Componentize renderer-side phishing classifier [1]

Initial componentization of chrome/renderer/safe_browsing
code.

In this patch:
 - move some initial files and tests to components/
   safe_browsing/content/renderer/phishing_classifier
   (in particular features* and test_utils code).
 - add new build config file,
 - update all relevant build files and includes.

BUG=1115700

Change-Id: I75e6ca754ad00135d8e6d9e466565885377b92f0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2437384
Commit-Queue: Tim Volodine <timvolodine@chromium.org>
Reviewed-by: Daniel Rubery <drubery@chromium.org>
Cr-Commit-Position: refs/heads/master@{#812800}

chrome/renderer/BUILD.gn

@@ -260,8 +260,6 @@ static_library("renderer") {
 
   if (safe_browsing_mode != 0) {
     sources += [
-      "safe_browsing/features.cc",
-      "safe_browsing/features.h",
       "safe_browsing/murmurhash3_util.cc",
       "safe_browsing/murmurhash3_util.h",
       "safe_browsing/phishing_classifier.cc",
@@ -280,6 +278,7 @@ static_library("renderer") {
     deps += [
       "//chrome/common/safe_browsing:proto",
       "//components/safe_browsing/content/renderer",
+      "//components/safe_browsing/content/renderer/phishing_classifier",
       "//components/safe_browsing/core:client_model_proto",
       "//components/safe_browsing/core:csd_proto",
       "//components/safe_browsing/core/common",
@@ -431,13 +430,12 @@ static_library("test_support") {
   sources = [
     "chrome_mock_render_thread.cc",
     "chrome_mock_render_thread.h",
-    "safe_browsing/test_utils.cc",
-    "safe_browsing/test_utils.h",
   ]
 
   deps = [
     ":renderer",
     "//chrome/common/search:mojo_bindings",
+    "//components/safe_browsing/content/renderer/phishing_classifier:unit_tests_support",
     "//content/test:test_support",
     "//extensions/buildflags",
     "//testing/gmock",

chrome/renderer/safe_browsing/phishing_classifier.cc

@@ -21,12 +21,12 @@
 #include "base/threading/thread_task_runner_handle.h"
 #include "cc/paint/skia_paint_canvas.h"
 #include "chrome/common/url_constants.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/phishing_dom_feature_extractor.h"
 #include "chrome/renderer/safe_browsing/phishing_term_feature_extractor.h"
 #include "chrome/renderer/safe_browsing/phishing_url_feature_extractor.h"
 #include "chrome/renderer/safe_browsing/scorer.h"
 #include "components/paint_preview/common/paint_preview_tracker.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "components/safe_browsing/core/proto/csd.pb.h"
 #include "content/public/renderer/render_frame.h"
 #include "content/public/renderer/render_thread.h"

chrome/renderer/safe_browsing/phishing_classifier_browsertest.cc

@@ -14,10 +14,10 @@
 #include "base/strings/utf_string_conversions.h"
 #include "base/test/test_discardable_memory_allocator.h"
 #include "chrome/renderer/chrome_content_renderer_client.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/murmurhash3_util.h"
 #include "chrome/renderer/safe_browsing/scorer.h"
 #include "chrome/test/base/chrome_render_view_test.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "components/safe_browsing/core/proto/client_model.pb.h"
 #include "components/safe_browsing/core/proto/csd.pb.h"
 #include "content/public/renderer/render_frame.h"

chrome/renderer/safe_browsing/phishing_classifier_delegate_browsertest.cc

@@ -8,12 +8,12 @@
 
 #include "base/bind.h"
 #include "base/strings/utf_string_conversions.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/phishing_classifier.h"
 #include "chrome/renderer/safe_browsing/scorer.h"
 #include "chrome/test/base/chrome_render_view_test.h"
 #include "chrome/test/base/chrome_unit_test_suite.h"
 #include "components/safe_browsing/content/common/safe_browsing.mojom-shared.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "components/safe_browsing/core/proto/csd.pb.h"
 #include "content/public/renderer/render_frame.h"
 #include "content/public/renderer/render_view.h"

chrome/renderer/safe_browsing/phishing_dom_feature_extractor.cc

@@ -15,7 +15,7 @@
 #include "base/threading/thread_task_runner_handle.h"
 #include "base/time/default_tick_clock.h"
 #include "base/time/time.h"
-#include "chrome/renderer/safe_browsing/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "content/public/renderer/render_view.h"
 #include "net/base/registry_controlled_domains/registry_controlled_domain.h"
 #include "third_party/blink/public/platform/web_string.h"

chrome/renderer/safe_browsing/phishing_dom_feature_extractor_browsertest.cc

@@ -12,9 +12,9 @@
 #include "base/memory/weak_ptr.h"
 #include "base/time/time.h"
 #include "chrome/renderer/chrome_content_renderer_client.h"
-#include "chrome/renderer/safe_browsing/features.h"
-#include "chrome/renderer/safe_browsing/test_utils.h"
 #include "chrome/test/base/chrome_render_view_test.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/test_utils.h"
 #include "content/public/common/content_switches.h"
 #include "content/public/renderer/render_frame.h"
 #include "content/public/test/test_utils.h"

chrome/renderer/safe_browsing/phishing_term_feature_extractor.cc

@@ -21,8 +21,8 @@
 #include "base/threading/thread_task_runner_handle.h"
 #include "base/time/default_tick_clock.h"
 #include "base/time/time.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/murmurhash3_util.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "crypto/sha2.h"
 
 namespace safe_browsing {

chrome/renderer/safe_browsing/phishing_term_feature_extractor_unittest.cc

@@ -23,9 +23,9 @@
 #include "base/threading/thread_task_runner_handle.h"
 #include "base/time/time.h"
 #include "build/build_config.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "chrome/renderer/safe_browsing/murmurhash3_util.h"
-#include "chrome/renderer/safe_browsing/test_utils.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/test_utils.h"
 #include "crypto/sha2.h"
 #include "testing/gmock/include/gmock/gmock.h"
 #include "testing/gtest/include/gtest/gtest.h"

chrome/renderer/safe_browsing/phishing_url_feature_extractor.cc

@@ -12,7 +12,7 @@
 #include "base/strings/string_split.h"
 #include "base/strings/string_util.h"
 #include "base/timer/elapsed_timer.h"
-#include "chrome/renderer/safe_browsing/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "net/base/registry_controlled_domains/registry_controlled_domain.h"
 #include "url/gurl.h"
 

chrome/renderer/safe_browsing/phishing_url_feature_extractor_unittest.cc

@@ -8,8 +8,8 @@
 #include <vector>
 #include "base/format_macros.h"
 #include "base/strings/stringprintf.h"
-#include "chrome/renderer/safe_browsing/features.h"
-#include "chrome/renderer/safe_browsing/test_utils.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/test_utils.h"
 #include "testing/gmock/include/gmock/gmock.h"
 #include "testing/gtest/include/gtest/gtest.h"
 #include "url/gurl.h"

chrome/renderer/safe_browsing/scorer.cc

@@ -15,8 +15,8 @@
 #include "base/strings/string_piece.h"
 #include "base/task/task_traits.h"
 #include "base/task/thread_pool.h"
-#include "chrome/renderer/safe_browsing/features.h"
 #include "components/safe_browsing/content/password_protection/visual_utils.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "components/safe_browsing/core/proto/client_model.pb.h"
 #include "components/safe_browsing/core/proto/csd.pb.h"
 #include "content/public/renderer/render_thread.h"

chrome/renderer/safe_browsing/scorer_unittest.cc

@@ -16,7 +16,7 @@
 #include "base/test/bind_test_util.h"
 #include "base/test/task_environment.h"
 #include "base/threading/thread.h"
-#include "chrome/renderer/safe_browsing/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "components/safe_browsing/core/proto/client_model.pb.h"
 #include "components/safe_browsing/core/proto/csd.pb.h"
 #include "testing/gmock/include/gmock/gmock.h"

chrome/test/BUILD.gn

@@ -5460,7 +5460,6 @@ test("unit_tests") {
       "../common/safe_browsing/ipc_protobuf_message_test_messages.h",
       "../common/safe_browsing/ipc_protobuf_message_unittest.cc",
       "../common/safe_browsing/mach_o_image_reader_mac_unittest.cc",
-      "../renderer/safe_browsing/features_unittest.cc",
       "../renderer/safe_browsing/murmurhash3_util_unittest.cc",
       "../renderer/safe_browsing/phishing_term_feature_extractor_unittest.cc",
       "../renderer/safe_browsing/phishing_url_feature_extractor_unittest.cc",

components/BUILD.gn

@@ -440,6 +440,10 @@ test("components_unittests") {
     deps += [ "//components/safe_browsing/android:unit_tests_mobile" ]
   }
 
+  if (safe_browsing_mode == 1) {
+    deps += [ "//components/safe_browsing/content/renderer/phishing_classifier:unit_tests" ]
+  }
+
   if (!is_ios) {
     deps += [
       "//components/safe_browsing/core/realtime:unit_tests",

components/safe_browsing/content/renderer/phishing_classifier/BUILD.gn

+# Copyright 2020 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.
+
+import("//build/config/features.gni")
+import("//components/safe_browsing/buildflags.gni")
+
+source_set("phishing_classifier") {
+  if (safe_browsing_mode != 0) {
+    sources = [
+      "features.cc",
+      "features.h",
+    ]
+    deps = [
+      "//components/safe_browsing/content/renderer",
+      "//components/safe_browsing/core/common",
+    ]
+  }
+}
+
+source_set("unit_tests") {
+  testonly = true
+  sources = [ "features_unittest.cc" ]
+  deps = [
+    ":phishing_classifier",
+    "//base:base",
+    "//components/safe_browsing/content/renderer/phishing_classifier:unit_tests_support",
+    "//testing/gtest",
+  ]
+}
+
+source_set("unit_tests_support") {
+  testonly = true
+  sources = [
+    "test_utils.cc",
+    "test_utils.h",
+  ]
+  deps = [
+    ":phishing_classifier",
+    "//testing/gmock",
+  ]
+}

chrome/renderer/safe_browsing/features.cc → components/safe_browsing/content/renderer/phishing_classifier/features.cc

@@ -2,7 +2,7 @@
 // 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/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 
 #include "base/metrics/histogram_macros.h"
 

chrome/renderer/safe_browsing/features.h → components/safe_browsing/content/renderer/phishing_classifier/features.h

@@ -22,8 +22,8 @@
 // just a thin wrapper around a map of feature name to value.  The entire set
 // of features for a URL is extracted before we do any scoring.
 
-#ifndef CHROME_RENDERER_SAFE_BROWSING_FEATURES_H_
-#define CHROME_RENDERER_SAFE_BROWSING_FEATURES_H_
+#ifndef COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_FEATURES_H_
+#define COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_FEATURES_H_
 
 #include <stddef.h>
 #include <string>
@@ -178,4 +178,4 @@ extern const char kPageTerm[];
 }  // namespace features
 }  // namespace safe_browsing
 
-#endif  // CHROME_RENDERER_SAFE_BROWSING_FEATURES_H_
+#endif  // COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_FEATURES_H_

chrome/renderer/safe_browsing/features_unittest.cc → components/safe_browsing/content/renderer/phishing_classifier/features_unittest.cc

@@ -2,13 +2,13 @@
 // 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/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 
 #include <stddef.h>
 
 #include "base/format_macros.h"
 #include "base/strings/stringprintf.h"
-#include "chrome/renderer/safe_browsing/test_utils.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/test_utils.h"
 #include "testing/gtest/include/gtest/gtest.h"
 
 namespace safe_browsing {
@@ -16,15 +16,15 @@ namespace safe_browsing {
 TEST(PhishingFeaturesTest, TooManyFeatures) {
   FeatureMap features;
   for (size_t i = 0; i < FeatureMap::kMaxFeatureMapSize; ++i) {
-    EXPECT_TRUE(features.AddBooleanFeature(
-        base::StringPrintf("Feature%" PRIuS, i)));
+    EXPECT_TRUE(
+        features.AddBooleanFeature(base::StringPrintf("Feature%" PRIuS, i)));
   }
   EXPECT_EQ(FeatureMap::kMaxFeatureMapSize, features.features().size());
 
   // Attempting to add more features should fail.
   for (size_t i = 0; i < 3; ++i) {
-    EXPECT_FALSE(features.AddBooleanFeature(
-        base::StringPrintf("Extra%" PRIuS, i)));
+    EXPECT_FALSE(
+        features.AddBooleanFeature(base::StringPrintf("Extra%" PRIuS, i)));
   }
   EXPECT_EQ(FeatureMap::kMaxFeatureMapSize, features.features().size());
 }

chrome/renderer/safe_browsing/test_utils.cc → components/safe_browsing/content/renderer/phishing_classifier/test_utils.cc

@@ -2,12 +2,12 @@
 // 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/test_utils.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/test_utils.h"
 
 #include <map>
 #include <string>
 
-#include "chrome/renderer/safe_browsing/features.h"
+#include "components/safe_browsing/content/renderer/phishing_classifier/features.h"
 #include "testing/gmock/include/gmock/gmock.h"
 
 namespace safe_browsing {

chrome/renderer/safe_browsing/test_utils.h → components/safe_browsing/content/renderer/phishing_classifier/test_utils.h

@@ -2,8 +2,8 @@
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
-#ifndef CHROME_RENDERER_SAFE_BROWSING_TEST_UTILS_H_
-#define CHROME_RENDERER_SAFE_BROWSING_TEST_UTILS_H_
+#ifndef COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_TEST_UTILS_H_
+#define COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_TEST_UTILS_H_
 
 namespace safe_browsing {
 class FeatureMap;
@@ -16,4 +16,4 @@ void ExpectFeatureMapsAreEqual(const FeatureMap& first,
 
 }  // namespace safe_browsing
 
-#endif  // CHROME_RENDERER_SAFE_BROWSING_TEST_UTILS_H_
+#endif  // COMPONENTS_SAFE_BROWSING_CONTENT_RENDERER_PHISHING_CLASSIFIER_TEST_UTILS_H_