[Httpunit-commit] CVS: httpunit/src/com/meterware/httpunit HTMLParser.java,NONE,1.1 JTidyHTMLParser.java,NONE,1.1 NekoHTMLParser.java,NONE,1.1 FormControl.java,1.28,1.29 HTMLPage.java,1.12,1.13 HttpUnitOptions.java,1.29,1.30 NodeUtils.java,1.12,1.13 ParsedHTML.java,1.30,1.31 WebLink.java,1.32,1.33 WebResponse.java,1.91,1.92

[Russell G. <rus...@us...>]

Update of /cvsroot/httpunit/httpunit/src/com/meterware/httpunit
In directory usw-pr-cvs1:/tmp/cvs-serv10393/src/com/meterware/httpunit

Modified Files:
	FormControl.java HTMLPage.java HttpUnitOptions.java 
	NodeUtils.java ParsedHTML.java WebLink.java WebResponse.java 
Added Files:
	HTMLParser.java JTidyHTMLParser.java NekoHTMLParser.java 
Log Message:
Added NekoHTML support

***** Error reading new file[Errno 2] No such file or directory: 'HTMLParser.java'
***** Error reading new file[Errno 2] No such file or directory: 'JTidyHTMLParser.java'
***** Error reading new file[Errno 2] No such file or directory: 'NekoHTMLParser.java'
Index: FormControl.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/FormControl.java,v
retrieving revision 1.28
retrieving revision 1.29
diff -u -r1.28 -r1.29
--- FormControl.java	13 Oct 2002 23:53:48 -0000	1.28
+++ FormControl.java	14 Oct 2002 04:19:06 -0000	1.29
@@ -272,9 +272,9 @@
     static FormControl newFormParameter( WebForm form, Node node ) {
         if (node.getNodeType() != Node.ELEMENT_NODE) {
             return null;
-        } else if (node.getNodeName().equals( "textarea" )) {
+        } else if (node.getNodeName().equalsIgnoreCase( "textarea" )) {
             return new TextAreaFormControl( form, node );
-        } else if (node.getNodeName().equals( "select" )) {
+        } else if (node.getNodeName().equalsIgnoreCase( "select" )) {
             return new SelectionFormControl( form, node );
         } else if (node.getNodeName().equalsIgnoreCase( "button" )) {
             final String type = NodeUtils.getNodeAttribute( node, "type", "submit" );
@@ -285,7 +285,7 @@
             } else {
                 return new Button( form, node );
             }
-        } else if (!node.getNodeName().equals( "input" )) {
+        } else if (!node.getNodeName().equalsIgnoreCase( "input" )) {
             return null;
         } else {
             final String type = NodeUtils.getNodeAttribute( node, "type", "text" );
@@ -991,8 +991,9 @@
 
             _options = new Option[ nl.getLength() ];
             for (int i = 0; i < _options.length; i++) {
-                _options[i] = new Option( getValue( nl.item(i).getFirstChild() ),
-                                          getOptionValue( nl.item(i) ),
+                final String displayedText = getValue( nl.item(i).getFirstChild() ).trim();
+                _options[i] = new Option( displayedText,
+                                          getOptionValue( nl.item(i), displayedText ),
                                           nl.item(i).getAttributes().getNamedItem( "selected" ) != null );
                 _options[i].setIndex( this, i );
             }
@@ -1136,12 +1137,12 @@
         }
 
 
-        private String getOptionValue( Node optionNode ) {
+        private String getOptionValue( Node optionNode, String displayedText ) {
             NamedNodeMap nnm = optionNode.getAttributes();
             if (nnm.getNamedItem( "value" ) != null) {
                 return getValue( nnm.getNamedItem( "value" ) );
             } else {
-                return getValue( optionNode.getFirstChild() );
+                return displayedText;
             }
         }
 

Index: HTMLPage.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/HTMLPage.java,v
retrieving revision 1.12
retrieving revision 1.13
diff -u -r1.12 -r1.13
--- HTMLPage.java	13 Oct 2002 23:53:48 -0000	1.12
+++ HTMLPage.java	14 Oct 2002 04:19:06 -0000	1.13
@@ -21,22 +21,18 @@
 *******************************************************************************************************************/
 import com.meterware.httpunit.scripting.NamedDelegate;
 import com.meterware.httpunit.scripting.ScriptableDelegate;
-
-import java.io.ByteArrayInputStream;
-import java.io.IOException;
-import java.io.UnsupportedEncodingException;
-import java.net.MalformedURLException;
-import java.net.URL;
-import java.util.Vector;
-import java.util.ArrayList;
-
 import org.w3c.dom.Document;
 import org.w3c.dom.Element;
 import org.w3c.dom.Node;
 import org.w3c.dom.NodeList;
-import org.w3c.tidy.Tidy;
 import org.xml.sax.SAXException;
 
+import java.io.IOException;
+import java.net.MalformedURLException;
+import java.net.URL;
+import java.util.ArrayList;
+import java.util.Vector;
+
 
 /**
  * This class represents an HTML page returned from a request.
@@ -49,7 +45,7 @@
     private Scriptable _scriptable;
 
 
-    public HTMLPage( WebResponse response, URL url, String parentTarget, String pageText, String characterSet ) throws SAXException {
+    HTMLPage( WebResponse response, URL url, String parentTarget, String pageText, String characterSet ) throws IOException, SAXException {
         super( response, url, parentTarget, getDOM( url, pageText ), characterSet );
         setBaseAttributes();
     }
@@ -239,24 +235,8 @@
 //---------------------------------- private members --------------------------------
 
 
-    private static Node getDOM( URL url, String pageText ) throws SAXException {
-        try {
-            return getParser( url ).parseDOM( new ByteArrayInputStream( pageText.getBytes( getUTFEncodingName() ) ), null );
-        } catch (UnsupportedEncodingException e) {
-            throw new RuntimeException( "UTF-8 encoding failed" );
-        }
-    }
-
-
-    private static String _utfEncodingName;
-
-    private static String getUTFEncodingName() {
-        if (_utfEncodingName == null) {
-            String versionNum = System.getProperty( "java.version" );
-            if (versionNum.startsWith( "1.1" )) _utfEncodingName = "UTF8";
-            else _utfEncodingName = "UTF-8";
-        }
-        return _utfEncodingName;
+    private static Node getDOM( URL url, String pageText ) throws IOException, SAXException {
+        return HttpUnitOptions.getHTMLParser().getDocument( url, pageText );
     }
 
 
@@ -281,16 +261,5 @@
         }
     }
 
-
-    private static Tidy getParser( URL url ) {
-        Tidy tidy = new Tidy();
-        tidy.setCharEncoding( org.w3c.tidy.Configuration.UTF8 );
-        tidy.setQuiet( true );
-        tidy.setShowWarnings( HttpUnitOptions.getParserWarningsEnabled() );
-        if (!HttpUnitOptions.getHtmlErrorListeners().isEmpty()) {
-            tidy.setErrout(new JTidyPrintWriter( url ));
-        }
-        return tidy;
-    }
 
 }

Index: HttpUnitOptions.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/HttpUnitOptions.java,v
retrieving revision 1.29
retrieving revision 1.30
diff -u -r1.29 -r1.30
--- HttpUnitOptions.java	9 Oct 2002 13:50:07 -0000	1.29
+++ HttpUnitOptions.java	14 Oct 2002 04:19:06 -0000	1.30
@@ -27,7 +27,7 @@
 /**
  * A collection of global options to control HttpUnit's behavior.
  *
- * @author <a href="mailto:rus...@ac...">Russell Gold</a>
+ * @author <a href="mailto:rus...@ht...">Russell Gold</a>
  * @author <a href="mailto:dg...@ss...">Dave Glowacki</a>
  * @author <a href="mailto:bx...@bi...">Benoit Xhenseval</a>
  **/
@@ -61,6 +61,11 @@
     }
 
 
+    public static HTMLParser getHTMLParser() {
+        return _htmlParser;
+    }
+
+
     /**
      * Returns true if HttpUnit is accepting and saving cookies. The default is to accept them.
      */
@@ -506,10 +511,30 @@
 
     private static boolean _exceptionsThrownOnScriptError = true;
 
+    private static HTMLParser _htmlParser;
 
+    private static HTMLParser _jtidyParser;
+
+    private static HTMLParser _nekoParser;
 
     static {
         _listeners = new Vector();
         reset();
+
+        try {
+            _jtidyParser = (HTMLParser) Class.forName( "com.meterware.httpunit.JTidyHTMLParser" ).newInstance();
+        } catch (InstantiationException e) {
+        } catch (IllegalAccessException e) {
+        } catch (ClassNotFoundException e) {
+        }
+
+        try {
+            _nekoParser = (HTMLParser) Class.forName( "com.meterware.httpunit.NekoHTMLParser" ).newInstance();
+        } catch (InstantiationException e) {
+        } catch (IllegalAccessException e) {
+        } catch (ClassNotFoundException e) {
+        }
+
+        _htmlParser = (_nekoParser == null) ? _jtidyParser : _nekoParser;
     }
 }

Index: NodeUtils.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/NodeUtils.java,v
retrieving revision 1.12
retrieving revision 1.13
diff -u -r1.12 -r1.13
--- NodeUtils.java	5 Aug 2002 15:21:19 -0000	1.12
+++ NodeUtils.java	14 Oct 2002 04:19:06 -0000	1.13
@@ -109,7 +109,7 @@
                 return true;
             }
             public void processTextNodeValue( String value ) {
-                sb.append( convertNBSP( value ) );
+                sb.append( HttpUnitOptions.getHTMLParser().getCleanedText( value ) );
             }
         } );
         return sb.toString();
@@ -136,15 +136,6 @@
 
             pushNodeList( node.getChildNodes(), pendingNodes );
         }
-    }
-
-
-    final static private char NBSP = (char) 160;   // non-breaking space, defined by JTidy
-
-
-    private static String convertNBSP( String text ) {
-        if (text == null) return "";
-        return text.replace( NBSP, ' ' );
     }
 
 

Index: ParsedHTML.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/ParsedHTML.java,v
retrieving revision 1.30
retrieving revision 1.31
diff -u -r1.30 -r1.31
--- ParsedHTML.java	4 Oct 2002 02:16:00 -0000	1.30
+++ ParsedHTML.java	14 Oct 2002 04:19:06 -0000	1.31
@@ -27,7 +27,7 @@
 import org.w3c.dom.*;
 
 /**
- * @author <a href="mailto:rus...@ac...">Russell Gold</a>
+ * @author <a href="mailto:rus...@ht...">Russell Gold</a>
  * @author <a href="mailto:bx...@bi...">Benoit Xhenseval</a>
  **/
 class ParsedHTML {
@@ -375,7 +375,7 @@
     private boolean isLinkAnchor( Node node ) {
         if (node.getNodeType() != Node.ELEMENT_NODE) {
             return false;
-        } else if (!node.getNodeName().equals( "a" ) && !node.getNodeName().equals( "area" )) {
+        } else if (!node.getNodeName().equalsIgnoreCase( "a" ) && !node.getNodeName().equalsIgnoreCase( "area" )) {
             return false;
         } else {
             return (node.getAttributes().getNamedItem( "href" ) != null);

Index: WebLink.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/WebLink.java,v
retrieving revision 1.32
retrieving revision 1.33
diff -u -r1.32 -r1.33
--- WebLink.java	13 Oct 2002 23:53:48 -0000	1.32
+++ WebLink.java	14 Oct 2002 04:19:06 -0000	1.33
@@ -66,7 +66,7 @@
      * Returns the text value of this link.
      **/
     public String asText() {
-        if (getNode().getNodeName().equals( "area" )) {
+        if (getNode().getNodeName().equalsIgnoreCase( "area" )) {
             return getAttribute( "alt" );
         } else if (!getNode().hasChildNodes()) {
             return "";

Index: WebResponse.java
===================================================================
RCS file: /cvsroot/httpunit/httpunit/src/com/meterware/httpunit/WebResponse.java,v
retrieving revision 1.91
retrieving revision 1.92
diff -u -r1.91 -r1.92
--- WebResponse.java	13 Oct 2002 23:53:48 -0000	1.91
+++ WebResponse.java	14 Oct 2002 04:19:06 -0000	1.92
@@ -39,7 +39,7 @@
 /**
  * A response to a web request from a web server.
  *
- * @author <a href="mailto:rus...@ac...">Russell Gold</a>
+ * @author <a href="mailto:rus...@ht...">Russell Gold</a>
  * @author <a href="mailto:DRE...@or...">Drew Varner</a>
  * @author <a href="mailto:dg...@ss...">Dave Glowacki</a>
  * @author <a href="mailto:bx...@bi...">Benoit Xhenseval</a>
@@ -1088,7 +1088,8 @@
                 if (!isHTML()) throw new NotHTMLException( getContentType() );
                 _page = new HTMLPage( this, _url, _frameName, getText(), getCharacterSet() );
             } catch (IOException e) {
-                throw new SAXException( e );
+                e.printStackTrace();
+                throw new RuntimeException( e.toString() );
             }
         }
         return _page;