alvinalexander.com | career | drupal | java | mac | mysql | perl | scala | uml | unix  

JMeter example source code file (XPathExtractor.java)

This example JMeter source code file (XPathExtractor.java) is included in the DevDaily.com "Java Source Code Warehouse" project. The intent of this project is to help you "Learn Java by Example" TM.

Java - JMeter tags/keywords

assertionresult, booleanproperty, bytearrayinputstream, dom, io, ioexception, ioexception, list, node, non-nls-1, non-nls-1, parser, sampleresult, sax, string, string, transformerexception, transformerexception, util, xml

The JMeter XPathExtractor.java source code

/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 *
 */
package org.apache.jmeter.extractor;

import java.io.ByteArrayInputStream;
import java.io.IOException;
import java.io.Serializable;
import java.io.UnsupportedEncodingException;
import java.io.StringWriter;
import java.util.ArrayList;
import java.util.List;

import javax.xml.parsers.ParserConfigurationException;
import javax.xml.transform.TransformerException;

import org.apache.jmeter.assertions.AssertionResult;
import org.apache.jmeter.processor.PostProcessor;
import org.apache.jmeter.samplers.SampleResult;
import org.apache.jmeter.testelement.AbstractScopedTestElement;
import org.apache.jmeter.testelement.property.BooleanProperty;
import org.apache.jmeter.threads.JMeterContext;
import org.apache.jmeter.threads.JMeterVariables;
import org.apache.jmeter.util.TidyException;
import org.apache.jmeter.util.XPathUtil;
import org.apache.jorphan.logging.LoggingManager;
import org.apache.jorphan.util.JMeterError;
import org.apache.jorphan.util.JOrphanUtils;
import org.apache.log.Logger;
import org.apache.xpath.XPathAPI;
import org.apache.xpath.objects.XObject;
import org.w3c.dom.Document;
import org.w3c.dom.Element;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import org.xml.sax.SAXException;

import javax.xml.transform.OutputKeys;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.dom.DOMSource;
import javax.xml.transform.stream.StreamResult;

//@see org.apache.jmeter.extractor.TestXPathExtractor for unit tests

/**
 * Extracts text from (X)HTML response using XPath query language
 * Example XPath queries:
 * <dl>
 * <dt>/html/head/title
 *     <dd>extracts Title from HTML response
 * <dt>//form[@name='countryForm']//select[@name='country']/option[text()='Czech Republic'])/@value
 *     <dd>extracts value attribute of option element that match text 'Czech Republic'
 *                 inside of select element with name attribute  'country' inside of
 *                 form with name attribute 'countryForm'</dd>
 * <dt>//head
 *     <dd>extracts the XML fragment for head node.
 * <dt>//head/text()
 *     <dd>extracts the text content for head node.
 * </dl>
 */
 /* This file is inspired by RegexExtractor.
 * author <a href="mailto:hpaluch@gitus.cz">Henryk Paluch
 *            of <a href="http://www.gitus.com">Gitus a.s.
 *
 * See Bugzilla: 37183
 */
public class XPathExtractor extends AbstractScopedTestElement implements
        PostProcessor, Serializable {
    private static final Logger log = LoggingManager.getLoggerForClass();

    private static final long serialVersionUID = 240L;

    private static final String MATCH_NR = "matchNr"; // $NON-NLS-1$

    //+ JMX file attributes
    private static final String XPATH_QUERY     = "XPathExtractor.xpathQuery"; // $NON-NLS-1$
    private static final String REFNAME         = "XPathExtractor.refname"; // $NON-NLS-1$
    private static final String DEFAULT         = "XPathExtractor.default"; // $NON-NLS-1$
    private static final String TOLERANT        = "XPathExtractor.tolerant"; // $NON-NLS-1$
    private static final String NAMESPACE       = "XPathExtractor.namespace"; // $NON-NLS-1$
    private static final String QUIET           = "XPathExtractor.quiet"; // $NON-NLS-1$
    private static final String REPORT_ERRORS   = "XPathExtractor.report_errors"; // $NON-NLS-1$
    private static final String SHOW_WARNINGS   = "XPathExtractor.show_warnings"; // $NON-NLS-1$
    private static final String DOWNLOAD_DTDS   = "XPathExtractor.download_dtds"; // $NON-NLS-1$
    private static final String WHITESPACE      = "XPathExtractor.whitespace"; // $NON-NLS-1$
    private static final String VALIDATE        = "XPathExtractor.validate"; // $NON-NLS-1$
    private static final String FRAGMENT        = "XPathExtractor.fragment"; // $NON-NLS-1$
    //- JMX file attributes


    private String concat(String s1,String s2){
        return new StringBuilder(s1).append("_").append(s2).toString(); // $NON-NLS-1$
    }

    private String concat(String s1, int i){
        return new StringBuilder(s1).append("_").append(i).toString(); // $NON-NLS-1$
    }

    /**
     * Do the job - extract value from (X)HTML response using XPath Query.
     * Return value as variable defined by REFNAME. Returns DEFAULT value
     * if not found.
     */
    public void process() {
        JMeterContext context = getThreadContext();
        final SampleResult previousResult = context.getPreviousResult();
        if (previousResult == null){
            return;
        }
        JMeterVariables vars = context.getVariables();
        String refName = getRefName();
        vars.put(refName, getDefaultValue());
        final String matchNR = concat(refName,MATCH_NR);
        int prevCount=0; // number of previous matches
        try {
            prevCount=Integer.parseInt(vars.get(matchNR));
        } catch (NumberFormatException e) {
            // ignored
        }
        vars.put(matchNR, "0"); // In case parse fails // $NON-NLS-1$
        vars.remove(concat(refName,"1")); // In case parse fails // $NON-NLS-1$

        List<SampleResult> samples = getSampleList(previousResult);
        try{
            List<String> matches = new ArrayList();
            for (SampleResult res : samples) {
                Document d = parseResponse(res);
                getValuesForXPath(d,getXPathQuery(),matches);
            }
            final int matchCount = matches.size();
            vars.put(matchNR, String.valueOf(matchCount));
            if (matchCount > 0){
                String value = matches.get(0);
                if (value != null) {
                    vars.put(refName, value);
                }
                for(int i=0; i < matchCount; i++){
                    value = matches.get(i);
                    if (value != null) {
                        vars.put(concat(refName,i+1),matches.get(i));
                    }
                }
            }
            vars.remove(concat(refName,matchCount+1)); // Just in case
            // Clear any other remaining variables
            for(int i=matchCount+2; i <= prevCount; i++) {
                vars.remove(concat(refName,i));
            }
        }catch(IOException e){// e.g. DTD not reachable
            final String errorMessage = "IOException on ("+getXPathQuery()+")";
            log.error(errorMessage,e);
            AssertionResult ass = new AssertionResult(getName());
            ass.setError(true);
            ass.setFailureMessage(new StringBuilder("IOException: ").append(e.getLocalizedMessage()).toString());
            previousResult.addAssertionResult(ass);
            previousResult.setSuccessful(false);
        } catch (ParserConfigurationException e) {// Should not happen
            final String errrorMessage = "ParserConfigurationException while processing ("+getXPathQuery()+")";
            log.error(errrorMessage,e);
            throw new JMeterError(errrorMessage,e);
        } catch (SAXException e) {// Can happen for bad input document
            log.warn("SAXException while processing ("+getXPathQuery()+") "+e.getLocalizedMessage());
            addAssertionFailure(previousResult, e, false); // Should this also fail the sample?
        } catch (TransformerException e) {// Can happen for incorrect XPath expression
            log.warn("TransformerException while processing ("+getXPathQuery()+") "+e.getLocalizedMessage());
            addAssertionFailure(previousResult, e, false);
        } catch (TidyException e) {
            addAssertionFailure(previousResult, e, true); // fail the sample
        }
    }

    private void addAssertionFailure(final SampleResult previousResult,
            final Throwable thrown, final boolean setFailed) {
        AssertionResult ass = new AssertionResult(thrown.getClass().getSimpleName()); // $NON-NLS-1$
        ass.setFailure(true);
        ass.setFailureMessage(getXPathQuery()+" => "+thrown.getLocalizedMessage());
        previousResult.addAssertionResult(ass);
        if (setFailed){
            previousResult.setSuccessful(false);
        }
    }

    /*============= object properties ================*/
    public void setXPathQuery(String val){
        setProperty(XPATH_QUERY,val);
    }

    public String getXPathQuery(){
        return getPropertyAsString(XPATH_QUERY);
    }

    public void setRefName(String refName) {
        setProperty(REFNAME, refName);
    }

    public String getRefName() {
        return getPropertyAsString(REFNAME);
    }

    public void setDefaultValue(String val) {
        setProperty(DEFAULT, val);
    }

    public String getDefaultValue() {
        return getPropertyAsString(DEFAULT);
    }

    public void setTolerant(boolean val) {
        setProperty(new BooleanProperty(TOLERANT, val));
    }

    public boolean isTolerant() {
        return getPropertyAsBoolean(TOLERANT);
    }

    public void setNameSpace(boolean val) {
        setProperty(new BooleanProperty(NAMESPACE, val));
    }

    public boolean useNameSpace() {
        return getPropertyAsBoolean(NAMESPACE);
    }

    public void setReportErrors(boolean val) {
            setProperty(REPORT_ERRORS, val, false);
    }

    public boolean reportErrors() {
        return getPropertyAsBoolean(REPORT_ERRORS, false);
    }

    public void setShowWarnings(boolean val) {
        setProperty(SHOW_WARNINGS, val, false);
    }

    public boolean showWarnings() {
        return getPropertyAsBoolean(SHOW_WARNINGS, false);
    }

    public void setQuiet(boolean val) {
        setProperty(QUIET, val, true);
    }

    public boolean isQuiet() {
        return getPropertyAsBoolean(QUIET, true);
    }

    /**
     * Should we return fragment as text, rather than text of fragment?
     * @return true if we should return fragment rather than text
     */
    public boolean getFragment() {
        return getPropertyAsBoolean(FRAGMENT, false);
    }

    /**
     * Should we return fragment as text, rather than text of fragment?
     * @param selected true to return fragment.
     */
    public void setFragment(boolean selected) {
        setProperty(FRAGMENT, selected, false);
    }

    /*================= internal business =================*/
    /**
     * Converts (X)HTML response to DOM object Tree.
     * This version cares of charset of response.
     * @param result
     * @return
     *
     */
    private Document parseResponse(SampleResult result)
      throws UnsupportedEncodingException, IOException, ParserConfigurationException,SAXException,TidyException
    {
      //TODO: validate contentType for reasonable types?

      // NOTE: responseData encoding is server specific
      //       Therefore we do byte -> unicode -> byte conversion
      //       to ensure UTF-8 encoding as required by XPathUtil
      String unicodeData = result.getResponseDataAsString();
      // convert unicode String -> UTF-8 bytes
      byte[] utf8data = unicodeData.getBytes("UTF-8"); // $NON-NLS-1$
      ByteArrayInputStream in = new ByteArrayInputStream(utf8data);
      boolean isXML = JOrphanUtils.isXML(utf8data);
      // this method assumes UTF-8 input data
      return XPathUtil.makeDocument(in,false,false,useNameSpace(),isTolerant(),isQuiet(),showWarnings(),reportErrors()
              ,isXML, isDownloadDTDs());
    }

    /**
     * Extract value from Document d by XPath query.
     * @param d the document
     * @param query the query to execute
     * @param matchStrings list of matched strings (may include nulls)
     *
     * @throws TransformerException
     */
    private void getValuesForXPath(Document d,String query, List<String> matchStrings)
        throws TransformerException {
        String val = null;
        XObject xObject = XPathAPI.eval(d, query);
        final int objectType = xObject.getType();
        if (objectType == XObject.CLASS_NODESET) {
            NodeList matches = xObject.nodelist();
            int length = matches.getLength();
            for (int i = 0 ; i < length; i++) {
                Node match = matches.item(i);
                if ( match instanceof Element){
                    if (getFragment()){
                        val = getValueForNode(match);
                    } else {
                        // elements have empty nodeValue, but we are usually interested in their content
                        final Node firstChild = match.getFirstChild();
                        if (firstChild != null) {
                            val = firstChild.getNodeValue();
                        } else {
                            val = match.getNodeValue(); // TODO is this correct?
                        }
                    }
                } else {
                   val = match.getNodeValue();
                }
                matchStrings.add(val);
            }
        } else if (objectType == XObject.CLASS_NULL
                || objectType == XObject.CLASS_UNKNOWN
                || objectType == XObject.CLASS_UNRESOLVEDVARIABLE) {
            log.warn("Unexpected object type: "+xObject.getTypeString()+" returned for: "+getXPathQuery());
        } else {
            val = xObject.toString();
            matchStrings.add(val);
      }
    }

    public void setWhitespace(boolean selected) {
        setProperty(WHITESPACE, selected, false);
    }

    public boolean isWhitespace() {
        return getPropertyAsBoolean(WHITESPACE, false);
    }

    public void setValidating(boolean selected) {
        setProperty(VALIDATE, selected);
    }

    public boolean isValidating() {
        return getPropertyAsBoolean(VALIDATE, false);
    }

    public void setDownloadDTDs(boolean selected) {
        setProperty(DOWNLOAD_DTDS, selected, false);
    }

    public boolean isDownloadDTDs() {
        return getPropertyAsBoolean(DOWNLOAD_DTDS, false);
    }

    private String getValueForNode(Node node) {
        StringWriter sw = new StringWriter();
        try {
            Transformer t = TransformerFactory.newInstance().newTransformer();
            t.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
            t.transform(new DOMSource(node), new StreamResult(sw));
        } catch (TransformerException e) {
            sw.write(e.getMessageAndLocation());
        }
        return sw.toString();
    }
}

Other JMeter examples (source code examples)

Here is a short list of links related to this JMeter XPathExtractor.java source code file:

... this post is sponsored by my books ...

#1 New Release!

FP Best Seller

 

new blog posts

 

Copyright 1998-2024 Alvin Alexander, alvinalexander.com
All Rights Reserved.

A percentage of advertising revenue from
pages under the /java/jwarehouse URI on this website is
paid back to open source projects.