Only show Reader Mode infobar on long articles

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 "base/metrics/histogram.h"
 
 #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"
@@ skipping 34 matching lines @@
 // Returns whether this update is the last one for the page.
 bool IsLast(bool is_loaded) {
   if (GetDistillerHeuristicsType() == DistillerHeuristicsType::ADABOOST_MODEL)
     return is_loaded;
 
   return true;
 }
 
 bool IsDistillablePageAdaboost(WebDocument& doc,
                                const DistillablePageDetector* detector,
+                               const DistillablePageDetector* long_page,

[mdjones, 2016/02/18 17:17:21]

If we add any more of these, we can probably just pass an array of page
detectors.

[wychen, 2016/02/18 21:47:23]

Acknowledged. It's hard to say whether we need more, or how
it would be used. If it's not ANDed together, having named
variable for each one is easier to understand.

                                bool is_last) {
   WebDistillabilityFeatures features = doc.distillabilityFeatures();
   GURL parsed_url(doc.url());
   if (!parsed_url.is_valid()) {
     return false;
   }
-  bool distillable = detector->Classify(CalculateDerivedFeatures(
+  std::vector<double> derived = CalculateDerivedFeatures(
     features.openGraph,
     parsed_url,
     features.elementCount,
     features.anchorCount,
     features.formCount,
     features.mozScore,
     features.mozScoreAllSqrt,
     features.mozScoreAllLinear
-  ));
+  );
+  bool distillable = detector->Classify(derived);
+  bool long_article = long_page->Classify(derived);
 
   int bucket = static_cast<unsigned>(features.isMobileFriendly) |
       (static_cast<unsigned>(distillable) << 1);
   if (is_last) {
     UMA_HISTOGRAM_ENUMERATION("DomDistiller.PageDistillableAfterLoading",
         bucket, 4);
   } else {
     UMA_HISTOGRAM_ENUMERATION("DomDistiller.PageDistillableAfterParsing",
         bucket, 4);
   }
-  return distillable && (!features.isMobileFriendly);
+  return distillable && long_article && (!features.isMobileFriendly);
 }
 
 bool IsDistillablePage(WebDocument& doc, bool is_last) {
   switch (GetDistillerHeuristicsType()) {
     case DistillerHeuristicsType::ALWAYS_TRUE:
       return true;
     case DistillerHeuristicsType::OG_ARTICLE:
       return doc.distillabilityFeatures().openGraph;
     case DistillerHeuristicsType::ADABOOST_MODEL:
-      return IsDistillablePageAdaboost(
-          doc, DistillablePageDetector::GetNewModel(), is_last);
+      return IsDistillablePageAdaboost(doc,
+          DistillablePageDetector::GetNewModel(),
+          DistillablePageDetector::GetLongPageModel(), is_last);
     case DistillerHeuristicsType::NONE:
     default:
       return false;
   }
 }
 
 }  // namespace
 
 DistillabilityAgent::DistillabilityAgent(
     content::RenderFrame* render_frame)
@@ skipping 19 matching lines @@
 
   bool is_last = IsLast(is_loaded);
   Send(new FrameHostMsg_Distillability(routing_id(),
       IsDistillablePage(doc, is_last), is_last));
 }
 
 
 DistillabilityAgent::~DistillabilityAgent() {}
 
 }  // namespace dom_distiller