Thread: [Htmlparser-cvs] htmlparser/src/org/htmlparser Node.java,1.50,1.51 PrototypicalNodeFactory.java,1.10

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Update of /cvsroot/htmlparser/htmlparser/src/org/htmlparser
In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv32670/src/org/htmlparser

Modified Files:
	Node.java PrototypicalNodeFactory.java Tag.java 
Log Message:
Part four of a multiphase refactoring.
Most internals now use the Tag interface.
This interface has been broadened to add set/get scanner and set/get endtag.
Removed the org.htmlparser.tags.Tag class and moved the remaining (minor) functionality
to the TagNode class. So now tags inherit directly from TagNode or CompositeTag.
** NOTE: If you have subclassed org.htmlparser.tags.Tag, use org.htmlparser.nodes.TagNode now.**
Removed deprecated methods getTagBegin/getTagEnd and deleted unused classes:
PeekingIterator and it's Implementation.


Index: Node.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/Node.java,v
retrieving revision 1.50
retrieving revision 1.51
diff -C2 -d -r1.50 -r1.51
*** Node.java	14 Jun 2004 00:06:51 -0000	1.50
--- Node.java	2 Jul 2004 00:49:26 -0000	1.51
***************
*** 27,30 ****
--- 27,31 ----
  package org.htmlparser;
  
+ import org.htmlparser.lexer.Page;
  import org.htmlparser.util.NodeList;
  import org.htmlparser.util.ParserException;
***************
*** 171,174 ****
--- 172,186 ----
  
      /**
+      * Get the page this node came from.
+      * @return The page that supplied this node.
+      */
+     public Page getPage ();
+ 
+     /**
+      * Set the page this node came from.
+      * @param page The page that supplied this node.
+      */
+     public void setPage (Page page);
+     /**
       * Apply the visitor to this node.
       * @param visitor The visitor to this node.

Index: PrototypicalNodeFactory.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/PrototypicalNodeFactory.java,v
retrieving revision 1.10
retrieving revision 1.11
diff -C2 -d -r1.10 -r1.11
*** PrototypicalNodeFactory.java	26 Jun 2004 11:25:01 -0000	1.10
--- PrototypicalNodeFactory.java	2 Jul 2004 00:49:26 -0000	1.11
***************
*** 43,46 ****
--- 43,47 ----
  import org.htmlparser.nodes.TextNode;
  import org.htmlparser.nodes.RemarkNode;
+ import org.htmlparser.nodes.TagNode;
  import org.htmlparser.tags.AppletTag;
  import org.htmlparser.tags.BaseHrefTag;
***************
*** 106,109 ****
--- 107,115 ----
  
      /**
+      * The prototypical tag node.
+      */
+     protected Tag mTag;
+ 
+     /**
       * The list of tags to return.
       * The list is keyed by tag name.
***************
*** 129,132 ****
--- 135,139 ----
          mText = new TextNode (null, 0, 0);
          mRemark = new RemarkNode (null, 0, 0);
+         mTag = new TagNode (null, 0, 0, null);
          if (!empty)
              registerTags ();
***************
*** 137,141 ****
       * @param tag The single tag to register in the otherwise empty factory.
       */
!     public PrototypicalNodeFactory (org.htmlparser.tags.Tag tag)
      {
          this (true);
--- 144,148 ----
       * @param tag The single tag to register in the otherwise empty factory.
       */
!     public PrototypicalNodeFactory (Tag tag)
      {
          this (true);
***************
*** 147,151 ****
       * @param tags The tags to register in the otherwise empty factory.
       */
!     public PrototypicalNodeFactory (org.htmlparser.tags.Tag[] tags)
      {
          this (true);
--- 154,158 ----
       * @param tags The tags to register in the otherwise empty factory.
       */
!     public PrototypicalNodeFactory (Tag[] tags)
      {
          this (true);
***************
*** 207,213 ****
       * Registers the given tag under every id the tag has.
       * @param tag The tag to register (subclass of
!      * {@link org.htmlparser.tags.Tag}).
       */
!     public void registerTag (org.htmlparser.tags.Tag tag)
      {
          String ids[];
--- 214,220 ----
       * Registers the given tag under every id the tag has.
       * @param tag The tag to register (subclass of
!      * {@link Tag}).
       */
!     public void registerTag (Tag tag)
      {
          String ids[];
***************
*** 222,228 ****
       * Unregisters the given tag from every id the tag has.
       * @param tag The tag to unregister (subclass of
!      * {@link org.htmlparser.tags.Tag}).
       */
!     public void unregisterTag (org.htmlparser.tags.Tag tag)
      {
          String ids[];
--- 229,235 ----
       * Unregisters the given tag from every id the tag has.
       * @param tag The tag to unregister (subclass of
!      * {@link Tag}).
       */
!     public void unregisterTag (Tag tag)
      {
          String ids[];
***************
*** 234,260 ****
  
      /**
-      * Register a tag.
-      * Registers the given tag under the tag {@link Tag#getTagName() name}.
-      * @param tag The tag to register (implements {@link org.htmlparser.Tag}).
-      */
-     public void registerTag (Tag tag)
-     {
-         put (tag.getTagName (), tag);
-     }
- 
-     /**
-      * Unregister a tag.
-      * Unregisters the given tag from the tag {@link Tag#getTagName() name}.
-      * @param tag The tag to unregister (implements {@link org.htmlparser.Tag}).
-      */
-     public void unregisterTag (Tag tag)
-     {
-         remove (tag.getTagName ());
-     }
- 
-     /**
       * Register all known tags in the tag package.
       * Registers tags from the {@link org.htmlparser.tags tag package} by
!      * calling {@link #registerTag(org.htmlparser.tags.Tag) registerTag()}.
       * @return 'this' nodefactory as a convenience.
       */
--- 241,247 ----
  
      /**
       * Register all known tags in the tag package.
       * Registers tags from the {@link org.htmlparser.tags tag package} by
!      * calling {@link #registerTag(Tag) registerTag()}.
       * @return 'this' nodefactory as a convenience.
       */
***************
*** 337,340 ****
--- 324,352 ----
      }
  
+     /**
+      * Get the object being used to generate generic tag nodes.
+      * These are returned from {@link createTagNode} when no specific tag
+      * is found in the registered tag list.
+      * @return The prototype for {@link Tag} nodes.
+      */
+     public Tag getTagPrototype ()
+     {
+         return (mTag);
+     }
+ 
+     /**
+      * Set the object to be used to generate tag nodes.
+      * These are returned from {@link createTagNode} when no specific tag
+      * is found in the registered tag list.
+      * @param remark The prototype for {@link Tag} nodes.
+      */
+     public void setTagPrototype (Tag tag)
+     {
+         if (null == tag)
+             throw new IllegalArgumentException ("tag prototype node cannot be null");
+         else
+             mTag = tag;
+     }
+ 
      //
      // NodeFactory interface
***************
*** 354,361 ****
          {
              ret = (Text)(getTextPrototype ().clone ());
!             if (ret instanceof AbstractNode)
!                 ((AbstractNode)ret).setPage (page);
!             else
!                 ret.setText (page.getText (start, end));
              ret.setStartPosition (start);
              ret.setEndPosition (end);
--- 366,370 ----
          {
              ret = (Text)(getTextPrototype ().clone ());
!             ret.setPage (page);
              ret.setStartPosition (start);
              ret.setEndPosition (end);
***************
*** 384,398 ****
          {
              ret = (Remark)(getRemarkPrototype ().clone ());
!             if (ret instanceof AbstractNode)
!                 ((AbstractNode)ret).setPage (page);
!             else
!             {
!                 first = start + 4; // <!--
!                 last = end - 3; // -->
!                 if (first >= last)
!                     ret.setText ("");
!                 else
!                     ret.setText (page.getText (first, last));
!             }
              ret.setStartPosition (start);
              ret.setEndPosition (end);
--- 393,397 ----
          {
              ret = (Remark)(getRemarkPrototype ().clone ());
!             ret.setPage (page);
              ret.setStartPosition (start);
              ret.setEndPosition (end);
***************
*** 445,450 ****
                          {
                              ret = (Tag)prototype.clone ();
!                             if (ret instanceof AbstractNode)
!                                 ((AbstractNode)ret).setPage (page);
                              ret.setStartPosition (start);
                              ret.setEndPosition (end);
--- 444,448 ----
                          {
                              ret = (Tag)prototype.clone ();
!                             ret.setPage (page);
                              ret.setStartPosition (start);
                              ret.setEndPosition (end);
***************
*** 460,465 ****
          }
          if (null == ret)
!             // generate a generic node
!             ret = new org.htmlparser.tags.Tag (page, start, end, attributes);
  
          return (ret);
--- 458,475 ----
          }
          if (null == ret)
!         {   // generate a generic node
!             try
!             {
!                 ret = (Tag)getTagPrototype ().clone ();
!                 ret.setPage (page);
!                 ret.setStartPosition (start);
!                 ret.setEndPosition (end);
!                 ret.setAttributesEx (attributes);
!             }
!             catch (CloneNotSupportedException cnse)
!             {
!                 ret = new TagNode (page, start, end, attributes);
!             }
!         }
  
          return (ret);

Index: Tag.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/Tag.java,v
retrieving revision 1.3
retrieving revision 1.4
diff -C2 -d -r1.3 -r1.4
*** Tag.java	24 May 2004 16:18:12 -0000	1.3
--- Tag.java	2 Jul 2004 00:49:26 -0000	1.4
***************
*** 27,32 ****
--- 27,35 ----
  package org.htmlparser;
  
+ import java.util.Hashtable;
  import java.util.Vector;
  
+ import org.htmlparser.scanners.Scanner;
+ 
  /**
   * Identifies what a Tag such as &lt;XXX xxx yyy="zzz"&gt; can do.
***************
*** 96,99 ****
--- 99,128 ----
       */
      public void setAttributesEx (Vector attribs);
+     
+     /**
+      * Gets the attributes in the tag.
+      * This is not the preferred  method to get attributes, see {@link
+      * #getAttributesEx getAttributesEx} which returns a list of {@link
+      * Attribute} objects, which offer more information than the simple
+      * <code>String</code> objects available from this <code>Hashtable</code>.
+      * @return Returns a list of name/value pairs representing the attributes.
+      * These are not in order, the keys (names) are converted to uppercase and the values
+      * are not quoted, even if they need to be. The table <em>will</em> return
+      * <code>null</code> if there was no value for an attribute (no equals
+      * sign or nothing to the right of the equals sign). A special entry with
+      * a key of SpecialHashtable.TAGNAME ("$<TAGNAME>$") holds the tag name.
+      * The conversion to uppercase is performed with an ENGLISH locale.
+      * @deprecated Use getAttributesEx() instead.
+      */
+     public Hashtable getAttributes ();
+ 
+     /**
+      * Sets the attributes.
+      * A special entry with a key of SpecialHashtable.TAGNAME ("$<TAGNAME>$")
+      * sets the tag name.
+      * @param attributes The attribute collection to set.
+      * @deprecated Use setAttributesEx() instead.
+      */
+     public void setAttributes (Hashtable attributes);
  
      /**
***************
*** 119,122 ****
--- 148,158 ----
  
      /**
+      * Return the name of this tag.
+      * @return The tag name or null if this tag contains nothing or only
+      * whitespace.
+      */
+     public String getRawTagName ();
+ 
+     /**
       * Determines if the given tag breaks the flow of text.
       * @return <code>true</code> if following text would start on a new line,
***************
*** 152,154 ****
--- 188,254 ----
       */
      public void setEmptyXmlTag (boolean emptyXmlTag);
+ 
+     /**
+      * Return the set of names handled by this tag.
+      * Since this a a generic tag, it has no ids.
+      * @return The names to be matched that create tags of this type.
+      */
+     public String[] getIds ();
+ 
+     /**
+      * Return the set of tag names that cause this tag to finish.
+      * These are the normal (non end tags) that if encountered while
+      * scanning (a composite tag) will cause the generation of a virtual
+      * tag.
+      * Since this a a non-composite tag, the default is no enders.
+      * @return The names of following tags that stop further scanning.
+      */
+     public String[] getEnders ();
+ 
+     /**
+      * Return the set of end tag names that cause this tag to finish.
+      * These are the end tags that if encountered while
+      * scanning (a composite tag) will cause the generation of a virtual
+      * tag.
+      * Since this a a non-composite tag, it has no end tag enders.
+      * @return The names of following end tags that stop further scanning.
+      */
+     public String[] getEndTagEnders ();
+ 
+     /**
+      * Get the end tag for this (composite) tag.
+      * For a non-composite tag this always returns <code>null</code>.
+      * @return The tag that terminates this composite tag, i.e. &lt;/HTML&gt;.
+      */
+     public Tag getEndTag ();
+ 
+     /**
+      * Set the end tag for this (composite) tag.
+      * For a non-composite tag this is a no-op.
+      * @param end The tag that terminates this composite tag, i.e. &lt;/HTML&gt;.
+      */
+     public void setEndTag (Tag end);
+ 
+     /**
+      * Return the scanner associated with this tag.
+      * @return The scanner associated with this tag.
+      */
+     public Scanner getThisScanner ();
+ 
+     /**
+      * Set the scanner associated with this tag.
+      * @param scanner The scanner for this tag.
+      */
+     public void setThisScanner (Scanner scanner);
+     
+     /**
+      * Get the line number where this tag starts.
+      * @return The (zero based) line number in the page where this tag starts.
+      */
+     public int getStartingLineNumber ();
+     /**
+      * Get the line number where this tag ends.
+      * @return The (zero based) line number in the page where this tag ends.
+      */
+     public int getEndingLineNumber ();
  }

Thread: [Htmlparser-cvs] htmlparser/src/org/htmlparser Node.java,1.50,1.51 PrototypicalNodeFactory.java,1.10

htmlparser-cvs