cziegeler 2003/07/12 06:49:56
Modified: lib jars.xml
src/blocks/portal/java/org/apache/cocoon/portal/transformation
RSSTransformer.java
Added: src/blocks/html/lib jtidy-04aug2000r7-dev.jar
Removed: src/blocks/html/lib .cvsignore
lib/optional jtidy-04aug2000r7-dev.jar
Log:
Removing direct dependency from portal block to jtidy
Revision Changes Path
1.3 +0 -0 cocoon-2.1/src/blocks/html/lib/jtidy-04aug2000r7-dev.jar
<<Binary file>>
1.65 +2 -2 cocoon-2.1/lib/jars.xml
Index: jars.xml
===================================================================
RCS file: /home/cvs/cocoon-2.1/lib/jars.xml,v
retrieving revision 1.64
retrieving revision 1.65
diff -u -r1.64 -r1.65
--- jars.xml 11 Jul 2003 20:25:30 -0000 1.64
+++ jars.xml 12 Jul 2003 13:49:56 -0000 1.65
@@ -496,7 +496,7 @@
<title>Transform HTML to XML</title>
<description>Tidy is a HTML syntax checker and pretty printer.</description>
<used-by>HTML generator (html block), RSSTransformer (Portal block)</used-by>
- <lib>optional/jtidy-04aug2000r7-dev.jar</lib>
+ <lib>html/lib/jtidy-04aug2000r7-dev.jar</lib>
<homepage>http://lempinen.net/sami/jtidy/</homepage>
</file>
1.2 +108 -34
cocoon-2.1/src/blocks/portal/java/org/apache/cocoon/portal/transformation/RSSTransformer.java
Index: RSSTransformer.java
===================================================================
RCS file:
/home/cvs/cocoon-2.1/src/blocks/portal/java/org/apache/cocoon/portal/transformation/RSSTransformer.java,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- RSSTransformer.java 11 Jul 2003 14:17:02 -0000 1.1
+++ RSSTransformer.java 12 Jul 2003 13:49:56 -0000 1.2
@@ -50,8 +50,22 @@
*/
package org.apache.cocoon.portal.transformation;
+import java.io.ByteArrayInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.util.Map;
+
+import org.apache.avalon.framework.component.Component;
+import org.apache.avalon.framework.component.ComponentException;
+import org.apache.avalon.framework.parameters.Parameters;
+import org.apache.cocoon.ProcessingException;
+import org.apache.cocoon.components.sax.XMLDeserializer;
+import org.apache.cocoon.components.sax.XMLSerializer;
+import org.apache.cocoon.environment.SourceResolver;
import org.apache.cocoon.transformation.AbstractSAXTransformer;
-import org.w3c.tidy.Tidy;
+import org.apache.cocoon.xml.IncludeXMLConsumer;
+import org.apache.cocoon.xml.XMLConsumer;
+import org.apache.excalibur.xmlizer.XMLizer;
import org.xml.sax.Attributes;
import org.xml.sax.SAXException;
@@ -66,6 +80,18 @@
public final class RSSTransformer
extends AbstractSAXTransformer {
+ /** The xmlizer for converting html to xml */
+ protected XMLizer xmlizer;
+
+ /** The xml serializer */
+ protected XMLSerializer serializer;
+
+ /** The xml deserializer */
+ protected XMLDeserializer deserializer;
+
+ /** The filter */
+ protected XMLConsumer filter;
+
/**
* receive notification of start element event.
**/
@@ -85,47 +111,95 @@
if ("description".equals(name)) {
final String text = this.endTextRecording();
final String html = "<html><body>"+text+"</body></html>";
-
+
+ boolean parsed = false;
try {
- final Tidy xhtmlconvert = new Tidy();
- xhtmlconvert.setXmlOut(true);
- xhtmlconvert.setXHTML(true);
- xhtmlconvert.setShowWarnings(false);
- org.w3c.dom.Document doc = xhtmlconvert.parseDOM(new
java.io.ByteArrayInputStream(html.getBytes()), null);
- org.w3c.dom.NodeList node =
org.apache.cocoon.xml.dom.DOMUtil.selectNodeList(doc, "/html/body/*");
- if (null != node) {
- for(int i = 0; i < node.getLength(); i++) {
- this.sendEvents(node.item(i));
- }
- } else {
- this.sendTextEvent(text);
- }
- } catch (Exception e) {
+ InputStream inputStream = new ByteArrayInputStream(html.getBytes());
+ this.xmlizer.toSAX(inputStream,
+ "text/html",
+ null,
+ this.serializer);
+ // if no exception occurs, everything is fine!
+ parsed = true;
+ } catch (Exception ignore) {
+ }
+ if ( parsed ) {
+ this.deserializer.setConsumer( this.filter );
+ this.deserializer.deserialize( this.serializer.getSAXFragment());
+ } else {
this.sendTextEvent(text);
}
}
super.endElement(uri,name,raw);
}
- /**
- * Replace occurence of searchString in source with replacement. If
- * replacement is null remove the occurences.
+ /* (non-Javadoc)
+ * @see org.apache.avalon.excalibur.pool.Recyclable#recycle()
+ */
+ public void recycle() {
+ this.manager.release( (Component) this.xmlizer );
+ this.manager.release( this.serializer );
+ this.manager.release( this.deserializer );
+ this.xmlizer = null;
+ this.serializer = null;
+ this.deserializer = null;
+ this.filter = null;
+ super.recycle();
+ }
+
+ /* (non-Javadoc)
+ * @see
org.apache.cocoon.sitemap.SitemapModelComponent#setup(org.apache.cocoon.environment.SourceResolver,
java.util.Map, java.lang.String, org.apache.avalon.framework.parameters.Parameters)
*/
- public static String replace(String source,
- String searchString,
- String replacement) {
- if (source != null && searchString != null) {
- int pos;
- int l = searchString.length();
- if (replacement == null) replacement = "";
- do {
- pos = source.indexOf(searchString);
- if (pos != -1) {
- source = source.substring(0, pos) + replacement +
source.substring(pos + l);
- }
- } while (pos != -1);
+ public void setup(SourceResolver resolver,
+ Map objectModel,
+ String src,
+ Parameters par)
+ throws ProcessingException, SAXException, IOException {
+ super.setup(resolver, objectModel, src, par);
+ try {
+ this.xmlizer = (XMLizer)this.manager.lookup(XMLizer.ROLE);
+ this.serializer =
(XMLSerializer)this.manager.lookup(XMLSerializer.ROLE);
+ this.deserializer =
(XMLDeserializer)this.manager.lookup(XMLDeserializer.ROLE);
+ } catch (ComponentException ce) {
+ throw new ProcessingException("Unable to lookup component.", ce);
}
- return source;
+ }
+
+ class HTMLFilter extends IncludeXMLConsumer {
+
+ int bodyCount = 0;
+
+ public HTMLFilter(XMLConsumer consumer) {
+ super(consumer);
+ }
+
+ public void startElement(String uri, String local, String qName, Attributes
attr) throws SAXException {
+ if (bodyCount > 0 ) {
+ super.startElement(uri, local, qName, attr);
+ }
+ if ("body".equalsIgnoreCase(local)) {
+ bodyCount++;
+ }
+ }
+
+ public void endElement(String uri, String local, String qName) throws
SAXException {
+ if ("body".equalsIgnoreCase(local)) {
+ bodyCount--;
+ }
+ if (bodyCount > 0 ) {
+ super.endElement(uri, local, qName );
+ }
+ }
+
+ }
+
+ /* (non-Javadoc)
+ * @see
org.apache.cocoon.transformation.AbstractSAXTransformer#setupTransforming()
+ */
+ public void setupTransforming()
+ throws IOException, ProcessingException, SAXException {
+ super.setupTransforming();
+ this.filter = new HTMLFilter( this.xmlConsumer );
}
}
