Add pause/resume functionality to precache

components/precache/core/precache_fetcher.cc

 // 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.
 
 #include "components/precache/core/precache_fetcher.h"
 
 #include <algorithm>
 #include <limits>
 #include <string>
 #include <utility>
 #include <vector>
 
 #include "base/bind.h"
+#include "base/bind_helpers.h"
 #include "base/callback.h"
 #include "base/command_line.h"
 #include "base/compiler_specific.h"
 #include "base/containers/hash_tables.h"
+#include "base/location.h"
 #include "base/logging.h"
 #include "base/memory/ptr_util.h"
+#include "base/memory/ref_counted.h"
 #include "base/metrics/histogram_macros.h"
+#include "components/precache/core/precache_database.h"
 #include "components/precache/core/precache_switches.h"
 #include "components/precache/core/proto/precache.pb.h"
 #include "net/base/completion_callback.h"
 #include "net/base/escape.h"
 #include "net/base/io_buffer.h"
 #include "net/base/load_flags.h"
 #include "net/base/net_errors.h"
 #include "net/http/http_response_headers.h"
 #include "net/url_request/url_fetcher_response_writer.h"
 #include "net/url_request/url_request_context_getter.h"
@@ skipping 223 matching lines @@
   response_bytes_ = source->GetReceivedResponseContentLength();
   network_response_bytes_ = source->GetTotalReceivedBytes();
   callback_.Run(*this);
 }
 
 PrecacheFetcher::PrecacheFetcher(
     const std::vector<std::string>& starting_hosts,
     net::URLRequestContextGetter* request_context,
     const GURL& config_url,
     const std::string& manifest_url_prefix,
-    PrecacheFetcher::PrecacheDelegate* precache_delegate)
+    PrecacheFetcher::PrecacheDelegate* precache_delegate,
+    base::WeakPtr<PrecacheDatabase> precache_database,
+    scoped_refptr<base::SingleThreadTaskRunner> database_task_runner)
     : starting_hosts_(starting_hosts),
       request_context_(request_context),
       config_url_(config_url),
       manifest_url_prefix_(manifest_url_prefix),
       precache_delegate_(precache_delegate),
       config_(new PrecacheConfigurationSettings),
+      precache_database_(precache_database),
+      database_task_runner_(database_task_runner),
       total_response_bytes_(0),
       network_response_bytes_(0),
       num_manifest_urls_to_fetch_(0),
       pool_(kMaxParallelFetches) {
   DCHECK(request_context_.get());  // Request context must be non-NULL.
   DCHECK(precache_delegate_);  // Precache delegate must be non-NULL.
 
   DCHECK_NE(GURL(), GetDefaultConfigURL())
       << "Could not determine the precache config settings URL.";
   DCHECK_NE(std::string(), GetDefaultManifestURLPrefix())
@@ skipping 22 matching lines @@
                                  num_manifest_urls_to_fetch_ * 100);
   UMA_HISTOGRAM_PERCENTAGE("Precache.Fetch.PercentCompleted",
                            percent_completed);
   UMA_HISTOGRAM_CUSTOM_COUNTS("Precache.Fetch.ResponseBytes.Total",
                               total_response_bytes_, 1, kMaxResponseBytes, 100);
   UMA_HISTOGRAM_CUSTOM_COUNTS("Precache.Fetch.ResponseBytes.Network",
                               network_response_bytes_, 1, kMaxResponseBytes,
                               100);
 }
 
+
+
+
+void PrecacheFetcher::Init(
+    const base::TimeDelta& max_age) {
+  // Fetch unfinished work that is no older than |max_age|.
+  scoped_refptr<UnfinishedWork> unfinished_work =
+      new UnfinishedWork(base::TimeTicks::Now() - max_age);
+  database_task_runner_->PostTaskAndReply(

[sclittle, 2016/05/10 00:01:26]

I'm not sure that UnfinishedWork needs to be refcounted or passed as a
scoped_refptr.

Could you just use PostTaskAndReplyWithResult, where the result is a
std::unique_ptr<UnfinishedWork>?
https://code.google.com/p/chromium/codesearch#chromium/src/base/task_runner_u...

[bengr, 2016/05/19 01:25:44]

Done.

+      FROM_HERE,
+      base::Bind(&PrecacheDatabase::GetUnfinishedWork, precache_database_,
+                 unfinished_work),
+      base::Bind(&PrecacheFetcher::OnInitDone, AsWeakPtr(),
+                 unfinished_work));
+}
+
+void PrecacheFetcher::OnInitDone(
+    scoped_refptr<UnfinishedWork> unfinished_work) {
+  if (unfinished_work->manifests.empty() &&
+      unfinished_work->resources.empty()) {
+    if (unfinished_work->total_bytes > 0) {
+      //TODO(rajendrant): report expired unfinished job.

[sclittle, 2016/05/10 00:01:26]

nit: add space between "//" and "TODO", and capitalize "report".

[bengr, 2016/05/19 01:25:44]

Done.

+    }
+  } else {
+    // Starting hosts are cleared because this is a continuation.
+    manifest_urls_to_fetch_ = unfinished_work->manifests;

[sclittle, 2016/05/10 00:01:26]

Use std::move on the lists here to avoid copying them all.

[bengr, 2016/05/19 01:25:44]

Done.

+    resource_urls_to_fetch_ = unfinished_work->resources;
+    total_response_bytes_ = unfinished_work->total_bytes;
+    network_response_bytes_ = unfinished_work->network_bytes;
+    num_manifest_urls_to_fetch_ = unfinished_work->num_manifest_urls;
+    start_time_ = unfinished_work->start_time;
+  }
+  precache_delegate_->OnInitDone();
+}
+
+void PrecacheFetcher::Shutdown() {
+  scoped_refptr<UnfinishedWork> unfinished_work = new UnfinishedWork(
+      manifest_urls_to_fetch_, resource_urls_to_fetch_,
+      total_response_bytes_, network_response_bytes_,
+      num_manifest_urls_to_fetch_, start_time_);
+  database_task_runner_->PostTaskAndReply(
+      FROM_HERE,
+      base::Bind(&PrecacheDatabase::SaveUnfinishedWork, precache_database_,
+                 unfinished_work),
+      base::Bind(&PrecacheFetcher::OnShutdownDone, AsWeakPtr()));
+}
+
+void PrecacheFetcher::OnShutdownDone() {
+  precache_delegate_->OnShutdownDone();
+}
+
 void PrecacheFetcher::Start() {
   GURL config_url =
       config_url_.is_empty() ? GetDefaultConfigURL() : config_url_;
 
   DCHECK(config_url.is_valid()) << "Config URL not valid: "
                                 << config_url.possibly_invalid_spec();
 
-  start_time_ = base::TimeTicks::Now();
+  if (start_time_.is_null())
+    start_time_ = base::TimeTicks::Now();
 
   // Fetch the precache configuration settings from the server.
   DCHECK(pool_.IsEmpty()) << "All parallel requests should be available";
   VLOG(3) << "Fetching " << config_url;
   pool_.Add(base::WrapUnique(new Fetcher(
       request_context_.get(), config_url,
       base::Bind(&PrecacheFetcher::OnConfigFetchComplete,
                  base::Unretained(this)),
       false /* is_resource_request */, std::numeric_limits<int32_t>::max())));
 }
@@ skipping 65 matching lines @@
                              : manifest_url_prefix_;
     DCHECK_NE(std::string(), prefix)
         << "Could not determine the precache manifest URL prefix.";
 
     // Keep track of manifest URLs that are being fetched, in order to elide
     // duplicates.
     base::hash_set<std::string> seen_manifest_urls;
 
     // Attempt to fetch manifests for starting hosts up to the maximum top sites
     // count. If a manifest does not exist for a particular starting host, then
-    // the fetch will fail, and that starting host will be ignored.
-    int64_t rank = 0;
-    for (const std::string& host : starting_hosts_) {
-      ++rank;
-      if (rank > config_->top_sites_count())
-        break;
-      AppendManifestURLIfNew(prefix, host, &seen_manifest_urls,
-                             &manifest_urls_to_fetch_);
+    // the fetch will fail, and that starting host will be ignored. Starting
+    // hosts are not added if this is a continuation from a previous precache
+    // session.
+    if (manifest_urls_to_fetch_.empty()) {

[sclittle, 2016/05/10 00:01:26]

What about if the UnfinishedWork already finished fetching the manifests, and
only has resource URLs left? It would try to fetch all the manifests again here,
wouldn't it?

[bengr, 2016/05/19 01:25:44]

Good catch. Thanks.

+      int64_t rank = 0;
+      for (const std::string& host : starting_hosts_) {
+        ++rank;
+        if (rank > config_->top_sites_count())
+          break;
+        AppendManifestURLIfNew(prefix, host, &seen_manifest_urls,
+                               &manifest_urls_to_fetch_);
+      }
+
+      for (const std::string& host : config_->forced_site()) {
+        AppendManifestURLIfNew(prefix, host, &seen_manifest_urls,
+                               &manifest_urls_to_fetch_);
+      }
     }
-
-    for (const std::string& host : config_->forced_site())
-      AppendManifestURLIfNew(prefix, host, &seen_manifest_urls,
-                             &manifest_urls_to_fetch_);
-
     num_manifest_urls_to_fetch_ = manifest_urls_to_fetch_.size();
   }
   pool_.Delete(source);
 
   StartNextFetch();
 }
 
 void PrecacheFetcher::OnManifestFetchComplete(const Fetcher& source) {
   UpdateStats(source.response_bytes(), source.network_response_bytes());
   if (source.network_url_fetcher() == nullptr) {
     pool_.DeleteAll();  // Cancel any other ongoing request.
   } else {
     PrecacheManifest manifest;
 
     if (ParseProtoFromFetchResponse(*source.network_url_fetcher(), &manifest)) {
       const int32_t len =
           std::min(manifest.resource_size(), config_->top_resources_count());
       for (int i = 0; i < len; ++i) {
-        if (manifest.resource(i).has_url()) {
+        if (manifest.resource(i).has_url())
           resource_urls_to_fetch_.push_back(GURL(manifest.resource(i).url()));
-        }
       }
     }
   }
 
   pool_.Delete(source);
   StartNextFetch();
 }
 
 void PrecacheFetcher::OnResourceFetchComplete(const Fetcher& source) {
   UpdateStats(source.response_bytes(), source.network_response_bytes());
   pool_.Delete(source);
   // The resource has already been put in the cache during the fetch process, so
   // nothing more needs to be done for the resource.
   StartNextFetch();
 }
 
 void PrecacheFetcher::UpdateStats(int64_t response_bytes,
                                   int64_t network_response_bytes) {
   total_response_bytes_ += response_bytes;
   network_response_bytes_ += network_response_bytes;
 }
 
+void PrecacheFetcher::AssignWorkForTest(const std::list<GURL>& manifests,
+                                        const std::list<GURL>& resources,
+                                        size_t total_response_bytes,
+                                        size_t network_response_bytes,
+                                        int num_manifest_urls_to_fetch,
+                                        const base::TimeTicks& start_time) {
+  DCHECK(manifest_urls_to_fetch_.empty());
+  DCHECK(resource_urls_to_fetch_.empty());
+  manifest_urls_to_fetch_ = manifests;
+  resource_urls_to_fetch_ = resources;
+  total_response_bytes_ = total_response_bytes;
+  network_response_bytes_ = network_response_bytes;
+  num_manifest_urls_to_fetch_ = num_manifest_urls_to_fetch;
+  start_time_ = start_time;
+}
+
 }  // namespace precache