[CacheStorage] Give cache directories unique names

content/browser/cache_storage/cache_storage.cc

 // Copyright 2014 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/cache_storage/cache_storage.h"
 
 #include <string>
 
 #include "base/barrier_closure.h"
 #include "base/files/file_util.h"
 #include "base/files/memory_mapped_file.h"
+#include "base/guid.h"
 #include "base/location.h"
 #include "base/memory/ref_counted.h"
 #include "base/metrics/histogram_macros.h"
 #include "base/numerics/safe_conversions.h"
 #include "base/sha1.h"
 #include "base/single_thread_task_runner.h"
 #include "base/stl_util.h"
 #include "base/strings/string_number_conversions.h"
 #include "base/strings/string_util.h"
 #include "base/thread_task_runner_handle.h"
 #include "content/browser/cache_storage/cache_storage.pb.h"
 #include "content/browser/cache_storage/cache_storage_cache.h"
 #include "content/browser/cache_storage/cache_storage_scheduler.h"
 #include "content/public/browser/browser_thread.h"
 #include "net/base/directory_lister.h"
 #include "net/base/net_errors.h"
 #include "net/url_request/url_request_context_getter.h"
 #include "storage/browser/blob/blob_storage_context.h"
 #include "storage/browser/quota/quota_manager_proxy.h"
 
 namespace content {
 
 namespace {
 
+std::string HexedHash(const std::string& value) {
+  std::string value_hash = base::SHA1HashString(value);
+  std::string valued_hexed_hash = base::ToLowerASCII(
+      base::HexEncode(value_hash.c_str(), value_hash.length()));
+  return valued_hexed_hash;
+}
+
 void CloseAllCachesDidCloseCache(const scoped_refptr<CacheStorageCache>& cache,
                                  const base::Closure& barrier_closure) {
   barrier_closure.Run();
 }
 
 }  // namespace
 
+class MigratedLegacyCacheDirectoryNameTest;

[jsbell, 2015/10/23 21:09:59]

Is this actually needed? (forward decls for friends do not appear to be
necessary if same namespace)

[jkarlin, 2015/10/23 22:57:41]

Done.

+
 const char CacheStorage::kIndexFileName[] = "index.txt";
 
-// Handles the loading and clean up of CacheStorageCache objects. The
-// callback of every public method is guaranteed to be called.
+// Handles the loading and clean up of CacheStorageCache objects.
 class CacheStorage::CacheLoader {
  public:
   typedef base::Callback<void(const scoped_refptr<CacheStorageCache>&)>
       CacheCallback;
   typedef base::Callback<void(bool)> BoolCallback;
   typedef base::Callback<void(scoped_ptr<std::vector<std::string>>)>
       StringVectorCallback;
 
   CacheLoader(
       base::SequencedTaskRunner* cache_task_runner,
@@ skipping 10 matching lines @@
   }
 
   virtual ~CacheLoader() {}
 
   // Creates a CacheStorageCache with the given name. It does not attempt to
   // load the backend, that happens lazily when the cache is used.
   virtual scoped_refptr<CacheStorageCache> CreateCache(
       const std::string& cache_name) = 0;
 
   // Deletes any pre-existing cache of the same name and then loads it.
-  virtual void CreateCache(const std::string& cache_name,
-                           const CacheCallback& callback) = 0;
+  virtual void PrepareNewCacheDestination(const std::string& cache_name,
+                                          const CacheCallback& callback) = 0;
 
   // After the backend has been deleted, do any extra house keeping such as
   // removing the cache's directory.
   virtual void CleanUpDeletedCache(const std::string& key,
                                    const BoolCallback& callback) = 0;
 
   // Writes the cache names (and sizes) to disk if applicable.
   virtual void WriteIndex(const StringVector& cache_names,
                           const BoolCallback& callback) = 0;
 
@@ skipping 26 matching lines @@
                     quota_manager_proxy,
                     blob_context,
                     origin) {}
 
   scoped_refptr<CacheStorageCache> CreateCache(
       const std::string& cache_name) override {
     return CacheStorageCache::CreateMemoryCache(
         origin_, request_context_getter_, quota_manager_proxy_, blob_context_);
   }
 
-  void CreateCache(const std::string& cache_name,
-                   const CacheCallback& callback) override {
+  void PrepareNewCacheDestination(const std::string& cache_name,
+                                  const CacheCallback& callback) override {
     scoped_refptr<CacheStorageCache> cache = CreateCache(cache_name);
     cache_refs_.insert(std::make_pair(cache_name, cache));
     callback.Run(cache);
   }
 
   void CleanUpDeletedCache(const std::string& cache_name,
                            const BoolCallback& callback) override {
     CacheRefMap::iterator it = cache_refs_.find(cache_name);
     DCHECK(it != cache_refs_.end());
     cache_refs_.erase(it);
@@ skipping 33 matching lines @@
                     request_context,
                     quota_manager_proxy,
                     blob_context,
                     origin),
         origin_path_(origin_path),
         weak_ptr_factory_(this) {}
 
   scoped_refptr<CacheStorageCache> CreateCache(
       const std::string& cache_name) override {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
+    DCHECK(ContainsKey(cache_name_to_cache_dir_, cache_name));
 
+    std::string cache_dir = cache_name_to_cache_dir_[cache_name];
+    base::FilePath cache_path = origin_path_.AppendASCII(cache_dir);
     return CacheStorageCache::CreatePersistentCache(
-        origin_, CreatePersistentCachePath(origin_path_, cache_name),
-        request_context_getter_, quota_manager_proxy_, blob_context_);
+        origin_, cache_path, request_context_getter_, quota_manager_proxy_,
+        blob_context_);
   }
 
-  void CreateCache(const std::string& cache_name,
-                   const CacheCallback& callback) override {
+  void PrepareNewCacheDestination(const std::string& cache_name,
+                                  const CacheCallback& callback) override {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
 
-    // 1. Delete the cache's directory if it exists.
-    // (CreateCacheDeleteFilesInPool)
-    // 2. Load the cache. (LoadCreateDirectoryInPool)
-
-    base::FilePath cache_path =
-        CreatePersistentCachePath(origin_path_, cache_name);
-
     PostTaskAndReplyWithResult(
         cache_task_runner_.get(), FROM_HERE,
-        base::Bind(&SimpleCacheLoader::CreateCachePrepDirInPool, cache_path),
-        base::Bind(&SimpleCacheLoader::CreateCachePreppedDir, cache_name,
-                   callback, weak_ptr_factory_.GetWeakPtr()));
+        base::Bind(&SimpleCacheLoader::PrepareNewCacheDirectoryInPool,
+                   origin_path_),
+        base::Bind(&SimpleCacheLoader::PrepareNewCacheCreateCache,
+                   weak_ptr_factory_.GetWeakPtr(), cache_name, callback));
   }
 
-  static bool CreateCachePrepDirInPool(const base::FilePath& cache_path) {
-    if (base::PathExists(cache_path))
-      base::DeleteFile(cache_path, /* recursive */ true);
-    return base::CreateDirectory(cache_path);
+  // Runs on the cache_task_runner_.
+  static std::string PrepareNewCacheDirectoryInPool(
+      const base::FilePath& origin_path) {
+    std::string cache_dir;
+    base::FilePath cache_path;
+    do {
+      cache_dir = base::GenerateGUID();
+      cache_path = origin_path.AppendASCII(cache_dir);
+    } while (base::PathExists(cache_path));
+
+    return base::CreateDirectory(cache_path) ? cache_dir : "";
   }
 
-  static void CreateCachePreppedDir(const std::string& cache_name,
-                                    const CacheCallback& callback,
-                                    base::WeakPtr<SimpleCacheLoader> loader,
-                                    bool success) {
-    if (!success || !loader) {
+  void PrepareNewCacheCreateCache(const std::string& cache_name,
+                                  const CacheCallback& callback,
+                                  const std::string& cache_dir) {
+    if (cache_dir.empty()) {
       callback.Run(scoped_refptr<CacheStorageCache>());
       return;
     }
 
-    callback.Run(loader->CreateCache(cache_name));
+    cache_name_to_cache_dir_[cache_name] = cache_dir;
+    callback.Run(CreateCache(cache_name));
   }
 
   void CleanUpDeletedCache(const std::string& cache_name,
                            const BoolCallback& callback) override {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
-
-    // 1. Delete the cache's directory. (CleanUpDeleteCacheDirInPool)
+    DCHECK(ContainsKey(cache_name_to_cache_dir_, cache_name));
 
     base::FilePath cache_path =
-        CreatePersistentCachePath(origin_path_, cache_name);
+        origin_path_.AppendASCII(cache_name_to_cache_dir_[cache_name]);
+    cache_name_to_cache_dir_.erase(cache_name);
+
     cache_task_runner_->PostTask(
         FROM_HERE,
         base::Bind(&SimpleCacheLoader::CleanUpDeleteCacheDirInPool, cache_path,
                    callback, base::ThreadTaskRunnerHandle::Get()));
   }
 
   static void CleanUpDeleteCacheDirInPool(
       const base::FilePath& cache_path,
       const BoolCallback& callback,
       const scoped_refptr<base::SingleThreadTaskRunner>& original_task_runner) {
     bool rv = base::DeleteFile(cache_path, true);
     original_task_runner->PostTask(FROM_HERE, base::Bind(callback, rv));
   }
 
   void WriteIndex(const StringVector& cache_names,
                   const BoolCallback& callback) override {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
 
     // 1. Create the index file as a string. (WriteIndex)
     // 2. Write the file to disk. (WriteIndexWriteToFileInPool)
 
     CacheStorageIndex index;
     index.set_origin(origin_.spec());
 
     for (size_t i = 0u, max = cache_names.size(); i < max; ++i) {
+      DCHECK(ContainsKey(cache_name_to_cache_dir_, cache_names[i]));
+
       CacheStorageIndex::Cache* index_cache = index.add_cache();
       index_cache->set_name(cache_names[i]);
+      index_cache->set_cache_dir(cache_name_to_cache_dir_[cache_names[i]]);
     }
 
     std::string serialized;
     bool success = index.SerializeToString(&serialized);
     DCHECK(success);
 
     base::FilePath tmp_path = origin_path_.AppendASCII("index.txt.tmp");
     base::FilePath index_path =
         origin_path_.AppendASCII(CacheStorage::kIndexFileName);
 
@@ skipping 23 matching lines @@
   void LoadIndex(scoped_ptr<std::vector<std::string>> names,
                  const StringVectorCallback& callback) override {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
 
     // 1. Read the file from disk. (LoadIndexReadFileInPool)
     // 2. Parse file and return the names of the caches (LoadIndexDidReadFile)
 
     base::FilePath index_path =
         origin_path_.AppendASCII(CacheStorage::kIndexFileName);
 
-    cache_task_runner_->PostTask(
-        FROM_HERE, base::Bind(&SimpleCacheLoader::LoadIndexReadFileInPool,
-                              index_path, base::Passed(names.Pass()), callback,
-                              base::ThreadTaskRunnerHandle::Get()));
+    PostTaskAndReplyWithResult(
+        cache_task_runner_.get(), FROM_HERE,
+        base::Bind(&SimpleCacheLoader::ReadAndMigrateIndexInPool, index_path),
+        base::Bind(&SimpleCacheLoader::LoadIndexDidReadFile,
+                   weak_ptr_factory_.GetWeakPtr(), base::Passed(&names),
+                   callback));
   }
 
-  static void LoadIndexReadFileInPool(
-      const base::FilePath& index_path,
-      scoped_ptr<std::vector<std::string>> names,
-      const StringVectorCallback& callback,
-      const scoped_refptr<base::SingleThreadTaskRunner>& original_task_runner) {
-    std::string body;
-    base::ReadFileToString(index_path, &body);
-
-    original_task_runner->PostTask(
-        FROM_HERE, base::Bind(&SimpleCacheLoader::LoadIndexDidReadFile,
-                              base::Passed(names.Pass()), callback, body));
-  }
-
-  static void LoadIndexDidReadFile(scoped_ptr<std::vector<std::string>> names,
-                                   const StringVectorCallback& callback,
-                                   const std::string& serialized) {
+  void LoadIndexDidReadFile(scoped_ptr<std::vector<std::string>> names,
+                            const StringVectorCallback& callback,
+                            const std::string& serialized) {
     DCHECK_CURRENTLY_ON(BrowserThread::IO);
 
     CacheStorageIndex index;
     if (index.ParseFromString(serialized)) {
       for (int i = 0, max = index.cache_size(); i < max; ++i) {
         const CacheStorageIndex::Cache& cache = index.cache(i);
+        DCHECK(cache.has_cache_dir());
         names->push_back(cache.name());
+        cache_name_to_cache_dir_[cache.name()] = cache.cache_dir();
       }
     }
 
     // TODO(jkarlin): Delete caches that are in the directory and not returned
     // in LoadIndex.
     callback.Run(names.Pass());
   }
 
  private:
+  friend class MigratedLegacyCacheDirectoryNameTest;
   ~SimpleCacheLoader() override {}
 
-  static std::string HexedHash(const std::string& value) {
-    std::string value_hash = base::SHA1HashString(value);
-    std::string valued_hexed_hash = base::ToLowerASCII(
-        base::HexEncode(value_hash.c_str(), value_hash.length()));
-    return valued_hexed_hash;
+  // Runs on cache_task_runner_
+  static std::string MigrateCachesIfNecessaryInPool(
+      const std::string& body,
+      const base::FilePath& index_path) {
+    CacheStorageIndex index;
+    base::FilePath origin_path = index_path.DirName();
+
+    if (!index.ParseFromString(body))
+      return body;
+
+    bool index_is_dirty = false;
+
+    // Look for caches that have no cache_dir. Give any such caches a directory
+    // with a random name and move them there. Then, rewrite the index file.
+    for (int i = 0, max = index.cache_size(); i < max; ++i) {
+      const CacheStorageIndex::Cache& cache = index.cache(i);
+      if (!cache.has_cache_dir()) {
+        // find a new home for the cache.

[jsbell, 2015/10/23 21:09:59]

Nit: Capitalize.

[jkarlin, 2015/10/23 22:57:41]

Done.

+        base::FilePath legacy_cache_path =
+            origin_path.AppendASCII(HexedHash(cache.name()));
+        std::string cache_dir;
+        base::FilePath cache_path;
+        do {
+          cache_dir = base::GenerateGUID();
+          cache_path = origin_path.AppendASCII(cache_dir);
+        } while (base::PathExists(cache_path));
+
+        if (!base::Move(legacy_cache_path, cache_path))
+          return "";

[jsbell, 2015/10/23 21:09:59]

At this point, some caches may have been moved so the index is bogus. It looks
like returning a non-parsable string would then run into the "Delete caches that
are in the directory and not returned" TODO above?

Maybe add a comment here?

And/or possibly define a const std::string error(""); up above to make the
various 'return ""' more self-documenting?

[jkarlin, 2015/10/23 22:57:41]

Done.

+
+        index.mutable_cache(0)->set_cache_dir(cache_dir);
+        index_is_dirty = true;
+      }
+    }
+
+    if (index_is_dirty) {
+      std::string new_body;
+      if (!index.SerializeToString(&new_body))
+        return "";
+      if (base::WriteFile(index_path, new_body.c_str(), new_body.size()) !=
+          base::checked_cast<int>(new_body.size()))
+        return "";
+      return new_body;
+    }
+
+    return body;
   }
 
-  static base::FilePath CreatePersistentCachePath(
-      const base::FilePath& origin_path,
-      const std::string& cache_name) {
-    return origin_path.AppendASCII(HexedHash(cache_name));
+  // Runs on cache_task_runner_
+  static std::string ReadAndMigrateIndexInPool(
+      const base::FilePath& index_path) {
+    std::string body;
+    base::ReadFileToString(index_path, &body);
+
+    return MigrateCachesIfNecessaryInPool(body, index_path);
   }
 
   const base::FilePath origin_path_;
+  std::map<std::string, std::string> cache_name_to_cache_dir_;
 
   base::WeakPtrFactory<SimpleCacheLoader> weak_ptr_factory_;
 };
 
 CacheStorage::CacheStorage(
     const base::FilePath& path,
     bool memory_only,
     base::SequencedTaskRunner* cache_task_runner,
     const scoped_refptr<net::URLRequestContextGetter>& request_context,
     const scoped_refptr<storage::QuotaManagerProxy>& quota_manager_proxy,
@@ skipping 198 matching lines @@
 }
 
 void CacheStorage::OpenCacheImpl(const std::string& cache_name,
                                  const CacheAndErrorCallback& callback) {
   scoped_refptr<CacheStorageCache> cache = GetLoadedCache(cache_name);
   if (cache.get()) {
     callback.Run(cache, CACHE_STORAGE_OK);
     return;
   }
 
-  cache_loader_->CreateCache(
+  cache_loader_->PrepareNewCacheDestination(
       cache_name, base::Bind(&CacheStorage::CreateCacheDidCreateCache,
                              weak_factory_.GetWeakPtr(), cache_name, callback));
 }
 
 void CacheStorage::CreateCacheDidCreateCache(
     const std::string& cache_name,
     const CacheAndErrorCallback& callback,
     const scoped_refptr<CacheStorageCache>& cache) {
   DCHECK_CURRENTLY_ON(BrowserThread::IO);
 
@@ skipping 22 matching lines @@
   DCHECK_CURRENTLY_ON(BrowserThread::IO);
   DCHECK(cache.get());
 
   // TODO(jkarlin): Handle !success.
 
   callback.Run(cache, CACHE_STORAGE_OK);
 }
 
 void CacheStorage::HasCacheImpl(const std::string& cache_name,
                                 const BoolAndErrorCallback& callback) {
-  bool has_cache = cache_map_.find(cache_name) != cache_map_.end();
-
+  bool has_cache = ContainsKey(cache_map_, cache_name);
   callback.Run(has_cache, CACHE_STORAGE_OK);
 }
 
 void CacheStorage::DeleteCacheImpl(const std::string& cache_name,
                                    const BoolAndErrorCallback& callback) {
   CacheMap::iterator it = cache_map_.find(cache_name);
   if (it == cache_map_.end()) {
     callback.Run(false, CACHE_STORAGE_ERROR_NOT_FOUND);
     return;
   }
@@ skipping 229 matching lines @@
     scoped_ptr<ServiceWorkerResponse> response,
     scoped_ptr<storage::BlobDataHandle> blob_data_handle) {
   base::WeakPtr<CacheStorage> cache_storage = weak_factory_.GetWeakPtr();
 
   callback.Run(error, response.Pass(), blob_data_handle.Pass());
   if (cache_storage)
     scheduler_->CompleteOperationAndRunNext();
 }
 
 }  // namespace content