Test distillability without JavaScript

components/dom_distiller/content/renderer/distillability_agent.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/content/common/distiller_messages.h"
+#include "components/dom_distiller/content/renderer/distillability_agent.h"
+#include "components/dom_distiller/core/distillable_page_detector.h"
+#include "components/dom_distiller/core/experiments.h"
+#include "components/dom_distiller/core/page_features.h"
+#include "content/public/renderer/render_frame.h"
+
+#include "third_party/WebKit/public/platform/WebDistillability.h"
+#include "third_party/WebKit/public/web/WebDocument.h"
+#include "third_party/WebKit/public/web/WebElement.h"
+#include "third_party/WebKit/public/web/WebLocalFrame.h"
+#include "third_party/WebKit/public/web/WebNode.h"
+#include "third_party/WebKit/public/web/WebNodeList.h"
+
+namespace dom_distiller {
+
+using namespace blink;
+
+namespace {
+
+// Returns whether it is necessary to send updates back to the browser.
+// The number of updates can be from 0 to 2. See the tests in
+// "distillable_page_utils_browsertest.cc".
+// Most heuristics types only require one update after parsing.
+// Adaboost is the only one doing the second update, which is after loading.
+bool needToUpdate(bool is_loaded) {
+  switch (GetDistillerHeuristicsType()) {
+    case DistillerHeuristicsType::ALWAYS_TRUE:
+      return !is_loaded;
+    case DistillerHeuristicsType::OG_ARTICLE:
+      return !is_loaded;
+    case DistillerHeuristicsType::ADABOOST_MODEL:
+      return true;
+    case DistillerHeuristicsType::NONE:
+    default:
+      return false;
+  }
+}
+
+// Returns whether this update is the last one for the page.
+bool isLast(bool is_loaded) {
+  switch (GetDistillerHeuristicsType()) {
+    case DistillerHeuristicsType::ALWAYS_TRUE:
+      return true;
+    case DistillerHeuristicsType::OG_ARTICLE:
+      return true;
+    case DistillerHeuristicsType::ADABOOST_MODEL:
+      return is_loaded;
+    case DistillerHeuristicsType::NONE:
+    default:
+      return true;
+  }
+}
+
+bool IsDistillablePageAdaboost(WebDocument& doc,
+                               const DistillablePageDetector* detector) {
+  WebDistillabilityFeatures features = doc.distillabilityFeatures();
+  GURL parsed_url(doc.url());
+  if (!parsed_url.is_valid()) {
+    return false;
+  }
+  // The adaboost model is only applied to non-mobile pages.
+  if (features.isMobileFriendly) {
+    return false;
+  }
+  return detector->Classify(CalculateDerivedFeatures(features, parsed_url));
+}
+
+bool IsDistillablePage(WebDocument& doc) {
+  switch (GetDistillerHeuristicsType()) {
+    case DistillerHeuristicsType::ALWAYS_TRUE:
+      return true;
+    case DistillerHeuristicsType::OG_ARTICLE:
+      return doc.distillabilityFeatures().openGraph;
+    case DistillerHeuristicsType::ADABOOST_MODEL:
+      return IsDistillablePageAdaboost(
+          doc, DistillablePageDetector::GetDefault());
+    case DistillerHeuristicsType::NONE:
+    default:
+      return false;
+  }
+}
+
+}  // namespace
+
+DistillabilityAgent::DistillabilityAgent(
+    content::RenderFrame* render_frame)
+    : RenderFrameObserver(render_frame) {
+}
+
+void DistillabilityAgent::DidMeaningfulLayout(
+    WebMeaningfulLayout layout_type) {
+  if (layout_type != WebMeaningfulLayout::FinishedParsing &&
+      layout_type != WebMeaningfulLayout::FinishedLoading) {
+    return;
+  }
+
+  DCHECK(render_frame());
+  if (!render_frame()->IsMainFrame()) return;
+  DCHECK(render_frame()->GetWebFrame());
+  WebDocument doc = render_frame()->GetWebFrame()->document();
+  if (doc.isNull() || doc.body().isNull()) return;
+
+  bool is_loaded = layout_type == WebMeaningfulLayout::FinishedLoading;
+  if (!needToUpdate(is_loaded)) return;
+
+  Send(new FrameHostMsg_Distillability(routing_id(),
+      IsDistillablePage(doc), isLast(is_loaded)));
+}
+
+
+DistillabilityAgent::~DistillabilityAgent() {}
+
+}  // namespace dom_distiller