DISALLOW_COPY_AND_ASSIGN(MHTMLGenerationManager);
+};
+
+#endif  // CHROME_BROWSER_DOWNLOAD_MHTML_GENERATION_MANAGER_H_
diff --git a/chrome/browser/renderer_host/chrome_render_message_filter.cc b/chrome/browser/renderer_host/chrome_render_message_filter.cc
index d7393aa..15852e2 100644
--- a/chrome/browser/renderer_host/chrome_render_message_filter.cc
+++ b/chrome/browser/renderer_host/chrome_render_message_filter.cc
@@ -10,6 +10,7 @@
 #include "chrome/browser/browser_process.h"
 #include "chrome/browser/content_settings/host_content_settings_map.h"
 #include "chrome/browser/content_settings/tab_specific_content_settings.h"
+#include "chrome/browser/download/mhtml_generation_manager.h"
 #include "chrome/browser/extensions/extension_event_router.h"
 #include "chrome/browser/extensions/extension_function_dispatcher.h"
 #include "chrome/browser/extensions/extension_message_service.h"
@@ -98,6 +99,7 @@ bool ChromeRenderMessageFilter::OnMessageReceived(const IPC::Message& message,
     IPC_MESSAGE_HANDLER(ViewHostMsg_CanTriggerClipboardWrite,
                         OnCanTriggerClipboardWrite)
     IPC_MESSAGE_HANDLER(ViewHostMsg_ClearPredictorCache, OnClearPredictorCache)
+    IPC_MESSAGE_HANDLER(ViewHostMsg_SavedPageAsMHTML, OnSavedPageAsMHTML)
     IPC_MESSAGE_UNHANDLED(handled = false)
   IPC_END_MESSAGE_MAP()
 
@@ -137,10 +139,10 @@ void ChromeRenderMessageFilter::OverrideThreadForMessage(
     case ExtensionHostMsg_AddListener::ID:
     case ExtensionHostMsg_RemoveListener::ID:
     case ExtensionHostMsg_CloseChannel::ID:
-      *thread = BrowserThread::UI;
-      break;
     case ViewHostMsg_UpdatedCacheStats::ID:
+    case ViewHostMsg_SavedPageAsMHTML::ID:
       *thread = BrowserThread::UI;
+      break;
     default:
       break;
   }
@@ -500,3 +502,10 @@ void ChromeRenderMessageFilter::OnSetCookie(const IPC::Message& message,
   AutomationResourceMessageFilter::SetCookiesForUrl(
       render_process_id_, message.routing_id(), url, cookie);
 }
+
+
+void ChromeRenderMessageFilter::OnSavedPageAsMHTML(int job_id, bool success) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
+  g_browser_process->mhtml_generation_manager()->
+      MHTMLGenerated(job_id, success);
+}
diff --git a/chrome/browser/renderer_host/chrome_render_message_filter.h b/chrome/browser/renderer_host/chrome_render_message_filter.h
index 867b824..55590d2 100644
--- a/chrome/browser/renderer_host/chrome_render_message_filter.h
+++ b/chrome/browser/renderer_host/chrome_render_message_filter.h
@@ -124,6 +124,7 @@ class ChromeRenderMessageFilter : public BrowserMessageFilter {
                    const GURL& url,
                    const GURL& first_party_for_cookies,
                    const std::string& cookie);
+  void OnSavedPageAsMHTML(int job_id, bool success);
 
   int render_process_id_;
   ProfileId profile_id_;