Performance improvement on parsing larger HTML pages.

jhy · jhy · commit 8c43a813f98d · 2015-08-01T17:32:00.000-07:00
diff --git a/CHANGES b/CHANGES
@@ -6,6 +6,10 @@ jsoup changelog
 
  * When fetching XML URLs, automatically switch to the XML parser instead of the HTML parser.
    <https://github.com/jhy/jsoup/pull/574>
+
+ * Performance improvement on parsing larger HTML pages. On Android KitKat, around 1.7x times faster. On Android
+   Lollipop, ~ 1.3x faster. Improvements largely from re-ordering the HtmlTreeBuilder methods based on analysis of
+   various websites; also from further memory reduction for nodes with no children, and other tweaks.
    
  * Fixed an issue in Element.getElementSiblingIndex (and related methods) where sibling elements with the same content
    would incorrectly have the same sibling index.
diff --git a/src/main/java/org/jsoup/helper/StringUtil.java b/src/main/java/org/jsoup/helper/StringUtil.java
@@ -2,6 +2,7 @@
 
 import java.net.MalformedURLException;
 import java.net.URL;
+import java.util.Arrays;
 import java.util.Collection;
 import java.util.Iterator;
 
@@ -153,6 +154,10 @@ public static boolean in(String needle, String... haystack) {
         return false;
     }
 
+    public static boolean inSorted(String needle, String[] haystack) {
+        return Arrays.binarySearch(haystack, needle) >= 0;
+    }
+
     /**
      * Create a new absolute URL, from a provided existing absolute URL and a relative URL component.
      * @param base the existing absolulte base URL
diff --git a/src/main/java/org/jsoup/nodes/Element.java b/src/main/java/org/jsoup/nodes/Element.java
@@ -284,6 +284,7 @@ public Element appendChild(Node child) {
 
         // was - Node#addChildren(child). short-circuits an array create and a loop.
         reparentChild(child);
+        ensureChildNodes();
         childNodes.add(child);
         child.setSiblingIndex(childNodes.size() - 1);
         return this;
diff --git a/src/main/java/org/jsoup/nodes/Node.java b/src/main/java/org/jsoup/nodes/Node.java
@@ -18,6 +18,7 @@
 
  @author Jonathan Hedley, jonathan@hedley.net */
 public abstract class Node implements Cloneable {
+    private static final List<Node> EMPTY_NODES = Collections.emptyList();
     Node parentNode;
     List<Node> childNodes;
     Attributes attributes;
@@ -33,7 +34,7 @@ protected Node(String baseUri, Attributes attributes) {
         Validate.notNull(baseUri);
         Validate.notNull(attributes);
         
-        childNodes = new ArrayList<Node>(4);
+        childNodes = EMPTY_NODES;
         this.baseUri = baseUri.trim();
         this.attributes = attributes;
     }
@@ -46,7 +47,7 @@ protected Node(String baseUri) {
      * Default constructor. Doesn't setup base uri, children, or attributes; use with caution.
      */
     protected Node() {
-        childNodes = Collections.emptyList();
+        childNodes = EMPTY_NODES;
         attributes = null;
     }
 
@@ -428,6 +429,7 @@ protected void addChildren(Node... children) {
         //most used. short circuit addChildren(int), which hits reindex children and array copy
         for (Node child: children) {
             reparentChild(child);
+            ensureChildNodes();
             childNodes.add(child);
             child.setSiblingIndex(childNodes.size()-1);
         }
@@ -438,11 +440,18 @@ protected void addChildren(int index, Node... children) {
         for (int i = children.length - 1; i >= 0; i--) {
             Node in = children[i];
             reparentChild(in);
+            ensureChildNodes();
             childNodes.add(index, in);
         }
         reindexChildren(index);
     }
 
+    protected void ensureChildNodes() {
+        if (childNodes == EMPTY_NODES) {
+            childNodes = new ArrayList<Node>(4);
+        }
+    }
+
     protected void reparentChild(Node child) {
         if (child.parentNode != null)
             child.parentNode.removeChild(child);
diff --git a/src/main/java/org/jsoup/parser/HtmlTreeBuilderState.java b/src/main/java/org/jsoup/parser/HtmlTreeBuilderState.java
