Add pause/resume functionality to precache

components/precache/core/precache_fetcher.h

 // Copyright 2013 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.
 
 #ifndef COMPONENTS_PRECACHE_CORE_PRECACHE_FETCHER_H_
 #define COMPONENTS_PRECACHE_CORE_PRECACHE_FETCHER_H_
 
 #include <stdint.h>
 
 #include <algorithm>
 #include <list>
 #include <memory>
 #include <string>
 #include <vector>
 
 #include "base/callback.h"
 #include "base/macros.h"
-#include "base/memory/ref_counted.h"
+#include "base/memory/weak_ptr.h"
+#include "base/single_thread_task_runner.h"
 #include "base/time/time.h"
 #include "components/precache/core/fetcher_pool.h"
 #include "net/url_request/url_fetcher.h"
 #include "net/url_request/url_fetcher_delegate.h"
 #include "url/gurl.h"
 
+namespace base {
+class SingleThreadTaskRunner;
+}
+
 namespace net {
 class URLRequestContextGetter;
 }
 
 namespace precache {
 
 class PrecacheConfigurationSettings;
+class PrecacheUnfinishedWork;
 
 // Visible for testing.
 extern const int kNoTracking;
 
 // Public interface to code that fetches resources that the user is likely to
 // want to fetch in the future, putting them in the network stack disk cache.
 // Precaching is intended to be done when Chrome is not actively in use, likely
 // hours ahead of the time when the resources are actually needed.
 //
 // This class takes as input a prioritized list of URL domains that the user
@@ skipping 10 matching lines @@
 // manifests are fetched from. These can be set by using command line switches
 // or by providing default values.
 //
 // Sample interaction:
 //
 // class MyPrecacheFetcherDelegate : public PrecacheFetcher::PrecacheDelegate {
 //  public:
 //   void PrecacheResourcesForTopURLs(
 //       net::URLRequestContextGetter* request_context,
 //       const std::list<GURL>& top_urls) {
-//     fetcher_.reset(new PrecacheFetcher(request_context, top_urls, this));
+//     fetcher_.reset(new PrecacheFetcher(...));
 //     fetcher_->Start();
 //   }
 //
+//   void Cancel() {
+//     std::unique_ptr<PrecacheUnfinishedWork> unfinished_work =
+//         fetcher_->GetUnfinishedWork();
+//     fetcher_.reset();
+//   }
+//
 //   virtual void OnDone() {
 //     // Do something when precaching is done.
 //   }
 //
 //  private:
 //   std::unique_ptr<PrecacheFetcher> fetcher_;
 // };
-class PrecacheFetcher {
+class PrecacheFetcher : public base::SupportsWeakPtr<PrecacheFetcher> {
  public:
   class PrecacheDelegate {
    public:
     // Called when the fetching of resources has finished, whether the resources
     // were fetched or not. If the PrecacheFetcher is destroyed before OnDone is
     // called, then precaching will be canceled and OnDone will not be called.
     virtual void OnDone() = 0;
+
   };
 
   // Visible for testing.
   class Fetcher;
 
   // Constructs a new PrecacheFetcher. The |starting_hosts| parameter is a
   // prioritized list of hosts that the user commonly visits. These hosts are
   // used by a server side component to construct a list of resource URLs that
   // the user is likely to fetch.
-  PrecacheFetcher(const std::vector<std::string>& starting_hosts,
-                  net::URLRequestContextGetter* request_context,
-                  const GURL& config_url,
-                  const std::string& manifest_url_prefix,
-                  PrecacheDelegate* precache_delegate);
+  PrecacheFetcher(

[sclittle, 2016/05/19 01:59:46]

Could this be separated into two constructors, with one taking just an
unfinished_work and the other taking the stuff to build one?

Alternatively, you could just put all the |config_url| and stuff into the
unfinished work itself, and just have one constructor with the
|unfinished_work|.

[bengr, 2016/05/19 23:19:31]

Done.

+      const std::vector<std::string>& starting_hosts,
+      net::URLRequestContextGetter* request_context,
+      const GURL& config_url,
+      const std::string& manifest_url_prefix,
+      const PrecacheUnfinishedWork* unfinished_work,

[sclittle, 2016/05/19 01:59:45]

Could you just pass in an unfinished_work with the starting_hosts already set?

[bengr, 2016/05/19 23:19:31]

Yes.

+      PrecacheDelegate* precache_delegate);
 
   virtual ~PrecacheFetcher();
 
   // Starts fetching resources to precache. URLs are fetched sequentially. Can
   // be called from any thread. Start should only be called once on a
   // PrecacheFetcher instance.
   void Start();
 
+  std::unique_ptr<PrecacheUnfinishedWork> GetUnfinishedWork();
+
+  // Used for testing to put the fetcher into a state where it has work to do.
+  void AssignWorkForTest(const std::list<GURL>& manifests,
+                         const std::list<GURL>& resources,
+                         size_t total_response_bytes,
+                         size_t network_response_bytes,
+                         size_t num_manifest_urls_to_fetch,
+                         const base::Time& start_time);
+
  private:
   // Fetches the next resource or manifest URL, if any remain. Fetching is done
   // sequentially and depth-first: all resources are fetched for a manifest
   // before the next manifest is fetched. This is done to limit the length of
   // the |resource_urls_to_fetch_| list, reducing the memory usage.
   void StartNextFetch();
 
   void StartNextManifestFetch();
   void StartNextResourceFetch();
 
   // Called when the precache configuration settings have been fetched.
   // Determines the list of manifest URLs to fetch according to the list of
   // |starting_hosts_| and information from the precache configuration settings.
   // If the fetch of the configuration settings fails, then precaching ends.
   void OnConfigFetchComplete(const Fetcher& source);
 
+  // Constructs manifest URLs using a manifest URL prefix, and lists of hosts.
+  void DetermineManifests();
+
   // Called when a precache manifest has been fetched. Builds the list of
   // resource URLs to fetch according to the URLs in the manifest. If the fetch
   // of a manifest fails, then it skips to the next manifest.
   void OnManifestFetchComplete(const Fetcher& source);
 
   // Called when a resource has been fetched.
   void OnResourceFetchComplete(const Fetcher& source);
 
   // Adds up the response sizes.
   void UpdateStats(int64_t response_bytes, int64_t network_response_bytes);
 
   // The prioritized list of starting hosts that the server will pick resource
   // URLs to be precached for.
   const std::vector<std::string> starting_hosts_;
 
   // The request context used when fetching URLs.
   const scoped_refptr<net::URLRequestContextGetter> request_context_;
 
+  // True if configuration settings have been retrieved, either from the network
+  // or from unfinished work.
+  bool has_config_;

[sclittle, 2016/05/19 01:59:45]

Could you just check if |config_| is non-null?

[bengr, 2016/05/19 23:19:31]

Done.

+
   // The custom URL to use when fetching the config. If not provided, the
   // default flag-specified URL will be used.
   const GURL config_url_;
 
   // The custom URL prefix to use when fetching manifests. If not provided, the
   // default flag-specified prefix will be used.
   const std::string manifest_url_prefix_;
 
   // Non-owning pointer. Should not be NULL.
   PrecacheDelegate* precache_delegate_;
 
   std::unique_ptr<PrecacheConfigurationSettings> config_;
 
   // Tally of the total number of bytes contained in URL fetches, including
   // config, manifests, and resources. This the number of bytes as they would be
   // compressed over the network.
   size_t total_response_bytes_;
 
   // Tally of the total number of bytes received over the network from URL
   // fetches (the same ones as in total_response_bytes_). This includes response
   // headers and intermediate responses such as 30xs.
   size_t network_response_bytes_;
 
   // Time when the prefetch was started.
-  base::TimeTicks start_time_;
+  base::Time start_time_;
 
-  int num_manifest_urls_to_fetch_;
+  size_t num_manifest_urls_to_fetch_;
   std::list<GURL> manifest_urls_to_fetch_;
   std::list<GURL> resource_urls_to_fetch_;
 
   FetcherPool<Fetcher> pool_;
 
   DISALLOW_COPY_AND_ASSIGN(PrecacheFetcher);
 };
 
 // Class that fetches a URL, and runs the specified callback when the fetch is
 // complete. This class exists so that a different method can be run in
@@ skipping 53 matching lines @@
   std::unique_ptr<net::URLFetcher> network_url_fetcher_;
   int64_t response_bytes_;
   int64_t network_response_bytes_;
 
   DISALLOW_COPY_AND_ASSIGN(Fetcher);
 };
 
 }  // namespace precache
 
 #endif  // COMPONENTS_PRECACHE_CORE_PRECACHE_FETCHER_H_