Only track http and https in duplicate resource handler metrics

content/browser/renderer_host/duplicate_content_resource_handler.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/renderer_host/duplicate_content_resource_handler.h"
 
 #include <set>
 
 #include "base/lazy_instance.h"
 #include "base/logging.h"
@@ skipping 33 matching lines @@
 DuplicateContentResourceHandler::DuplicateContentResourceHandler(
     scoped_ptr<ResourceHandler> next_handler,
     ResourceType::Type resource_type,
     net::URLRequest* request)
     : LayeredResourceHandler(next_handler.Pass()),
       resource_type_(resource_type),
       bytes_read_(0),
       request_(request),
       pmurhash_ph1_(0),
       pmurhash_pcarry_(0) {
+  // Ignore everything that's not http/https. Specifically, exclude data and
+  // blob URLs which can be generated by content and cause the maintained sets
+  // to grow without bounds.
+  const GURL& url = request_->url();
+  track_request_ = url.SchemeIs("http") || url.SchemeIs("https");
 }
 
 DuplicateContentResourceHandler::~DuplicateContentResourceHandler() {
 }
 
 bool DuplicateContentResourceHandler::OnWillRead(int request_id,
                                                  net::IOBuffer** buf,
                                                  int* buf_size,
                                                  int min_size) {
   DCHECK_EQ(-1, min_size);
 
   if (!next_handler_->OnWillRead(request_id, buf, buf_size, min_size))
     return false;
   read_buffer_ = *buf;
   return true;
 }
 
 bool DuplicateContentResourceHandler::OnReadCompleted(int request_id,
                                                       int bytes_read,
                                                       bool* defer) {
-  PMurHash32_Process(&pmurhash_ph1_, &pmurhash_pcarry_,
-                      read_buffer_->data(), bytes_read);
-  bytes_read_ += bytes_read;
+  if (track_request_) {
+    PMurHash32_Process(&pmurhash_ph1_, &pmurhash_pcarry_,
+                       read_buffer_->data(), bytes_read);
+    bytes_read_ += bytes_read;
+  }
   return next_handler_->OnReadCompleted(request_id, bytes_read, defer);
 }
 
 bool DuplicateContentResourceHandler::OnResponseCompleted(
     int request_id,
     const net::URLRequestStatus& status,
     const std::string& security_info) {
-
-  if (status.is_success())
-    RecordContentMetrics();
+  if (track_request_) {
+    if (status.is_success())
+      RecordContentMetrics();
+  }
 
   return next_handler_->OnResponseCompleted(request_id, status, security_info);
 }
 
 void DuplicateContentResourceHandler::RecordContentMetrics() {
   MH_UINT32 contents_hash = PMurHash32_Result(pmurhash_ph1_,
                                               pmurhash_pcarry_, bytes_read_);
 
   bool is_http_or_https = request_->url().SchemeIs("http") ||
                           request_->url().SchemeIs("https");
   UMA_HISTOGRAM_BOOLEAN("Duplicate.IsHttpOrHttps", is_http_or_https);
 
   // Combine the contents_hash with the url, so we can test if future content
   // identical resources have the same original url or not.
-  MH_UINT32 hashed_with_url;
   const std::string& url_spec = request_->url().spec();
   PMurHash32_Process(&pmurhash_ph1_, &pmurhash_pcarry_,
                      url_spec.data(), url_spec.length());
-  hashed_with_url = PMurHash32_Result(pmurhash_ph1_, pmurhash_pcarry_,
-                                      url_spec.length() + bytes_read_);
+  MH_UINT32 hashed_with_url = PMurHash32_Result(
+      pmurhash_ph1_, pmurhash_pcarry_, url_spec.length() + bytes_read_);
 
   DVLOG(4) << "url: " << url_spec;
   DVLOG(4) << "contents hash: " << contents_hash;
   DVLOG(4) << "hash with url: " << hashed_with_url;
 
   std::set<MH_UINT32>* content_matches =
       GlobalDuplicateRecords::GetInstance()->content_matches();
   std::set<MH_UINT32>* content_and_url_matches =
       GlobalDuplicateRecords::GetInstance()->content_and_url_matches();
 
@@ skipping 14 matching lines @@
                                resource_type_, ResourceType::LAST_TYPE);
   }
   content_matches->insert(contents_hash);
   content_and_url_matches->insert(hashed_with_url);
 
   bytes_read_ = 0;
   read_buffer_ = NULL;
 }
 
 }  // namespace content