Add read-only access to POST data for webRequest's onBeforeRequest

chrome/browser/extensions/api/web_request/web_request_api.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 "chrome/browser/extensions/api/web_request/web_request_api.h"
 
 #include <algorithm>
 
 #include "base/bind.h"
 #include "base/json/json_writer.h"
 #include "base/metrics/histogram.h"
 #include "base/string_number_conversions.h"
 #include "base/string_util.h"
 #include "base/time.h"
 #include "base/utf_string_conversions.h"
 #include "base/values.h"
 #include "chrome/browser/browser_process.h"
 #include "chrome/browser/chrome_content_browser_client.h"
 #include "chrome/browser/extensions/api/declarative_webrequest/webrequest_rule.h"
 #include "chrome/browser/extensions/api/declarative_webrequest/webrequest_rules_registry.h"
 #include "chrome/browser/extensions/api/web_navigation/web_navigation_api_helpers.h"
+#include "chrome/browser/extensions/api/web_request/upload_data_presenter.h"
 #include "chrome/browser/extensions/api/web_request/web_request_api_constants.h"
 #include "chrome/browser/extensions/api/web_request/web_request_api_helpers.h"
 #include "chrome/browser/extensions/api/web_request/web_request_time_tracker.h"
 #include "chrome/browser/extensions/event_router.h"
 #include "chrome/browser/extensions/extension_info_map.h"
 #include "chrome/browser/extensions/extension_prefs.h"
 #include "chrome/browser/extensions/extension_service.h"
 #include "chrome/browser/extensions/extension_tab_id_map.h"
 #include "chrome/browser/profiles/profile.h"
 #include "chrome/browser/profiles/profile_manager.h"
 #include "chrome/browser/renderer_host/chrome_render_message_filter.h"
 #include "chrome/common/extensions/api/web_request.h"
 #include "chrome/common/extensions/event_filtering_info.h"
 #include "chrome/common/extensions/extension.h"
 #include "chrome/common/extensions/extension_constants.h"
 #include "chrome/common/extensions/extension_error_utils.h"
 #include "chrome/common/extensions/extension_messages.h"
+#include "chrome/common/extensions/features/feature.h"
 #include "chrome/common/extensions/url_pattern.h"
 #include "chrome/common/url_constants.h"
 #include "content/public/browser/browser_message_filter.h"
 #include "content/public/browser/browser_thread.h"
 #include "content/public/browser/render_process_host.h"
 #include "content/public/browser/resource_request_info.h"
 #include "googleurl/src/gurl.h"
 #include "grit/generated_resources.h"
 #include "net/base/auth.h"
 #include "net/base/net_errors.h"
+#include "net/base/upload_data.h"
+#include "net/base/upload_element.h"
 #include "net/http/http_response_headers.h"
 #include "net/url_request/url_request.h"
 #include "ui/base/l10n/l10n_util.h"
 
+using base::DictionaryValue;
+using base::ListValue;
+using base::StringValue;
+using chrome::VersionInfo;
 using content::BrowserMessageFilter;
 using content::BrowserThread;
 using content::ResourceRequestInfo;
 using extensions::Extension;
+using extensions::Feature;
 
 using extensions::web_navigation_api_helpers::GetFrameId;
 
 namespace helpers = extension_web_request_api_helpers;
 namespace keys = extension_web_request_api_constants;
 namespace web_request = extensions::api::web_request;
 
 namespace {
 
 // List of all the webRequest events.
-static const char* const kWebRequestEvents[] = {
+const char* const kWebRequestEvents[] = {
   keys::kOnBeforeRedirect,
   keys::kOnBeforeRequest,
   keys::kOnBeforeSendHeaders,
   keys::kOnCompleted,
   keys::kOnErrorOccurred,
   keys::kOnSendHeaders,
   keys::kOnAuthRequired,
   keys::kOnResponseStarted,
   keys::kOnHeadersReceived,
 };
 
 #define ARRAYEND(array) (array + arraysize(array))
 
+// Access to request body (crbug.com/91191/) is currently only enabled in dev
+// and canary channels. This function caches the release channel info from
+// chrome::VersionInfo. Don't use this outside IO thread.

[battre, 2012/09/09 22:08:35]

where is this caching? I don't understand why we cannot use it outside the IO
thread.

[vabr (Chromium), 2012/09/12 11:18:08]

Thanks for catching that. The comment was obsolete, from before I switched to
using Feature. I removed the no longer applicable part.

+bool IsWebRequestBodyDataAccessEnabled() {
+  return Feature::GetCurrentChannel() <= VersionInfo::CHANNEL_DEV;
+}
+
 bool IsWebRequestEvent(const std::string& event_name) {
   return std::find(kWebRequestEvents, ARRAYEND(kWebRequestEvents),
                    event_name) != ARRAYEND(kWebRequestEvents);
 }
 
 // Returns whether |request| has been triggered by an extension in
 // |extension_info_map|.
 bool IsRequestFromExtension(const net::URLRequest* request,
                             const ExtensionInfoMap* extension_info_map) {
   // |extension_info_map| is NULL for system-level requests.
@@ skipping 60 matching lines @@
                  base::Uint64ToString(request->identifier()));
   out->SetString(keys::kUrlKey, request->url().spec());
   out->SetString(keys::kMethodKey, request->method());
   out->SetInteger(keys::kFrameIdKey, frame_id_for_extension);
   out->SetInteger(keys::kParentFrameIdKey, parent_frame_id_for_extension);
   out->SetInteger(keys::kTabIdKey, tab_id);
   out->SetString(keys::kTypeKey, helpers::ResourceTypeToString(resource_type));
   out->SetDouble(keys::kTimeStampKey, base::Time::Now().ToDoubleT() * 1000);
 }
 
+// Extracts the body from |request| and writes the data into |out|.
+void ExtractRequestInfoBody(const net::URLRequest* request,
+                            DictionaryValue* out) {
+  if (request->method() != "POST" && request->method() != "PUT")
+    return;  // Need to exit without "out->Set(keys::kRequestBodyKey, ...);" .
+
+  DictionaryValue* requestBody = new DictionaryValue();
+  out->Set(keys::kRequestBodyKey, requestBody);
+
+  // Get the data presenters, ordered by being interesting.

[battre, 2012/09/09 22:08:35]

"ordered by being interesting"? grammar and what does that mean? by specificity?

[vabr (Chromium), 2012/09/12 11:18:08]

I refined that comment, and also corrected the two in-line comments inside the
definition of presenters[].

+  extensions::RawDataPresenter raw_data_presenter;
+  extensions::ParsedDataPresenter parsed_data_presenter(request);

[battre, 2012/09/09 22:08:35]

nit: swap previous two lines?

[vabr (Chromium), 2012/09/12 11:18:08]

Done.

+  extensions::UploadDataPresenter* const presenters[] = {
+    &parsed_data_presenter,    // Second -- any parseable forms?
+    &raw_data_presenter        // Third  -- any data at all?
+  };
+  // Keys for the results of the corresponding presenters.
+  static const char* const kKeys[] = {
+    keys::kRequestBodyFormDataKey,
+    keys::kRequestBodyRawKey
+  };
+
+  const std::vector<net::UploadElement>* elements =
+      request->get_upload()->elements();
+  bool some_succeeded = false;
+  for (size_t i = 0; !some_succeeded && i < arraysize(presenters); ++i) {
+    std::vector<net::UploadElement>::const_iterator element;
+    for (element = elements->begin(); element != elements->end(); ++element)
+      presenters[i]->FeedNext(*element);
+    if (presenters[i]->Succeeded()) {
+      requestBody->Set(kKeys[i], presenters[i]->Result().release());
+      some_succeeded = true;
+    }
+  }
+  if (!some_succeeded)
+    requestBody->SetString(keys::kRequestBodyErrorKey, "Unknown error.");
+}
+
 // Converts a HttpHeaders dictionary to a |name|, |value| pair. Returns
 // true if successful.
 bool FromHeaderDictionary(const DictionaryValue* header_value,
                           std::string* name,
                           std::string* value) {
   if (!header_value->GetString(keys::kHeaderNameKey, name))
     return false;
 
   // We require either a "value" or a "binaryValue" entry.
   if (!(header_value->HasKey(keys::kHeaderValueKey) ^
@@ skipping 228 matching lines @@
       return false;
 
     if (str == "requestHeaders")
       *extra_info_spec |= REQUEST_HEADERS;
     else if (str == "responseHeaders")
       *extra_info_spec |= RESPONSE_HEADERS;
     else if (str == "blocking")
       *extra_info_spec |= BLOCKING;
     else if (str == "asyncBlocking")
       *extra_info_spec |= ASYNC_BLOCKING;
+    else if (str == "requestBody")
+      *extra_info_spec |=
+          IsWebRequestBodyDataAccessEnabled() ? REQUEST_BODY : 0;
     else
       return false;
 
     // BLOCKING and ASYNC_BLOCKING are mutually exclusive.
     if ((*extra_info_spec & BLOCKING) && (*extra_info_spec & ASYNC_BLOCKING))
       return false;
   }
   return true;
 }
 
@@ skipping 70 matching lines @@
 
   int extra_info_spec = 0;
   std::vector<const EventListener*> listeners =
       GetMatchingListeners(profile, extension_info_map, keys::kOnBeforeRequest,
                            request, &extra_info_spec);
   if (!listeners.empty() &&
       !GetAndSetSignaled(request->identifier(), kOnBeforeRequest)) {
     ListValue args;
     DictionaryValue* dict = new DictionaryValue();
     ExtractRequestInfo(request, dict);
+    if (extra_info_spec & ExtraInfoSpec::REQUEST_BODY)
+      ExtractRequestInfoBody(request, dict);
     args.Append(dict);
 
     initialize_blocked_requests |=
         DispatchEvent(profile, request, listeners, args);
   }
 
   if (!initialize_blocked_requests)
     return net::OK;  // Nobody saw a reason for modifying the request.
 
   blocked_requests_[request->identifier()].event = kOnBeforeRequest;
@@ skipping 1331 matching lines @@
       } else if ((*it)->name().find("AdBlock") != std::string::npos) {
         adblock = true;
       } else {
         other = true;
       }
     }
   }
 
   host->Send(new ExtensionMsg_UsingWebRequestAPI(adblock, adblock_plus, other));
 }