Switch to the new client-side phishing model that uses Murmurhash for word hashes.

chrome/renderer/safe_browsing/phishing_term_feature_extractor_unittest.cc

 // 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.
 
 #include "chrome/renderer/safe_browsing/phishing_term_feature_extractor.h"
 
 #include <string>
 
 #include "base/bind.h"
 #include "base/callback.h"
 #include "base/hash_tables.h"
 #include "base/memory/scoped_ptr.h"
 #include "base/message_loop.h"
 #include "base/string16.h"
 #include "base/stringprintf.h"
 #include "base/time.h"
 #include "base/utf_string_conversions.h"
 #include "crypto/sha2.h"
 #include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/mock_feature_extractor_clock.h"
+#include "chrome/renderer/safe_browsing/murmurhash3_util.h"
 #include "testing/gmock/include/gmock/gmock.h"
 #include "testing/gtest/include/gtest/gtest.h"
 
 using ::testing::ContainerEq;
 using ::testing::Return;
 
 namespace safe_browsing {
 
 class PhishingTermFeatureExtractorTest : public ::testing::Test {
  protected:
@@ skipping 23 matching lines @@
     words.insert("multi");
     words.insert("word");
     words.insert("test");
     words.insert("capitalization");
     words.insert("space");
     words.insert("separator");
     words.insert("punctuation");
     words.insert("\xe4\xbd\xa0\xe5\xa5\xbd");
     words.insert("\xe5\x86\x8d\xe8\xa7\x81");
 
+    static const uint32 kMurmurHash3Seed = 2777808611U;
     for (base::hash_set<std::string>::iterator it = words.begin();
          it != words.end(); ++it) {
-      word_hashes_.insert(crypto::SHA256HashString(*it));
+      word_hashes_.insert(MurmurHash3String(*it, kMurmurHash3Seed));
     }
 
     extractor_.reset(new PhishingTermFeatureExtractor(
         &term_hashes_,
         &word_hashes_,
         3 /* max_words_per_term */,
+        kMurmurHash3Seed,
         &clock_));
   }
 
   // Runs the TermFeatureExtractor on |page_text|, waiting for the
   // completion callback.  Returns the success boolean from the callback.
   bool ExtractFeatures(const string16* page_text, FeatureMap* features) {
     success_ = false;
     extractor_->ExtractFeatures(
         page_text,
         features,
@@ skipping 22 matching lines @@
 
   void QuitExtraction() {
     extractor_->CancelPendingExtraction();
     msg_loop_.Quit();
   }
 
   MessageLoop msg_loop_;
   MockFeatureExtractorClock clock_;
   scoped_ptr<PhishingTermFeatureExtractor> extractor_;
   base::hash_set<std::string> term_hashes_;
-  base::hash_set<std::string> word_hashes_;
+  base::hash_set<uint32> word_hashes_;
   bool success_;  // holds the success value from ExtractFeatures
 };
 
 TEST_F(PhishingTermFeatureExtractorTest, ExtractFeatures) {
   // This test doesn't exercise the extraction timing.
   EXPECT_CALL(clock_, Now()).WillRepeatedly(Return(base::TimeTicks::Now()));
 
   string16 page_text = ASCIIToUTF16("blah");
   FeatureMap expected_features;  // initially empty
 
@@ skipping 175 matching lines @@
   // Now extract normally and make sure nothing breaks.
   EXPECT_TRUE(ExtractFeatures(page_text.get(), &features));
 
   FeatureMap expected_features;
   expected_features.AddBooleanFeature(features::kPageTerm +
                                       std::string("multi word test"));
   EXPECT_THAT(features.features(), ContainerEq(expected_features.features()));
 }
 
 }  // namespace safe_browsing