javaxmlsaxcore-media

Is it possible to merge XML-Elements with SAX (coremedia CAE filter)

Given is:

a XML structure like

<span class="abbreviation">AGB<span class"explanation">Allgemeine Geschäftsbedingungen</span></span>

and the result after the transformation should be:

<abbr title="Allgemeine Geschäftsbedingungen">AGB</abbr>

I know that SAX is an event-based XML-parser, and with methods like

I can capture events (like open-a-tag, close-a-tag) and with

I can extract the text between the tags.

My Question is:

Can i create a transformation mentioned above (is it possible)?

My Problem is:

Solution

  • The answer is yes it's possible!

    The main argument/hint you can get from this StackOverflow-link

    here is what has to be done:

    1. you have to remember the states, at which span-tag the sax parser is located ("class=abbreviation" or "class=explanation")
    2. you have to extract the content of the tags (this can be done with the #character method)
    3. When you know the state of the sax parser and the content, you can create a new abbr-tag
    4. all other tags, have to accede without any modification

    For completeness here is the source code of the coremedia cae filter:

    import com.coremedia.blueprint.cae.richtext.filter.FilterFactory;
import com.coremedia.xml.Filter;
import org.apache.commons.lang3.StringUtils;
import org.xml.sax.Attributes;
import org.xml.sax.SAXException;
import org.xml.sax.helpers.AttributesImpl;

import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;

public class GlossaryFilter extends Filter implements FilterFactory {
  private static final String SPAN = "span";
  private static final String CLASS = "class";

  private boolean isAbbreviation = false;
  private boolean isExplanation = false;
  private String abbreviation;
  private String currentUri;
  private boolean spanExplanationClose = false;
  private boolean spanAbbreviationClose = false;

  @Override
  public Filter getInstance(final HttpServletRequest request, final HttpServletResponse response) {
    return new GlossaryFilter();
  }

  @Override
  public void startElement(final String uri, final String localName, final String qName,
      final Attributes attributes) throws SAXException {
    if (isSpanAbbreviationTag(qName, attributes)) {
      isAbbreviation = true;
    } else if (isSpanExplanationTag(qName, attributes)) {
      isExplanation = true;
      currentUri = uri;
    } else {
      super.startElement(uri, localName, qName, attributes);
    }
  }

  private boolean isSpanExplanationTag(final String qName, final Attributes attributes) {
    //noinspection OverlyComplexBooleanExpression
    return StringUtils.isNotEmpty(qName) && qName.equalsIgnoreCase(SPAN) && (
        attributes.getLength() > 0) && attributes.getValue(CLASS).equals("explanation");
  }

  private boolean isSpanAbbreviationTag(final String qName, final Attributes attributes) {
    //noinspection OverlyComplexBooleanExpression
    return StringUtils.isNotEmpty(qName) && qName.equalsIgnoreCase(SPAN) && (
        attributes.getLength() > 0) && attributes.getValue(CLASS).equals("abbreviation");
  }

  @Override
  public void endElement(final String uri, final String localName, final String qName)
      throws SAXException {
    if (spanExplanationClose) {
      spanExplanationClose = false;
    } else if (spanAbbreviationClose) {
      spanAbbreviationClose = false;
    } else {
      super.endElement(uri, localName, qName);
    }
  }

  @Override
  public void characters(final char[] ch, final int start, final int length) throws SAXException {
    if (isAbbreviation && isExplanation) {
      final String explanation = new String(ch, start, length);
      final AttributesImpl newAttributes = createAttributes(explanation);
      writeAbbrTag(newAttributes);
      changeState();
    } else if (isAbbreviation && !isExplanation) {
      abbreviation = new String(ch, start, length);
    } else {
      super.characters(ch, start, length);
    }
  }

  private void changeState() {
    isExplanation = false;
    isAbbreviation = false;
    spanExplanationClose = true;
    spanAbbreviationClose = true;
  }

  @SuppressWarnings("TypeMayBeWeakened")
  private void writeAbbrTag(final AttributesImpl newAttributes) throws SAXException {
    super.startElement(currentUri, "abbr", "abbr", newAttributes);
    super.characters(abbreviation.toCharArray(), 0, abbreviation.length());
    super.endElement(currentUri, "abbr", "abbr");
  }

  private AttributesImpl createAttributes(final String explanation) {
    final AttributesImpl newAttributes = new AttributesImpl();
    newAttributes.addAttribute(currentUri, "title", "abbr:title", "CDATA", explanation);
    return newAttributes;
  }
}

    The interesting stuff is in the methods:

    startElement(...)

    Here you store the state at which tag the sax-parser is located (more detailed: you store the state, which span-tag (the "class=abbreviation" or "class=explanation") was opened.

    You only store states. The mentioned span-tags will not be processed/filtered (the result is, they would be removed). Every other tag is processed with no filtering, they will be applied without modification (that's the else-block).

    endElement(...)

    Here you want only process every tag except (the mentioned span-tags). All these tags are applied without modification (the else-block). If the sax parser is located at a closed span-tag (with "class=abbreviation" or "class=explanation") you want to do nothing (except store the state)

    characters(...)

    In this method the magic (creating a tag with the parser) happens. Depending on the state:

    1. Sax parser is located at a span-tag with "class=explanation" (this means there was an open span-tag with "class=abbreviation" passed before) --> branch (isAbbreviation && isExplanation)
    2. Sax parser is located at the first span-tag (the span-tag with "class=abbreviation") --> branch (isAbbreviation && !isExplanation)
    3. every other character you find in any other tag --> branch else

    for state 3.

    simply copy the text you find

    for state 2.

    extract the content of the span-tag with "class=abbreviation" for later use

    for state 3.