Test distillability without JavaScript

third_party/WebKit/Source/core/dom/DocumentStatisticsCollector.cpp

Index: third_party/WebKit/Source/core/dom/DocumentStatisticsCollector.cpp
diff --git a/third_party/WebKit/Source/core/dom/DocumentStatisticsCollector.cpp b/third_party/WebKit/Source/core/dom/DocumentStatisticsCollector.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..76a408099a2c1c7cf72656761475b71850a395ec
--- /dev/null
+++ b/third_party/WebKit/Source/core/dom/DocumentStatisticsCollector.cpp
@@ -0,0 +1,95 @@
+// 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 "DocumentStatisticsCollector.h"
+
+#include "core/HTMLNames.h"
+#include "core/editing/EphemeralRange.h"
+#include "core/editing/iterators/TextIterator.h"
+#include "core/editing/iterators/WordAwareIterator.h"
+#include "core/inspector/ConsoleMessage.h"
+#include "platform/text/TextBreakIterator.h"
+#include "wtf/text/StringBuilder.h"
+
+using namespace WTF;
+using namespace Unicode;
+
+namespace blink {
+
+using namespace HTMLNames;
+
+
+DocumentStatisticsCollector::DocumentStatisticsCollector()
+    : m_statisticsCollected(false)
+    , m_readyToCollect(false)
+{
+}
+
+void DocumentStatisticsCollector::collectStatistics(Document& document)
+{
+    if (!m_readyToCollect)
+        return;
+
+    if (m_statisticsCollected)
+        return;
+
+    if (!document.frame() || !document.frame()->isMainFrame())
+        return;
+
+    if (!document.hasFinishedParsing())
+        return;
+
+    ASSERT(document.body());
+
+    StringBuilder message;
+    message.append("Done: ");
+    message.appendNumber(m_statisticsCollected);
+
+    m_statisticsCollected = true;
+
+    DocumentStatistics statistics = {0};
+
+    // First, traverse the DOM tree and collect statistics on element counts and textContent length.
+    for (Node& node : NodeTraversal::descendantsOf(*document.body())) {
+        statistics.nodeCount++;
+        if (node.isElementNode()) {
+            statistics.elementCount++;
+            Element& element = toElement(node);
+            if (element.hasTagName(aTag))
+                statistics.anchorElementCount++;
+            else if (element.hasTagName(formTag))
+                statistics.formElementCount++;
+        } else if (node.isTextNode()) {
+            String text = toText(node).data();
+            statistics.textContentLength += text.length();
+        }
+    }
+
+    // Next, traverse the Layout tree and collect statistics on innerText length.
+    EphemeralRange range = EphemeralRange::rangeOfContents(*document.body());
+    WordAwareIterator it(range.startPosition(), range.endPosition());
+    for (; !it.atEnd(); it.advance()) {
+        statistics.innerTextLength += it.length();
+    }
+
+    message.append(". FinishedParsing: ");
+    message.appendNumber(document.hasFinishedParsing());
+    message.append(". Node: ");
+    message.appendNumber(statistics.nodeCount);
+    message.append(", elements: ");
+    message.appendNumber(statistics.elementCount);
+    message.append(", anchors: ");
+    message.appendNumber(statistics.anchorElementCount);
+    message.append(", forms: ");
+    message.appendNumber(statistics.formElementCount);
+    message.append(", textContentLength: ");
+    message.appendNumber(statistics.textContentLength);
+    message.append(", innerTextLength: ");
+    message.appendNumber(statistics.innerTextLength);
+    RefPtrWillBeRawPtr<ConsoleMessage> consoleMessage = ConsoleMessage::create(ConsoleAPIMessageSource, DebugMessageLevel, message.toString());
+    // document.addConsoleMessage(consoleMessage);

[mdjones, 2015/09/29 17:31:42]

Is this logging code permanent?

[wychen, 2015/09/29 21:46:57]

The WebKit part is not done yet. This is just for debugging only.

+}
+
+}