Save Page As MHTML

content/browser/download/save_package.cc

 // Copyright (c) 2012 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 "content/browser/download/save_package.h"
 
 #include <algorithm>
 
 #include "base/bind.h"
 #include "base/file_path.h"
@@ skipping 113 matching lines @@
     : content::WebContentsObserver(web_contents),
       file_manager_(NULL),
       download_manager_(NULL),
       download_(NULL),
       page_url_(GetUrlToBeSaved()),
       saved_main_file_path_(file_full_path),
       saved_main_directory_path_(directory_full_path),
       title_(web_contents->GetTitle()),
       start_tick_(base::TimeTicks::Now()),
       finished_(false),
+      mhtml_finishing_(false),
       user_canceled_(false),
       disk_error_occurred_(false),
       save_type_(save_type),
       all_save_items_count_(0),
       wait_state_(INITIALIZE),
       contents_id_(web_contents->GetRenderProcessHost()->GetID()),
       unique_id_(g_save_package_id++),
       wrote_to_completed_file_(false),
       wrote_to_failed_file_(false) {
   DCHECK(page_url_.is_valid());
-  DCHECK(save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML ||
-         save_type_ == content::SAVE_PAGE_TYPE_AS_COMPLETE_HTML);
+  DCHECK((save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML) ||
+         (save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML) ||
+         (save_type_ == content::SAVE_PAGE_TYPE_AS_COMPLETE_HTML));
   DCHECK(!saved_main_file_path_.empty() &&
          saved_main_file_path_.value().length() <= kMaxFilePathLength);
   DCHECK(!saved_main_directory_path_.empty() &&
          saved_main_directory_path_.value().length() < kMaxFilePathLength);
   InternalInit();
 }
 
 SavePackage::SavePackage(WebContents* web_contents)
     : content::WebContentsObserver(web_contents),
       file_manager_(NULL),
       download_manager_(NULL),
       download_(NULL),
       page_url_(GetUrlToBeSaved()),
       title_(web_contents->GetTitle()),
       start_tick_(base::TimeTicks::Now()),
       finished_(false),
+      mhtml_finishing_(false),
       user_canceled_(false),
       disk_error_occurred_(false),
       save_type_(content::SAVE_PAGE_TYPE_UNKNOWN),
       all_save_items_count_(0),
       wait_state_(INITIALIZE),
       contents_id_(web_contents->GetRenderProcessHost()->GetID()),
       unique_id_(g_save_package_id++),
       wrote_to_completed_file_(false),
       wrote_to_failed_file_(false) {
   DCHECK(page_url_.is_valid());
   InternalInit();
 }
 
 // This is for testing use. Set |finished_| as true because we don't want
 // method Cancel to be be called in destructor in test mode.
 // We also don't call InternalInit().
 SavePackage::SavePackage(WebContents* web_contents,
                          const FilePath& file_full_path,
                          const FilePath& directory_full_path)
     : content::WebContentsObserver(web_contents),
       file_manager_(NULL),
       download_manager_(NULL),
       download_(NULL),
       saved_main_file_path_(file_full_path),
       saved_main_directory_path_(directory_full_path),
       start_tick_(base::TimeTicks::Now()),
       finished_(true),
+      mhtml_finishing_(false),
       user_canceled_(false),
       disk_error_occurred_(false),
       save_type_(content::SAVE_PAGE_TYPE_UNKNOWN),
       all_save_items_count_(0),
       wait_state_(INITIALIZE),
       contents_id_(0),
       unique_id_(g_save_package_id++),
       wrote_to_completed_file_(false),
       wrote_to_failed_file_(false) {
 }
@@ skipping 59 matching lines @@
 
   file_manager_ = rdh->save_file_manager();
   DCHECK(file_manager_);
 
   download_manager_ = web_contents()->GetBrowserContext()->GetDownloadManager();
   DCHECK(download_manager_);
 
   download_stats::RecordSavePackageEvent(download_stats::SAVE_PACKAGE_STARTED);
 }
 
-bool SavePackage::Init() {
+bool SavePackage::Init(
+    const content::SaveFileDownloadCreatedCallback& download_created_callback) {
   // Set proper running state.
   if (wait_state_ != INITIALIZE)
     return false;
 
   wait_state_ = START_PROCESS;
 
   // Initialize the request context and resource dispatcher.
   content::BrowserContext* browser_context =
       web_contents()->GetBrowserContext();
   if (!browser_context) {
     NOTREACHED();
     return false;
   }
 
   // The download manager keeps ownership but adds us as an observer.
   download_ = download_manager_->CreateSavePackageDownloadItem(
-      saved_main_file_path_, page_url_,
-      browser_context->IsOffTheRecord(), this);
+      saved_main_file_path_,
+      page_url_,
+      browser_context->IsOffTheRecord(),
+      ((save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML) ?
+       "multipart/related" : "text/html"),
+      this);
+  if (!download_created_callback.is_null())
+    download_created_callback.Run(download_);
 
   // Check save type and process the save page job.
   if (save_type_ == content::SAVE_PAGE_TYPE_AS_COMPLETE_HTML) {
     // Get directory
     DCHECK(!saved_main_directory_path_.empty());
     GetAllSavableResourceLinksForCurrentPage();
+  } else if (save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML) {
+    web_contents()->GenerateMHTML(saved_main_file_path_, base::Bind(
+        &SavePackage::OnMHTMLGenerated, this));
   } else {
     wait_state_ = NET_FILES;
     SaveFileCreateInfo::SaveFileSource save_source = page_url_.SchemeIsFile() ?
         SaveFileCreateInfo::SAVE_FILE_FROM_FILE :
         SaveFileCreateInfo::SAVE_FILE_FROM_NET;
     SaveItem* save_item = new SaveItem(page_url_,
                                        GURL(),
                                        this,
                                        save_source);
     // Add this item to waiting list.
     waiting_item_queue_.push(save_item);
     all_save_items_count_ = 1;
     download_->SetTotalBytes(1);
 
     DoSavingProcess();
   }
 
   return true;
 }
 
+void SavePackage::OnMHTMLGenerated(const FilePath& path, int64 size) {
+  if (size <= 0) {
+    Cancel(false);
+    return;
+  }
+  wrote_to_completed_file_ = true;
+  download_->SetTotalBytes(size);
+  download_->UpdateProgress(size, size, DownloadItem::kEmptyFileHash);
+  // Must call OnAllDataSaved here in order for
+  // GDataDownloadObserver::ShouldUpload() to return true.
+  // ShouldCompleteDownload() may depend on the gdata uploader to finish.
+  download_->OnAllDataSaved(size, DownloadItem::kEmptyFileHash);
+  // GDataDownloadObserver is waiting for the upload to complete. When that
+  // happens, it will call download_->MaybeCompleteDownload(), which will call
+  // through our OnDownloadUpdated() allowing us to Finish().
+  // OnDownloadUpdated() may have been called in OnAllDataSaved(), so |this| may
+  // be deleted at this point.
+}
+
 // On POSIX, the length of |pure_file_name| + |file_name_ext| is further
 // restricted by NAME_MAX. The maximum allowed path looks like:
 // '/path/to/save_dir' + '/' + NAME_MAX.
 uint32 SavePackage::GetMaxPathLengthForDirectory(const FilePath& base_dir) {
 #if defined(OS_POSIX)
   return std::min(kMaxFilePathLength,
                   static_cast<uint32>(base_dir.value().length()) +
                   NAME_MAX + 1);
 #else
   return kMaxFilePathLength;
@@ skipping 375 matching lines @@
        it != saved_failed_items_.end(); ++it)
     save_ids.push_back(it->second->save_id());
 
   BrowserThread::PostTask(
       BrowserThread::FILE, FROM_HERE,
       base::Bind(&SaveFileManager::RemoveSavedFileFromFileMap,
                  file_manager_,
                  save_ids));
 
   if (download_) {
-    download_->OnAllDataSaved(all_save_items_count_,
-                              DownloadItem::kEmptyFileHash);
+    if (save_type_ != content::SAVE_PAGE_TYPE_AS_MHTML)
+      download_->OnAllDataSaved(all_save_items_count_,
+                                DownloadItem::kEmptyFileHash);
     download_->MarkAsComplete();
     FinalizeDownloadEntry();
   }
 }
 
 // Called for updating end state.
 void SavePackage::SaveFinished(int32 save_id, int64 size, bool is_success) {
   // Because we might have canceled this saving job before,
   // so we might not find corresponding SaveItem. Just ignore it.
   SaveItem* save_item = LookupItemInProcessBySaveId(save_id);
@@ skipping 47 matching lines @@
 
   save_item->Finish(0, false);
 
   PutInProgressItemToSavedMap(save_item);
 
   // Inform the DownloadItem to update UI.
   // We use the received bytes as number of saved files.
   if (download_)
     download_->UpdateProgress(completed_count(), CurrentSpeed(), "");
 
-  if (save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML ||
-      save_item->save_source() == SaveFileCreateInfo::SAVE_FILE_FROM_DOM) {
+  if ((save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML) ||
+      (save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML) ||
+      (save_item->save_source() == SaveFileCreateInfo::SAVE_FILE_FROM_DOM)) {
     // We got error when saving page. Treat it as disk error.
     Cancel(true);
   }
 
   if (canceled()) {
     DCHECK(finished_);
     return;
   }
 
   // Continue processing the save page job.
@@ skipping 85 matching lines @@
         wait_state_ = HTML_DATA;
         // All non-HTML resources have been finished, start all remaining
         // HTML files.
         SaveNextFile(true);
       }
     } else if (in_process_count()) {
       // Continue asking for HTML data.
       DCHECK(wait_state_ == HTML_DATA);
     }
   } else {
-    // Save as HTML only.
+    // Save as HTML only or MHTML.
     DCHECK(wait_state_ == NET_FILES);
-    DCHECK(save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML);
+    DCHECK((save_type_ == content::SAVE_PAGE_TYPE_AS_ONLY_HTML) ||
+           (save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML));
     if (waiting_item_queue_.size()) {
       DCHECK(all_save_items_count_ == waiting_item_queue_.size());
       SaveNextFile(false);
     }
   }
 }
 
 bool SavePackage::OnMessageReceived(const IPC::Message& message) {
   bool handled = true;
   IPC_BEGIN_MESSAGE_MAP(SavePackage, message)
@@ skipping 363 matching lines @@
   // The WebContents which owns this SavePackage may have disappeared during
   // the UI->FILE->UI thread hop of
   // GetSaveInfo->CreateDirectoryOnFileThread->ContinueGetSaveInfo.
   if (!web_contents())
     return;
 
   FilePath::StringType default_extension;
   if (can_save_as_complete)
     default_extension = kDefaultHtmlExtension;
 
-  // On ChromeOS, OnPathPicked is not invoked; SavePackageFilePickerChromeOS
-  // handles the the save.
   download_manager_->delegate()->ChooseSavePath(
-      web_contents(), suggested_path, default_extension, can_save_as_complete,
+      web_contents(),
+      suggested_path,
+      default_extension,
+      can_save_as_complete,
       base::Bind(&SavePackage::OnPathPicked, AsWeakPtr()));
 }
 
-void SavePackage::OnPathPicked(const FilePath& final_name,
-                               content::SavePageType type) {
+void SavePackage::OnPathPicked(
+    const FilePath& final_name,
+    content::SavePageType type,
+    const content::SaveFileDownloadCreatedCallback& download_created_callback) {
   // Ensure the filename is safe.
   saved_main_file_path_ = final_name;
   // TODO(asanka): This call may block on IO and shouldn't be made
   // from the UI thread.  See http://crbug.com/61827.
   net::GenerateSafeFileName(web_contents()->GetContentsMimeType(), false,
                             &saved_main_file_path_);
 
   saved_main_directory_path_ = saved_main_file_path_.DirName();
   save_type_ = type;
   if (save_type_ == content::SAVE_PAGE_TYPE_AS_COMPLETE_HTML) {
     // Make new directory for saving complete file.
     saved_main_directory_path_ = saved_main_directory_path_.Append(
         saved_main_file_path_.RemoveExtension().BaseName().value() +
         FILE_PATH_LITERAL("_files"));
   }
 
-  Init();
+  Init(download_created_callback);
 }
 
 void SavePackage::StopObservation() {
   DCHECK(download_);
   DCHECK(download_manager_);
 
   download_->RemoveObserver(this);
   download_ = NULL;
   download_manager_ = NULL;
 }
 
 void SavePackage::OnDownloadUpdated(DownloadItem* download) {
   DCHECK(download_);
   DCHECK(download_ == download);
   DCHECK(download_manager_);
 
   // Check for removal.
-  if (download->GetState() == DownloadItem::REMOVING)
+  if (download_->GetState() == DownloadItem::REMOVING) {
     StopObservation();
+  }
+
+  // MHTML saves may need to wait for GData to finish uploading.
+  if ((save_type_ == content::SAVE_PAGE_TYPE_AS_MHTML) &&
+      download_->AllDataSaved() &&
+      !download_->IsComplete() &&
+      !mhtml_finishing_ &&
+      download_manager_->delegate()->ShouldCompleteDownload(download_)) {
+    // Post a task to avoid re-entering OnDownloadUpdated. Set a flag to
+    // prevent double-calling Finish() in case another OnDownloadUpdated happens
+    // before Finish() runs.
+    mhtml_finishing_ = true;
+    BrowserThread::PostTask(BrowserThread::UI, FROM_HERE,
+        base::Bind(&SavePackage::Finish, this));

[Randy Smith (Not in Mondays), 2012/04/27 18:24:04]

This makes me twitchy--a complicated conditional, a PostTask(), and a boolean to
handle accidental calling repeats makes me want to look for a better way.  I
*think* the suggestion I made elsewhere of passing a callback to
DMD::ShouldCompleteDownload() will render this unnecessary, as that callback can
just call SavePackage::Finish() directly.  If not, I'd like to come back here
(or for you to come back here, or for us to come back here) and see if there's
some other way to think about this problem that would lead to simpler code.

+  }
 }
 
 void SavePackage::FinalizeDownloadEntry() {
   DCHECK(download_);
   DCHECK(download_manager_);
 
   download_manager_->SavePageDownloadFinished(download_);
   StopObservation();
 }