Add a new set of page features for distillability testing

components/dom_distiller/core/page_features_unittest.cc

 // Copyright 2015 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/dom_distiller/core/page_features.h"
 
 #include <string>
 #include <vector>
 
 #include "base/files/file_util.h"
 #include "base/json/json_reader.h"
 #include "base/json/json_writer.h"
 #include "base/memory/scoped_ptr.h"
 #include "base/path_service.h"
 #include "testing/gtest/include/gtest/gtest.h"
+#include "url/gurl.h"
 
 namespace dom_distiller {
 
 // This test uses input data of core features and the output of the training
 // pipeline's derived feature extraction to ensure that the extraction that is
 // done in Chromium matches that in the training pipeline.
 TEST(DomDistillerPageFeaturesTest, TestCalculateDerivedFeatures) {
   base::FilePath dir_source_root;
   EXPECT_TRUE(PathService::Get(base::DIR_SOURCE_ROOT, &dir_source_root));
   std::string input_data;
@@ skipping 62 matching lines @@
         bool bool_value;
         ASSERT_TRUE(derived_features->GetBoolean(value_index, &bool_value));
         expected_value = bool_value ? 1.0 : 0.0;
       }
       EXPECT_DOUBLE_EQ(derived[j], expected_value)
           << "incorrect value for entry with url " << entry_url
           << " for derived feature " << labels[j];
     }
   }
 }
+
+std::vector<double> DeriveFromPath(const GURL& url) {
+  return CalculateDerivedFeatures(
+    false, // bool openGraph
+    url,   // const GURL& url
+    0,     // unsigned elementCount
+    0,     // unsigned anchorCount
+    0,     // unsigned formCount
+    0,     // double mozScore
+    0,     // double mozScoreAllSqrt
+    0      // double mozScoreAllLinear
+  );
 }
+
+TEST(DomDistillerPageFeaturesTest, TestPath) {
+  GURL url("http://example.com/search/view/index/the-title-of-archive.php");
+
+  std::vector<double> derived(DeriveFromPath(url));
+  EXPECT_EQ(0, lround(derived[1]));
+  EXPECT_EQ(1, lround(derived[2]));
+  EXPECT_EQ(1, lround(derived[3]));
+  EXPECT_EQ(1, lround(derived[4]));
+  EXPECT_EQ(1, lround(derived[5]));
+  EXPECT_EQ(0, lround(derived[6]));
+  EXPECT_EQ(0, lround(derived[7]));
+  EXPECT_EQ(1, lround(derived[8]));
+  EXPECT_EQ(43, lround(derived[9]));
+  EXPECT_EQ(0, lround(derived[10]));
+  EXPECT_EQ(4, lround(derived[11]));
+  EXPECT_EQ(4, lround(derived[12]));
+  EXPECT_EQ(0, lround(derived[13]));
+  EXPECT_EQ(24, lround(derived[14]));
+}
+
+TEST(DomDistillerPageFeaturesTest, TestPath2) {
+  GURL url("http://example.com/phpbb/forum123/456.asp");
+
+  std::vector<double> derived(DeriveFromPath(url));
+  EXPECT_EQ(1, lround(derived[1]));
+  EXPECT_EQ(0, lround(derived[2]));
+  EXPECT_EQ(0, lround(derived[3]));
+  EXPECT_EQ(0, lround(derived[4]));
+  EXPECT_EQ(0, lround(derived[5]));
+  EXPECT_EQ(1, lround(derived[6]));
+  EXPECT_EQ(1, lround(derived[7]));
+  EXPECT_EQ(0, lround(derived[8]));
+  EXPECT_EQ(23, lround(derived[9]));
+  EXPECT_EQ(0, lround(derived[10]));
+  EXPECT_EQ(3, lround(derived[11]));
+  EXPECT_EQ(1, lround(derived[12]));
+  EXPECT_EQ(2, lround(derived[13]));
+  EXPECT_EQ(7, lround(derived[14]));
+}
+
+TEST(DomDistillerPageFeaturesTest, TestPath3) {
+  GURL url("https://example.com/");
+
+  std::vector<double> derived(DeriveFromPath(url));
+  EXPECT_EQ(0, lround(derived[1]));
+  EXPECT_EQ(0, lround(derived[2]));
+  EXPECT_EQ(0, lround(derived[3]));
+  EXPECT_EQ(0, lround(derived[4]));
+  EXPECT_EQ(0, lround(derived[5]));
+  EXPECT_EQ(0, lround(derived[6]));
+  EXPECT_EQ(0, lround(derived[7]));
+  EXPECT_EQ(0, lround(derived[8]));
+  EXPECT_EQ(1, lround(derived[9]));
+  EXPECT_EQ(1, lround(derived[10]));
+  EXPECT_EQ(0, lround(derived[11]));
+  EXPECT_EQ(0, lround(derived[12]));
+  EXPECT_EQ(0, lround(derived[13]));
+  EXPECT_EQ(0, lround(derived[14]));
+}
+}