Examples of NodeWalker

com.projity.grouping.core.summaries.NodeWalker
Traversal functor which goes recursively thru children
org.apache.nutch.util.NodeWalker

A utility class that allows the walking of any DOM tree using a stack instead of recursion. As the node tree is walked the next node is popped off of the stack and all of its children are automatically added to the stack to be called in tree order.

Currently this class is not thread safe. It is assumed that only one thread will be accessing the NodeWalker at any given time.

Examples of com.projity.grouping.core.summaries.NodeWalker

    // group's special summaries handled here
    if (context == null)
      context = field.specialFieldContext;


    Object object = node.getImpl();
    NodeWalker walkingVisitor = (NodeWalker) field.getSummaryVisitor((object instanceof GroupNodeImpl) ? field.getSummaryForGroup() : field
        .getSummary(),object instanceof Document);
    walkingVisitor.setNode(node);
    walkingVisitor.setNodeModel(nodeModel);
    walkingVisitor.setContext(context);
    walkingVisitor.setField(field);
    Object result = walkingVisitor.getSummary();
    if (result instanceof Double) { // convert to proper display type
      result = ClassUtils.doubleToObject((Double) result, field.getDisplayType());
    }
    if ((object instanceof GroupNodeImpl) && field.hasOptions()) { // TODO
                                    // should

View Full Code Here

Examples of com.projity.grouping.core.summaries.NodeWalker

    // group's special summaries handled here
    if (context == null)
      context = field.specialFieldContext;


    Object object = node.getImpl();
    NodeWalker walkingVisitor = (NodeWalker) field.getSummaryVisitor((object instanceof GroupNodeImpl) ? field.getSummaryForGroup() : field
        .getSummary(),object instanceof Document);
    walkingVisitor.setNode(node);
    walkingVisitor.setNodeModel(nodeModel);
    walkingVisitor.setContext(context);
    walkingVisitor.setField(field);
    Object result = walkingVisitor.getSummary();
    if (result instanceof Double) { // convert to proper display type
      result = ClassUtils.doubleToObject((Double) result, field.getDisplayType());
    }
    if ((object instanceof GroupNodeImpl) && field.hasOptions()) { // TODO
                                    // should

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

  // anchors
  private boolean getTextHelper(StringBuffer sb, Node node, 
                                             boolean abortOnNestedAnchors,
                                             int anchorDepth) {
    boolean abort = false;
    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
    
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      if ("script".equalsIgnoreCase(nodeName)) {
        walker.skipChildren();
      }
      if ("style".equalsIgnoreCase(nodeName)) {
        walker.skipChildren();
      }
      if (abortOnNestedAnchors && "a".equalsIgnoreCase(nodeName)) {
        anchorDepth++;
        if (anchorDepth > 1) {
          abort = true;
          break;
        }        
      }
      if (nodeType == Node.COMMENT_NODE) {
        walker.skipChildren();
      }
      if (nodeType == Node.TEXT_NODE) {
        // cleanup and trim the value
        String text = currentNode.getNodeValue();
        text = text.replaceAll("\\s+", " ");

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

   *
   * @return true if a title node was found, false otherwise
   */
  public boolean getTitle(StringBuffer sb, Node node) {
    
    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
  
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      if ("body".equalsIgnoreCase(nodeName)) { // stop after HEAD
        return false;

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

  }


  /** If Node contains a BASE tag then it's HREF is returned. */
  public URL getBase(Node node) {


    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
  
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      // is this node a BASE tag?
      if (nodeType == Node.ELEMENT_NODE) {

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

   * nekohtml).
   */
  public void getOutlinks(URL base, ArrayList outlinks, 
                                       Node node) {
    
    NodeWalker walker = new NodeWalker(node);
    while (walker.hasNext()) {
      
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();      
      NodeList children = currentNode.getChildNodes();
      int childLen = (children != null) ? children.getLength() : 0;

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

  // anchors
  private boolean getTextHelper(StringBuffer sb, Node node, 
                                             boolean abortOnNestedAnchors,
                                             int anchorDepth) {
    boolean abort = false;
    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
    
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      if ("script".equalsIgnoreCase(nodeName)) {
        walker.skipChildren();
      }
      if ("style".equalsIgnoreCase(nodeName)) {
        walker.skipChildren();
      }
      if (abortOnNestedAnchors && "a".equalsIgnoreCase(nodeName)) {
        anchorDepth++;
        if (anchorDepth > 1) {
          abort = true;
          break;
        }        
      }
      if (nodeType == Node.COMMENT_NODE) {
        walker.skipChildren();
      }
      if (nodeType == Node.TEXT_NODE) {
        // cleanup and trim the value
        String text = currentNode.getNodeValue();
        text = text.replaceAll("\\s+", " ");

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

   *
   * @return true if a title node was found, false otherwise
   */
  public boolean getTitle(StringBuffer sb, Node node) {
    
    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
  
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      if ("body".equalsIgnoreCase(nodeName)) { // stop after HEAD
        return false;

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

  }


  /** If Node contains a BASE tag then it's HREF is returned. */
  URL getBase(Node node) {


    NodeWalker walker = new NodeWalker(node);
    
    while (walker.hasNext()) {
  
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();
      
      // is this node a BASE tag?
      if (nodeType == Node.ELEMENT_NODE) {

View Full Code Here

Examples of org.apache.nutch.util.NodeWalker

   * nekohtml).
   */
  public void getOutlinks(URL base, ArrayList<Outlink> outlinks, 
                                       Node node) {
    
    NodeWalker walker = new NodeWalker(node);
    while (walker.hasNext()) {
      
      Node currentNode = walker.nextNode();
      String nodeName = currentNode.getNodeName();
      short nodeType = currentNode.getNodeType();      
      NodeList children = currentNode.getChildNodes();
      int childLen = (children != null) ? children.getLength() : 0;

View Full Code Here

0 1 2 3 4

TOP

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.