| Index: third_party/WebKit/Source/core/dom/DocumentStatisticsCollectorTest.cpp
|
| diff --git a/third_party/WebKit/Source/core/dom/DocumentStatisticsCollectorTest.cpp b/third_party/WebKit/Source/core/dom/DocumentStatisticsCollectorTest.cpp
|
| new file mode 100644
|
| index 0000000000000000000000000000000000000000..b465b668f207ff893d669033f3760265afd98274
|
| --- /dev/null
|
| +++ b/third_party/WebKit/Source/core/dom/DocumentStatisticsCollectorTest.cpp
|
| @@ -0,0 +1,128 @@
|
| +// 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 "config.h"
|
| +#include "core/dom/DocumentStatisticsCollector.h"
|
| +
|
| +#include "core/dom/Document.h"
|
| +#include "core/dom/DocumentVisibilityObserver.h"
|
| +#include "core/frame/FrameView.h"
|
| +#include "core/html/HTMLHeadElement.h"
|
| +#include "core/html/HTMLLinkElement.h"
|
| +#include "core/testing/DummyPageHolder.h"
|
| +#include "public/platform/WebDistillability.h"
|
| +#include <gmock/gmock.h>
|
| +#include <gtest/gtest.h>
|
| +
|
| +namespace blink {
|
| +
|
| +class DocumentStatisticsCollectorTest : public ::testing::Test {
|
| +protected:
|
| + void SetUp() override;
|
| +
|
| +#if ENABLE(OILPAN)
|
| + void TearDown() override
|
| + {
|
| + Heap::collectAllGarbage();
|
| + }
|
| +#endif
|
| +
|
| + Document& document() const { return m_dummyPageHolder->document(); }
|
| +
|
| + void setHtmlInnerHTML(const String&);
|
| +
|
| +private:
|
| + OwnPtr<DummyPageHolder> m_dummyPageHolder;
|
| +};
|
| +
|
| +void DocumentStatisticsCollectorTest::SetUp()
|
| +{
|
| + m_dummyPageHolder = DummyPageHolder::create(IntSize(800, 600));
|
| +}
|
| +
|
| +void DocumentStatisticsCollectorTest::setHtmlInnerHTML(const String& htmlContent)
|
| +{
|
| + document().documentElement()->setInnerHTML((htmlContent), ASSERT_NO_EXCEPTION);
|
| + document().view()->updateAllLifecyclePhases();
|
| +}
|
| +
|
| +// This test checks open graph articles can be recognized.
|
| +TEST_F(DocumentStatisticsCollectorTest, HasOpenGraphArticle)
|
| +{
|
| + setHtmlInnerHTML(
|
| + "<head>"
|
| + // Note the case-insensitive matching of the word "article".
|
| + " <meta property='og:type' content='arTiclE' />"
|
| + "</head>"
|
| + );
|
| + WebDistillabilityFeatures features = DocumentStatisticsCollector::collectStatistics(document());
|
| +
|
| + EXPECT_EQ(true, features.openGraph);
|
| +}
|
| +
|
| +// This test checks non-existence of open graph articles can be recognized.
|
| +TEST_F(DocumentStatisticsCollectorTest, NoOpenGraphArticle)
|
| +{
|
| + setHtmlInnerHTML(
|
| + "<head>"
|
| + " <meta property='og:type' content='movie' />"
|
| + "</head>"
|
| + );
|
| + WebDistillabilityFeatures features = DocumentStatisticsCollector::collectStatistics(document());
|
| +
|
| + EXPECT_EQ(false, features.openGraph);
|
| +}
|
| +
|
| +// This test checks element counts are correct.
|
| +TEST_F(DocumentStatisticsCollectorTest, CountElements)
|
| +{
|
| + setHtmlInnerHTML(
|
| + "<form>"
|
| + " <input type='text'>"
|
| + " <input type='password'>"
|
| + "</form>"
|
| + "<pre></pre>"
|
| + "<p><a> </a></p>"
|
| + "<ul><li><p><a> </a></p></li></ul>"
|
| + );
|
| + WebDistillabilityFeatures features = DocumentStatisticsCollector::collectStatistics(document());
|
| +
|
| + EXPECT_EQ(false, features.openGraph);
|
| +
|
| + EXPECT_EQ(10u, features.elementCount);
|
| + EXPECT_EQ(2u, features.anchorCount);
|
| + EXPECT_EQ(1u, features.formCount);
|
| + EXPECT_EQ(1u, features.textInputCount);
|
| + EXPECT_EQ(1u, features.passwordInputCount);
|
| + EXPECT_EQ(2u, features.pCount);
|
| + EXPECT_EQ(1u, features.preCount);
|
| +}
|
| +
|
| +// This test checks score calculations are correct.
|
| +TEST_F(DocumentStatisticsCollectorTest, CountScore)
|
| +{
|
| + setHtmlInnerHTML(
|
| + "<p class='menu' id='article'> 1 </p>" // trimmedTextContentLength = 1
|
| + "<ul><li><p>12</p></li></ul>" // trimmedTextContentLength = 2, skipped because under li
|
| + "<p class='menu'>123</p>" // trimmedTextContentLength = 3, skipped because unlikelyCandidates
|
| + "<p>"
|
| + "12345678901234567890123456789012345678901234567890"
|
| + "12345678901234567890123456789012345678901234567890"
|
| + "12345678901234567890123456789012345678901234"
|
| + "</p>" // trimmedTextContentLength = 144
|
| + "<p style='display:none'>12345</p>" // trimmedTextContentLength = 5, skipped because invisible
|
| + "<div style='visibility:hidden'><p>123456</p></div>" // trimmedTextContentLength = 6, skipped because invisible
|
| + "<p style='opacity:0'>1234567</p>" // trimmedTextContentLength = 7, skipped because invisible
|
| + "<p> <a href='#'> 12345 </a> 9 <b> </b> </p>" // trimmedTextContentLength = 9
|
| + );
|
| + WebDistillabilityFeatures features = DocumentStatisticsCollector::collectStatistics(document());
|
| +
|
| + const unsigned kParagraphLengthThreshold = 140;
|
| +
|
| + EXPECT_DOUBLE_EQ(features.mozScore, sqrt(144 - kParagraphLengthThreshold));
|
| + EXPECT_DOUBLE_EQ(features.mozScoreAllSqrt, 1 + sqrt(144) + sqrt(9));
|
| + EXPECT_DOUBLE_EQ(features.mozScoreAllLinear, 1 + 144 + 9);
|
| +}
|
| +
|
| +} // namespace blink
|
|
|
Chromium Code Reviews
Issue 1419033004:
Add feature extraction for distillability to Blink (Closed)
Base URL: https://chromium.googlesource.com/chromium/src.git@master