Get signals working in the EXTRA_DATA section of MHTML

content/browser/download/mhtml_generation_manager.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/mhtml_generation_manager.h"
 
 #include <map>
 #include <queue>
 #include <utility>
 
 #include "base/bind.h"
 #include "base/files/file.h"
 #include "base/guid.h"
 #include "base/macros.h"
 #include "base/memory/ptr_util.h"
 #include "base/metrics/histogram_macros.h"
 #include "base/scoped_observer.h"
 #include "base/stl_util.h"
 #include "base/strings/string_util.h"
 #include "base/strings/stringprintf.h"
 #include "base/time/time.h"
 #include "base/trace_event/trace_event.h"
 #include "content/browser/bad_message.h"
+#include "content/browser/download/mhtml_extra_data.h"
 #include "content/browser/frame_host/frame_tree_node.h"
 #include "content/browser/frame_host/render_frame_host_impl.h"
 #include "content/common/frame_messages.h"
 #include "content/public/browser/browser_thread.h"
 #include "content/public/browser/render_frame_host.h"
 #include "content/public/browser/render_process_host.h"
 #include "content/public/browser/render_process_host_observer.h"
 #include "content/public/browser/web_contents.h"
 #include "content/public/common/mhtml_generation_params.h"
 #include "net/base/mime_util.h"
 
+namespace {
+const char kContentLocation[] = "Content-Location: ";
+}  // namespace
+
 namespace content {
 
 // The class and all of its members live on the UI thread.  Only static methods
 // are executed on other threads.
 class MHTMLGenerationManager::Job : public RenderProcessHostObserver {
  public:
   Job(int job_id,
       WebContents* web_contents,
       const MHTMLGenerationParams& params,
       const GenerateMHTMLCallback& callback);
@@ skipping 49 matching lines @@
 
  private:
   // Writes the MHTML footer to the file and closes it.
   //
   // Note: The same |boundary| marker must be used for all "boundaries" -- in
   // the header, parts and footer -- that belong to the same MHTML document (see
   // also rfc1341, section 7.2.1, "boundary" description).
   static std::tuple<MhtmlSaveStatus, int64_t> CloseFileOnFileThread(
       MhtmlSaveStatus save_status,
       const std::string& boundary,
-      base::File file);
+      base::File file,
+      const std::string& extra_data);
   void AddFrame(RenderFrameHost* render_frame_host);
 
+  // If we have any extra data sections to write out, write them into the file
+  // while on the file thread.  Returns true for success.
+  static bool WriteExtraDataSectionOnFileThread(const std::string& boundary,
+                                                base::File& file,
+                                                const std::string& extra_data);
+
   // Creates a new map with values (content ids) the same as in
   // |frame_tree_node_to_content_id_| map, but with the keys translated from
   // frame_tree_node_id into a |site_instance|-specific routing_id.
   std::map<int, std::string> CreateFrameRoutingIdToContentId(
       SiteInstance* site_instance);
 
   // Id used to map renderer responses to jobs.
   // See also MHTMLGenerationManager::id_to_job_ map.
   const int job_id_;
 
@@ skipping 32 matching lines @@
   std::string salt_;
 
   // The callback to call once generation is complete.
   const GenerateMHTMLCallback callback_;
 
   // Whether the job is finished (set to true only for the short duration of
   // time between MHTMLGenerationManager::JobFinished is called and the job is
   // destroyed by MHTMLGenerationManager::OnFileClosed).
   bool is_finished_;
 
+  // Any extra data section that should be emitted into the output MHTML.
+  std::string extra_data_;
+
   // RAII helper for registering this Job as a RenderProcessHost observer.
   ScopedObserver<RenderProcessHost, MHTMLGenerationManager::Job>
       observed_renderer_process_host_;
 
   DISALLOW_COPY_AND_ASSIGN(Job);
 };
 
 MHTMLGenerationManager::Job::Job(int job_id,
                                  WebContents* web_contents,
                                  const MHTMLGenerationParams& params,
                                  const GenerateMHTMLCallback& callback)
     : job_id_(job_id),
       creation_time_(base::TimeTicks::Now()),
       params_(params),
       frame_tree_node_id_of_busy_frame_(FrameTreeNode::kFrameTreeNodeInvalidId),
       mhtml_boundary_marker_(net::GenerateMimeMultipartBoundary()),
       salt_(base::GenerateGUID()),
       callback_(callback),
       is_finished_(false),
       observed_renderer_process_host_(this) {
   DCHECK_CURRENTLY_ON(BrowserThread::UI);
   web_contents->ForEachFrame(base::Bind(
       &MHTMLGenerationManager::Job::AddFrame,
       base::Unretained(this)));  // Safe because ForEachFrame is synchronous.
 
   // Main frame needs to be processed first.
   DCHECK(!pending_frame_tree_node_ids_.empty());
   DCHECK(FrameTreeNode::GloballyFindByID(pending_frame_tree_node_ids_.front())
              ->parent() == nullptr);
+
+  // Save off any extra data
+  base::SupportsUserData::Data* found_user_data =
+      web_contents->GetUserData(content::kMHTMLExtraDataKey);

[Dmitry Titov, 2017/03/16 02:28:21]

It is better to follow other examples that generally look like this:

static Foo* Foo::FromWebContents(WebContents*) {}

See, for example DownloadUIAdapter or other classes that define FromWebContents
static. this allows to keep kFooKey in foo.cc file...

[fgorski, 2017/03/16 17:46:32]

+1. Details in my earlier comment.

[Pete Williamson, 2017/03/17 23:28:32]

Done.

[Pete Williamson, 2017/03/17 23:28:32]

Done.

+  MHTMLExtraData* found_data = static_cast<MHTMLExtraData*>(found_user_data);
+
+  if (found_data != nullptr)
+    extra_data_ = found_data->data_string;
 }
 
 MHTMLGenerationManager::Job::~Job() {
   DCHECK_CURRENTLY_ON(BrowserThread::UI);
 }
 
 std::map<int, std::string>
 MHTMLGenerationManager::Job::CreateFrameRoutingIdToContentId(
     SiteInstance* site_instance) {
   std::map<int, std::string> result;
@@ skipping 150 matching lines @@
     return;
   }
 
   // If no previous error occurred the boundary should be sent.
   BrowserThread::PostTaskAndReplyWithResult(
       BrowserThread::FILE, FROM_HERE,
       base::Bind(
           &MHTMLGenerationManager::Job::CloseFileOnFileThread, save_status,
           (save_status == MhtmlSaveStatus::SUCCESS ? mhtml_boundary_marker_
                                                    : std::string()),
-          base::Passed(&browser_file_)),
+          base::Passed(&browser_file_), extra_data_),
       callback);
 }
 
 bool MHTMLGenerationManager::Job::IsMessageFromFrameExpected(
     RenderFrameHostImpl* sender) {
   int sender_id = sender->frame_tree_node()->frame_tree_node_id();
   if (sender_id != frame_tree_node_id_of_busy_frame_)
     return false;
 
   // We only expect one message per frame - let's make sure subsequent messages
@@ skipping 25 matching lines @@
 
   // Report success if all frames have been processed.
   if (pending_frame_tree_node_ids_.empty())
     return MhtmlSaveStatus::SUCCESS;
 
   return SendToNextRenderFrame();
 }
 
 // static
 std::tuple<MhtmlSaveStatus, int64_t>
-MHTMLGenerationManager::Job::CloseFileOnFileThread(MhtmlSaveStatus save_status,
-                                                   const std::string& boundary,
-                                                   base::File file) {
+MHTMLGenerationManager::Job::CloseFileOnFileThread(
+    MhtmlSaveStatus save_status,
+    const std::string& boundary,
+    base::File file,
+    const std::string& extra_data) {
   DCHECK_CURRENTLY_ON(BrowserThread::FILE);
 
   // If no previous error occurred the boundary should have been provided.
   if (save_status == MhtmlSaveStatus::SUCCESS) {
     TRACE_EVENT0("page-serialization",
                  "MHTMLGenerationManager::Job MHTML footer writing");
     DCHECK(!boundary.empty());
+
+    // Write the extra data into a section of its own, if we have any.
+    if (!WriteExtraDataSectionOnFileThread(boundary, file, extra_data))
+      save_status = MhtmlSaveStatus::FILE_WRITTING_ERROR;

[Dmitry Titov, 2017/03/16 02:28:21]

I think once there is a FILE_WRITING_ERROR, there should be no further writes to
the file... likely early return here?

[Pete Williamson, 2017/03/17 23:28:32]

Done.

+
     std::string footer = base::StringPrintf("--%s--\r\n", boundary.c_str());
     DCHECK(base::IsStringASCII(footer));
     if (file.WriteAtCurrentPos(footer.data(), footer.size()) < 0)
       save_status = MhtmlSaveStatus::FILE_WRITTING_ERROR;
   }
 
   // If the file is still valid try to close it. Only update the status if that
   // won't hide an earlier error.
   int64_t file_size = -1;
   if (file.IsValid()) {
     file_size = file.GetLength();
     file.Close();
   } else if (save_status == MhtmlSaveStatus::SUCCESS) {
     save_status = MhtmlSaveStatus::FILE_CLOSING_ERROR;
   }
 
   return std::make_tuple(save_status, file_size);
 }
 
+// static
+bool MHTMLGenerationManager::Job::WriteExtraDataSectionOnFileThread(
+    const std::string& boundary,
+    base::File& file,
+    const std::string& extra_data) {
+  // Don't write an extra data section if there is none.
+  if (extra_data.empty())
+    return true;
+
+  // TODO: What value do I use for content location?
+  std::string content_location_value;

[Dmitry Titov, 2017/03/16 02:28:20]

line 345 in this file is an example of how we generate CIDs in MHTML. This
results in something like
cid:frame-62691-645341c4-62b3-478e-a8c5-e0dfccc3ca02@mhtml.blink

I think we can continue this with something like 
cid:extra-data-[GUID]@mhtml.blink

[Pete Williamson, 2017/03/17 23:28:33]

Done.

+
+  // Write a newline, then a boundary, then the content location, another
+  // newline, the extra data string, and end with a newline.
+  std::string extra_data_section = base::StringPrintf(
+      "--%s--\r\n%s%s\r\n%s\r\n", boundary.c_str(), kContentLocation,
+      content_location_value.c_str(), extra_data.c_str());
+  // Now actually write the string into the file.
+  DCHECK(base::IsStringASCII(extra_data_section));
+  if (file.WriteAtCurrentPos(extra_data_section.data(),

[fgorski, 2017/03/16 17:46:32]

return file.Write... >= 0;

[Pete Williamson, 2017/03/17 23:28:32]

Done.

+                             extra_data_section.size()) < 0)

[fgorski, 2017/03/16 17:46:32]

nit: {} if you don't apply above.

[Pete Williamson, 2017/03/17 23:28:32]

Acknowledged.

+    return false;
+
+  return true;
+}
+
 MHTMLGenerationManager* MHTMLGenerationManager::GetInstance() {
   return base::Singleton<MHTMLGenerationManager>::get();
 }
 
 MHTMLGenerationManager::MHTMLGenerationManager() : next_job_id_(0) {}
 
 MHTMLGenerationManager::~MHTMLGenerationManager() {
 }
 
 void MHTMLGenerationManager::SaveMHTML(WebContents* web_contents,
@@ skipping 149 matching lines @@
   return iter->second.get();
 }
 
 void MHTMLGenerationManager::RenderProcessExited(Job* job) {
   DCHECK_CURRENTLY_ON(BrowserThread::UI);
   DCHECK(job);
   JobFinished(job, MhtmlSaveStatus::RENDER_PROCESS_EXITED);
 }
 
 }  // namespace content