The optional Malware Details also collects HTTP cache information. See design...

chrome/browser/safe_browsing/malware_details_cache.cc

Index: chrome/browser/safe_browsing/malware_details_cache.cc
===================================================================
--- chrome/browser/safe_browsing/malware_details_cache.cc	(revision 0)
+++ chrome/browser/safe_browsing/malware_details_cache.cc	(revision 0)
@@ -0,0 +1,330 @@
+// 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.
+//
+// Implementation of the MalwareDetails class.
+
+#include "chrome/browser/safe_browsing/malware_details.h"
+
+#include "base/callback.h"
+#include "base/lazy_instance.h"
+#include "base/md5.h"
+#include "chrome/browser/net/chrome_url_request_context.h"
+#include "chrome/browser/profiles/profile.h"
+#include "chrome/browser/safe_browsing/malware_details_cache.h"
+#include "chrome/browser/safe_browsing/safe_browsing_service.h"
+#include "chrome/browser/safe_browsing/report.pb.h"
+#include "content/browser/browser_thread.h"
+#include "net/base/io_buffer.h"
+#include "net/disk_cache/disk_cache.h"
+#include "net/http/http_cache.h"
+#include "net/http/http_response_headers.h"
+#include "net/http/http_response_info.h"
+#include "net/url_request/url_request_context_getter.h"
+
+using safe_browsing::ClientMalwareReportRequest;
+
+// Only send small files for now, a better strategy would use the size
+// of the whole report and the user's bandwidth.
+static const int kMaxBodySizeBytes = 1024;
+
+MalwareDetailsCacheCollector::MalwareDetailsCacheCollector()
+    : cache_state_(STATE_NONE),
+      ALLOW_THIS_IN_INITIALIZER_LIST(
+          cache_callback_(this , &MalwareDetailsCacheCollector::CacheLoop)),
+      ALLOW_THIS_IN_INITIALIZER_LIST(
+          entry_callback_(
+              new net::CancelableCompletionCallback<
+              MalwareDetailsCacheCollector>(
+                  this, &MalwareDetailsCacheCollector::CacheLoop))) {
+}
+
+MalwareDetailsCacheCollector::~MalwareDetailsCacheCollector() {
+}
+
+void MalwareDetailsCacheCollector::StartCacheCollection(
+    net::URLRequestContextGetter* request_context_getter,
+    safe_browsing::ResourceMap* resources,
+    int* result,
+    Task* callback) {
+  // Start the data collection from the HTTP cache.
+  DVLOG(1) << "Getting disk_cache data for all urls...";
+  request_context_getter_ = request_context_getter;
+  resources_ = resources;
+  resources_it_ = resources_->begin();
+  cache_state_ = STATE_OPEN_CACHE;
+  result_ = result;
+  callback_ = callback;
+
+  AddRef();  // The cache_callback_ calls are asynchronous and not refcounted.
+
+  // Post a task in the message loop, so the callers don't need to
+  // check if we call their callback immediately.
+  BrowserThread::PostTask(
+      BrowserThread::IO, FROM_HERE,
+      NewRunnableMethod(
+          this, &MalwareDetailsCacheCollector::CacheLoop, net::OK));
+}
+
+bool MalwareDetailsCacheCollector::InProgress() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  return cache_state_ != STATE_NONE;
+}
+
+// The IO thread loop. Called once the DOM details are callected,
+// and then is passed to all the cache lookups as a callback.
+void MalwareDetailsCacheCollector::CacheLoop(int result) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  DCHECK(cache_state_ != STATE_NONE);
+
+  int rv = result;
+  do {
+    DVLOG(1) << "cache_state_:" << cache_state_;
+    CacheState state = cache_state_;
+    cache_state_ = STATE_NONE;
+    switch (state) {
+      case STATE_OPEN_CACHE:
+        rv = DoOpenCache();
+        break;
+      case STATE_OPEN_CACHE_COMPLETE:
+        rv = DoOpenCacheComplete(rv);
+        break;
+      case STATE_OPEN_ENTRY:
+        rv = DoOpenEntry();
+        break;
+      case STATE_OPEN_ENTRY_COMPLETE:
+        rv = DoOpenEntryComplete(rv);
+        break;
+      case STATE_READ_RESPONSE:
+        rv = DoReadResponse();
+        break;
+      case STATE_READ_RESPONSE_COMPLETE:
+        rv = DoReadResponseComplete(rv);
+        break;
+      case STATE_READ_DATA:
+        rv = DoReadData();
+        break;
+      case STATE_READ_DATA_COMPLETE:
+        rv = DoReadDataComplete(rv);
+        break;
+      default:
+        NOTREACHED() << "Bad state:" << state;
+        rv = net::ERR_FAILED;
+        break;
+    }
+  } while (rv != net::ERR_IO_PENDING && cache_state_ != STATE_NONE);
+
+  if (rv != net::ERR_IO_PENDING) {
+    DVLOG(1) << "CacheIO done.";
+    DoCacheDone(rv);
+  }
+}
+
+net::URLRequestContext* MalwareDetailsCacheCollector::GetURLRequestContext() {
+  if (!request_context_getter_) {
+    DVLOG(1) << "Missing request context getter";
+    return NULL;
+  }
+  return request_context_getter_->GetURLRequestContext();
+}
+
+int MalwareDetailsCacheCollector::DoOpenCache() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_OPEN_CACHE_COMPLETE;
+
+  net::URLRequestContext* context = GetURLRequestContext();
+  if (!context) {
+    DVLOG(1) << "Could not get URLRequestContext";
+    cache_state_ = STATE_NONE;
+    return net::ERR_FAILED;
+  }
+
+  if (!context->http_transaction_factory()) {
+    DVLOG(1) << "Could not get HTTP transaction factory";
+    cache_state_ = STATE_NONE;
+    return net::ERR_FAILED;
+  }
+
+  net::HttpCache* http_cache =
+      context->http_transaction_factory()->GetCache();
+  if (!http_cache) {
+    DVLOG(1) << "Could not get http cache";
+    cache_state_ = STATE_NONE;
+    return net::ERR_FAILED;
+  }
+
+  return http_cache->GetBackend(&cache_, &cache_callback_);
+}
+
+int MalwareDetailsCacheCollector::DoOpenCacheComplete(int result) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_OPEN_ENTRY;
+
+  if (result == net::ERR_FAILED || !cache_) {
+    DVLOG(1) << "Could not get disk cache";
+    cache_state_ = STATE_NONE;
+    return net::ERR_FAILED;
+  }
+
+  return net::OK;
+}
+
+int MalwareDetailsCacheCollector::DoOpenEntry() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_OPEN_ENTRY_COMPLETE;
+
+  if (resources_it_ == resources_->end()) {
+    cache_state_ = STATE_NONE;
+    return net::OK;
+  }
+
+  entry_ = NULL;  // Reset the entry.
+  // resources_it_ is an iterator to a map, key = URL, value =
+  // ClientMalwareReportRequest::Resource.
+  return cache_->OpenEntry(resources_it_->first, &entry_, &cache_callback_);
+}
+
+int MalwareDetailsCacheCollector::DoOpenEntryComplete(int rv) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  if (rv != net::OK || !entry_) {
+    DVLOG(1) << "Error opening entry, or entry not found: Advancing.";
+    return AdvanceEntry();
+  }
+
+  cache_state_ = STATE_READ_RESPONSE;
+  return net::OK;
+}
+
+int MalwareDetailsCacheCollector::DoReadResponse() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_READ_RESPONSE_COMPLETE;
+
+  // net::HttpCache::kResponseInfoIndex = 0.

[mattm, 2011/04/12 01:47:26]

Do you know if there is a reason why the enum where this is defined isn't
public?

[panayiotis, 2011/04/12 17:23:10]

I don't know :/

+  buf_len_ = entry_->GetDataSize(0);
+  if (!buf_len_) {
+    DVLOG(1) << "Empty response";
+    return AdvanceEntry();
+  }
+
+  buf_ = new net::IOBuffer(buf_len_);
+  return entry_->ReadData(0, 0, buf_, buf_len_, entry_callback_);
+}
+
+// Here, |result| is the length of the buffer read.
+int MalwareDetailsCacheCollector::DoReadResponseComplete(int result) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_READ_DATA;
+
+  if (!result || result != buf_len_) {
+    DVLOG(1) << "Empty buffer or buffer length mismatch.";
+    // Reset buf_ and buf_len_.
+    buf_ = NULL;
+    buf_len_ = 0;
+    return AdvanceEntry();
+  }
+
+  linked_ptr<ClientMalwareReportRequest::Resource>
+      pb_resource = resources_it_->second;
+  net::HttpResponseInfo response;
+  bool truncated;
+  if (net::HttpCache::ParseResponseInfo(buf_->data(), buf_len_, &response,
+                                        &truncated)) {
+    if (!truncated) {
+      if (response.headers) {
+        ClientMalwareReportRequest::HTTPResponse* pb_response =
+            pb_resource->mutable_response();
+        pb_response->mutable_firstline()->set_code(
+            response.headers->response_code());
+        void* iter = NULL;
+        std::string name, value;
+        while (response.headers->EnumerateHeaderLines(&iter, &name, &value)) {
+          ClientMalwareReportRequest::HTTPHeader* pb_header =
+              pb_response->add_headers();
+          pb_header->set_name(name);
+          // Strip any Set-Cookie headers.

[mattm, 2011/04/12 01:47:26]

Is this for privacy reasons?  Seems like it should be mentioned in the design
doc.

[panayiotis, 2011/04/12 17:23:10]

Update the doc. 

+          if (LowerCaseEqualsASCII(name, "set-cookie")) {
+            pb_header->set_value("");
+          } else {
+            pb_header->set_value(value);
+          }
+        }
+      } else {
+        DVLOG(1) << "Missing response headers.";
+      }
+    } else {
+      DVLOG(1) << "Response truncated";
+    }
+  } else {
+    DVLOG(1) << "Could not parse HTTP cache entry";
+  }
+
+  // Reset buf_ and buf_len_.
+  buf_ = NULL;
+  buf_len_ = 0;
+  return net::OK;
+}
+
+int MalwareDetailsCacheCollector::DoReadData() {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  cache_state_ = STATE_READ_DATA_COMPLETE;
+
+  // net::HttpCache::kResponseContentIndex = 1.
+  buf_len_ = entry_->GetDataSize(1);
+  if (!buf_len_) {
+    DVLOG(1) << "Empty data";
+    return AdvanceEntry();
+  }
+
+  buf_ = new net::IOBuffer(buf_len_);
+  return entry_->ReadData(1, 0, buf_, buf_len_, entry_callback_);

[mattm, 2011/04/12 01:47:26]

I don't quite grok the disk cache interface, but is it ok that we don't check
for / do something with sparse entries?

[panayiotis, 2011/04/12 17:23:10]

Do you mean that the entries we lookup could be spread in sparse locations on
the disk and we could cause too many seeks? 

I think since the majority of reports have <=3 entries, it shouldn't be a
problem. In addition, most of these URLs should have been added to the cache
very recently -- they are all part of the page that was loaded. So maybe there
cache is smart enough to keep these entries together ... 

[mattm, 2011/04/12 22:18:12]

I was refering to ReadSparseData in disk_cache.h.  It has a comment saying
trying to read sparse entries with the non-sparse function gives undefined
results, but I dunno if any of the cache entries this will get are sparse.

+}
+
+// Here, |result| is the length of the buffer read.
+int MalwareDetailsCacheCollector::DoReadDataComplete(int result) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+
+  if (result && result == buf_len_) {
+    linked_ptr<ClientMalwareReportRequest::Resource>
+        pb_resource = resources_it_->second;
+    ClientMalwareReportRequest::HTTPResponse* pb_response =
+        pb_resource->mutable_response();
+    if (buf_len_ <= kMaxBodySizeBytes) {  // Only send small bodies for now.
+      pb_response->set_body(buf_->data(), buf_len_);
+    }
+    pb_response->set_bodylength(buf_len_);
+    MD5Digest digest;
+    MD5Sum(buf_->data(), buf_len_, &digest);
+    pb_response->set_bodydigest(MD5DigestToBase16(digest));
+  }
+
+  // Reset buf_ and buf_len.
+  buf_ = NULL;
+  buf_len_ = 0;
+  return AdvanceEntry();
+}
+
+int MalwareDetailsCacheCollector::AdvanceEntry() {
+  // Close this entry, advance to the next one. Create a task so we don't
+  // take over the IO thread for too long.
+  cache_state_ = STATE_OPEN_ENTRY;
+  ++resources_it_;
+  if (entry_) {
+    entry_->Close();
+    entry_ = NULL;
+  }
+  BrowserThread::PostTask(
+      BrowserThread::IO, FROM_HERE,
+      NewRunnableMethod(
+          this, &MalwareDetailsCacheCollector::CacheLoop, net::OK));
+
+  return net::ERR_IO_PENDING;
+}
+
+void MalwareDetailsCacheCollector::DoCacheDone(int result) {
+  DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
+  *result_ = result;
+  BrowserThread::PostTask(BrowserThread::IO, FROM_HERE, callback_);
+
+  Release();
+  // We are now deleted.
+}
Property changes on: chrome/browser/safe_browsing/malware_details_cache.cc
___________________________________________________________________
Added: svn:eol-style
   + LF