user/src/com/google/gwt/benchmarks/BenchmarkReport.java - gwt - Git at Google

 /*
  * Copyright 2008 Google Inc.
  *
  * Licensed 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 com.google.gwt.benchmarks;

 import com.google.gwt.benchmarks.client.impl.BenchmarkResults;
 import com.google.gwt.benchmarks.client.impl.Trial;
 import com.google.gwt.benchmarks.rebind.BenchmarkGenerator;
 import com.google.gwt.core.ext.TreeLogger;
 import com.google.gwt.core.ext.typeinfo.HasAnnotations;
 import com.google.gwt.core.ext.typeinfo.HasMetaData;
 import com.google.gwt.core.ext.typeinfo.JClassType;
 import com.google.gwt.core.ext.typeinfo.JMethod;
 import com.google.gwt.core.ext.typeinfo.TypeOracle;
 import com.google.gwt.dev.util.Util;
 import com.google.gwt.util.tools.Utility;

 import junit.framework.TestCase;

 import org.w3c.dom.Document;
 import org.w3c.dom.Element;

 import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.text.BreakIterator;
 import java.text.DateFormat;
 import java.util.ArrayList;
 import java.util.Date;
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;

 import javax.xml.parsers.DocumentBuilder;
 import javax.xml.parsers.DocumentBuilderFactory;
 import javax.xml.parsers.ParserConfigurationException;

 /**
  * Generates a detailed report that contains the results of all of the
  * benchmark-related unit tests executed during a unit test session. The primary
  * user of this class is {@link com.google.gwt.junit.JUnitShell}.
  *
  * The report is in XML format. To view the XML reports, use benchmarkViewer.
  */
 public class BenchmarkReport {

   /**
    * Converts a set of test results for a single benchmark method into XML.
    */
   private class BenchmarkXml {

     private MetaData metaData;

     private List<BenchmarkResults> results;

     private TestCase test;

     BenchmarkXml(TestCase test, List<BenchmarkResults> results) {
       this.test = test;
       this.results = results;
       Map<String, MetaData> methodMetaData = testMetaData.get(test.getClass().toString());
       metaData = methodMetaData.get(test.getName());
     }

     Element toElement(Document doc) {
       Element benchmark = doc.createElement("benchmark");
       benchmark.setAttribute("class", test.getClass().getName());
       benchmark.setAttribute("name", metaData.getTestName());
       benchmark.setAttribute("description", metaData.getTestDescription());

       String sourceCode = metaData.getSourceCode();
       if (sourceCode != null) {
         Element sourceCodeElement = doc.createElement("source_code");
         sourceCodeElement.appendChild(doc.createTextNode(sourceCode));
         benchmark.appendChild(sourceCodeElement);
       }

       // TODO(tobyr): create target_code element

       for (BenchmarkResults result : results) {
         benchmark.appendChild(toElement(doc, result));
       }

       return benchmark;
     }

     private Element toElement(Document doc, BenchmarkResults result) {
       Element resultElement = doc.createElement("result");
       resultElement.setAttribute("host", result.getHost());
       resultElement.setAttribute("agent", result.getAgent());

       Throwable exception = result.getException();

       if (exception != null) {
         Element exceptionElement = doc.createElement("exception");
         exceptionElement.appendChild(doc.createTextNode(exception.toString()));
         resultElement.appendChild(exceptionElement);
       }

       List<Trial> trials = result.getTrials();

       for (Trial trial : trials) {
         Element trialElement = toElement(doc, trial);
         resultElement.appendChild(trialElement);
       }

       return resultElement;
     }

     private Element toElement(Document doc, Trial trial) {
       Element trialElement = doc.createElement("trial");

       Map<String, String> variables = trial.getVariables();

       for (Map.Entry<String, String> entry : variables.entrySet()) {
         Object name = entry.getKey();
         Object value = entry.getValue();
         Element variableElement = doc.createElement("variable");
         variableElement.setAttribute("name", name.toString());
         variableElement.setAttribute("value", value.toString());
         trialElement.appendChild(variableElement);
       }

       trialElement.setAttribute("timing",
           String.valueOf(trial.getRunTimeMillis()));

       return trialElement;
     }
   }

   /**
    * Parses .java source files to get source code for methods.
    */
   private class Parser {

     /**
      * Maps classes to the contents of their source files.
      */
     private Map<JClassType, String> classSources = new HashMap<JClassType, String>();

     /**
      * Returns the source code for the method of the given name.
      *
      * @param logger to log the process
      * @param method a not <code>null</code> method
      * @return <code>null</code> if the source code for the method can not be
      *         located
      */
     public String getMethod(TreeLogger logger, JMethod method) {
       JClassType clazz = method.getEnclosingType();

       if (!classSources.containsKey(clazz)) {
         char[] sourceContents = null;
         File sourceFile = findSourceFile(clazz);
         if (sourceFile != null) {
           sourceContents = Util.readFileAsChars(sourceFile);
           classSources.put(clazz, new String(sourceContents));
         }

         if (sourceContents == null) {
           classSources.put(clazz, null);
           String msg = "An unknown I/O exception occured while trying to read "
               + sourceFile.getAbsolutePath();
           logger.log(TreeLogger.WARN, msg, null);
         } else {
           classSources.put(clazz, new String(sourceContents));
           String msg = "BenchmarkReport read the contents of "
               + sourceFile.getAbsolutePath();
           TreeLogger branch = logger.branch(TreeLogger.DEBUG, msg, null);
           if (logger.isLoggable(TreeLogger.SPAM)) {
             branch.log(TreeLogger.SPAM, new String(sourceContents), null);
           }
         }
       }

       String source = classSources.get(clazz);

       if (source == null) {
         return source;
       }

       try {
         return source.substring(method.getDeclStart(), method.getDeclEnd() + 1);
       } catch (IndexOutOfBoundsException e) {
         logger.log(TreeLogger.WARN, "Unable to parse " + method.getName(), e);
         // Have seen this happen when the compiler read the source using one
         // character encoding, and then this Parser read it in a different
         // encoding. I don't know if there are other cases in which this can
         // occur.
         return null;
       }
     }
   }

   /**
    * Converts an entire report into XML.
    */
   private class ReportXml {

     private Map<String, Element> categoryElementMap = new HashMap<String, Element>();

     private Date date = new Date();

     private String version = "unknown";

     Element toElement(Document doc) {
       Element report = doc.createElement("gwt_benchmark_report");
       String dateString = DateFormat.getDateTimeInstance().format(date);
       report.setAttribute("date", dateString);
       report.setAttribute("gwt_version", version);

       // Add each test result into the report.
       // Add the category for the test result, if necessary.
       for (Map.Entry<TestCase, List<BenchmarkResults>> entry : testResults.entrySet()) {
         TestCase test = entry.getKey();
         List<BenchmarkResults> results = entry.getValue();
         BenchmarkXml xml = new BenchmarkXml(test, results);
         Element categoryElement = getCategoryElement(doc, report,
             xml.metaData.getCategory().getClassName());
         categoryElement.appendChild(xml.toElement(doc));
       }

       return report;
     }

     /**
      * Locates or creates the category element by the specified name.
      *
      * @param doc The document to search
      * @param report The report to which the category belongs
      * @param name The name of the category
      *
      * @return The matching category element
      */
     private Element getCategoryElement(Document doc, Element report, String name) {
       Element e = categoryElementMap.get(name);

       if (e != null) {
         return e;
       }

       Element categoryElement = doc.createElement("category");
       categoryElementMap.put(name, categoryElement);
       CategoryImpl category = testCategories.get(name);
       categoryElement.setAttribute("name", category.getName());
       categoryElement.setAttribute("description", category.getDescription());

       report.appendChild(categoryElement);

       return categoryElement;
     }
   }

   private static final String GWT_BENCHMARK_CATEGORY = "gwt.benchmark.category";

   private static final String GWT_BENCHMARK_DESCRIPTION = "gwt.benchmark.description";

   private static final String GWT_BENCHMARK_NAME = "gwt.benchmark.name";

   private static File findSourceFile(JClassType clazz) {
     final char separator = File.separator.charAt(0);
     String filePath = clazz.getPackage().getName().replace('.', separator)
         + separator + clazz.getSimpleSourceName() + ".java";
     String[] paths = getClassPath();

     for (int i = 0; i < paths.length; ++i) {
       File maybeSourceFile = new File(paths[i] + separator + filePath);

       if (maybeSourceFile.exists()) {
         return maybeSourceFile;
       }
     }

     return null;
   }

   private static String[] getClassPath() {
     String path = System.getProperty("java.class.path");
     return path.split(File.pathSeparator);
   }

   private static String getSimpleMetaData(HasMetaData hasMetaData, String name) {
     String[][] allValues = hasMetaData.getMetaData(name);

     if (allValues == null) {
       return null;
     }

     StringBuffer result = new StringBuffer();

     for (int i = 0; i < allValues.length; ++i) {
       String[] values = allValues[i];
       for (int j = 0; j < values.length; ++j) {
         result.append(values[j]);
         result.append(" ");
       }
     }

     String resultString = result.toString().trim();
     return resultString.equals("") ? null : resultString;
   }

   private Parser parser = new Parser();

   private Map<String, CategoryImpl> testCategories = new HashMap<String, CategoryImpl>();

   private Map<String, Map<String, MetaData>> testMetaData = new HashMap<String, Map<String, MetaData>>();

   private Map<TestCase, List<BenchmarkResults>> testResults = new HashMap<TestCase, List<BenchmarkResults>>();

   private TypeOracle typeOracle;

   public BenchmarkReport() {
   }

   /**
    * Adds the Benchmark to the report. All of the metadata about the benchmark
    * (category, name, description, etc...) is recorded from the TypeOracle.
    *
    * @param logger to log the process
    * @param deprecationBranch to log usages of old-style metadata
    * @param benchmarkClass The benchmark class to record. Must not be
    *          <code>null</code>.
    * @param typeOracle The <code>TypeOracle</code> for the compilation session
    *          must not be <code>null</code>.
    */
   public void addBenchmark(TreeLogger logger, TreeLogger deprecationBranch,
       JClassType benchmarkClass, TypeOracle typeOracle) {

     this.typeOracle = typeOracle;
     String categoryType = getBenchmarkCategory(deprecationBranch,
         benchmarkClass);

     Map<String, JMethod> zeroArgMethods = BenchmarkGenerator.getNotOverloadedTestMethods(benchmarkClass);
     Map<String, JMethod> parameterizedMethods = BenchmarkGenerator.getParameterizedTestMethods(
         benchmarkClass, TreeLogger.NULL);
     List<JMethod> testMethods = new ArrayList<JMethod>(zeroArgMethods.size()
         + parameterizedMethods.size());
     testMethods.addAll(zeroArgMethods.values());
     testMethods.addAll(parameterizedMethods.values());

     Map<String, MetaData> metaDataMap = testMetaData.get(benchmarkClass.toString());
     if (metaDataMap == null) {
       metaDataMap = new HashMap<String, MetaData>();
       testMetaData.put(benchmarkClass.toString(), metaDataMap);
     }

     // Add all of the benchmark methods
     for (JMethod method : testMethods) {
       String methodName = method.getName();
       String methodCategoryType = getBenchmarkCategory(deprecationBranch,
           method);
       if (methodCategoryType == null) {
         methodCategoryType = categoryType;
       }
       CategoryImpl methodCategory = getCategory(deprecationBranch,
           methodCategoryType);

       String methodSource = parser.getMethod(logger, method);
       StringBuffer sourceBuffer = (methodSource == null) ? null
           : new StringBuffer(methodSource);
       StringBuffer summary = new StringBuffer();
       StringBuffer comment = new StringBuffer();
       getComment(sourceBuffer, summary, comment);

       MetaData metaData = new MetaData(benchmarkClass.toString(), methodName,
           (sourceBuffer != null) ? sourceBuffer.toString() : null,
           methodCategory, methodName, summary.toString());
       metaDataMap.put(methodName, metaData);
     }
   }

   public void addBenchmarkResults(TestCase test, BenchmarkResults results) {
     List<BenchmarkResults> currentResults = testResults.get(test);
     if (currentResults == null) {
       currentResults = new ArrayList<BenchmarkResults>();
       testResults.put(test, currentResults);
     }
     currentResults.add(results);
   }

   /**
    * Generates reports for all of the benchmarks which were added to the
    * generator.
    *
    * @param outputPath The path to write the reports to.
    * @throws ParserConfigurationException If an error occurs during xml parsing
    * @throws IOException If anything goes wrong writing to outputPath
    */
   public void generate(String outputPath) throws ParserConfigurationException,
       IOException {

     // Don't generate a new report if no tests were actually run.
     if (testResults.size() == 0) {
       return;
     }

     DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
     DocumentBuilder builder = factory.newDocumentBuilder();
     Document doc = builder.newDocument();
     doc.appendChild(new ReportXml().toElement(doc));
     byte[] xmlBytes = Util.toXmlUtf8(doc);
     FileOutputStream fos = null;
     try {
       fos = new FileOutputStream(outputPath);
       fos.write(xmlBytes);
     } finally {
       Utility.close(fos);
     }

     // TODO(bruce) The code below is commented out because of GWT Issue 958.

     // // TODO(tobyr) Looks like indenting is busted
     // // http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6296446
     // // Not a big deal, since we don't intend to read the XML by hand anyway
     // TransformerFactory transformerFactory = TransformerFactory.newInstance();
     // // Think this can be used with JDK 1.5
     // // transformerFactory.setAttribute( "indent-number", new Integer(2) );
     // Transformer serializer = transformerFactory.newTransformer();
     // serializer.setOutputProperty(OutputKeys.METHOD, "xml");
     // serializer.setOutputProperty(OutputKeys.INDENT, "yes");
     // serializer
     // .setOutputProperty("{ http://xml.apache.org/xslt }indent-amount", "2");
     // BufferedOutputStream docOut = new BufferedOutputStream(
     // new FileOutputStream(outputPath));
     // serializer.transform(new DOMSource(doc), new StreamResult(docOut));
     // docOut.close();
   }

   private <T extends HasMetaData & HasAnnotations> String getBenchmarkCategory(
       TreeLogger deprecationBranch, T element) {
     String category = getSimpleMetaData(element, GWT_BENCHMARK_CATEGORY);
     if (category != null) {
       deprecationBranch.log(TreeLogger.WARN, GWT_BENCHMARK_CATEGORY + " has "
           + "been deprecated with no replacement.", null);
     }
     return category;
   }

   private CategoryImpl getCategory(TreeLogger deprecationBranch, String name) {
     CategoryImpl c = testCategories.get(name);

     if (c != null) {
       return c;
     }

     c = getCategoryMetaData(deprecationBranch, name);
     testCategories.put(name, c);
     return c;
   }

   private CategoryImpl getCategoryMetaData(TreeLogger deprecationBranch,
       String typeName) {
     if (typeName == null) {
       return new CategoryImpl(null, "", "");
     }

     JClassType categoryType = typeOracle.findType(typeName);

     if (categoryType == null) {
       return new CategoryImpl(typeName, "", "");
     }

     String categoryName = getSimpleMetaData(categoryType, GWT_BENCHMARK_NAME);
     String categoryDescription = getSimpleMetaData(categoryType,
         GWT_BENCHMARK_DESCRIPTION);

     if (categoryName != null || categoryDescription != null) {
       deprecationBranch.log(TreeLogger.WARN, GWT_BENCHMARK_NAME + " and "
           + GWT_BENCHMARK_DESCRIPTION + " have been deprecated with no "
           + "replacement", null);
     }

     categoryName = categoryName == null ? "" : categoryName;
     categoryDescription = categoryDescription == null ? ""
         : categoryDescription;

     return new CategoryImpl(typeName, categoryName, categoryDescription);
   }

   /**
    * Parses out the JavaDoc comment from a string of source code. Returns the
    * first sentence summary in <code>summary</code> and the body of the entire
    * comment (including the summary) in <code>comment</code>.
    *
    * @param sourceCode The source code of a function, including its comment.
    *          Modified to remove leading whitespace.
    * @param summary Modified to contain the first sentence of the comment.
    * @param comment Modified to contain the entire comment.
    */
   private void getComment(StringBuffer sourceCode, StringBuffer summary,
       StringBuffer comment) {

     if (sourceCode == null) {
       return;
     }

     summary.setLength(0);
     comment.setLength(0);

     String regex = "/\\*\\*(.(?!}-\\*/))*\\*/";

     Pattern p = Pattern.compile(regex, Pattern.DOTALL);
     Matcher m = p.matcher(sourceCode);

     // Early out if there is no javadoc comment.
     if (!m.find()) {
       return;
     }

     String commentStr = m.group();

     p = Pattern.compile("(/\\*\\*\\s*)" + // The comment header
         "(((\\s*\\**\\s*)([^\n\r]*)[\n\r]+)*)" // The comment body
     );

     m = p.matcher(commentStr);

     if (!m.find()) {
       return;
     }

     String stripped = m.group(2);

     p = Pattern.compile("^\\p{Blank}*\\**\\p{Blank}*", Pattern.MULTILINE);
     String bareComment = p.matcher(stripped).replaceAll("");

     BreakIterator iterator = BreakIterator.getSentenceInstance();
     iterator.setText(bareComment);
     int firstSentenceEnd = iterator.next();
     if (firstSentenceEnd == BreakIterator.DONE) {
       summary.append(bareComment);
     } else {
       summary.append(bareComment.substring(0, firstSentenceEnd));
     }

     comment.append(bareComment);

     // Measure the indentation width on the second line to infer what the
     // first line indent should be.
     p = Pattern.compile("[^\\r\\n]+[\\r\\n]+(\\s+)\\*", Pattern.MULTILINE);
     m = p.matcher(sourceCode);
     int indentLen = 0;
     if (m.find()) {
       String indent = m.group(1);
       indentLen = indent.length() - 1;
     }
     StringBuffer leadingIndent = new StringBuffer();
     for (int i = 0; i < indentLen; ++i) {
       leadingIndent.append(' ');
     }

     // By inserting at 0 here, we are assuming that sourceCode begins with
     // /**, which is actually a function of how JDT sees a declaration start.
     // If in the future, you see bogus indentation here, it means that this
     // assumption is bad.
     sourceCode.insert(0, leadingIndent);
   }
 }
	/*
	* Copyright 2008 Google Inc.
	*
	* Licensed 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 com.google.gwt.benchmarks;

	import com.google.gwt.benchmarks.client.impl.BenchmarkResults;
	import com.google.gwt.benchmarks.client.impl.Trial;
	import com.google.gwt.benchmarks.rebind.BenchmarkGenerator;
	import com.google.gwt.core.ext.TreeLogger;
	import com.google.gwt.core.ext.typeinfo.HasAnnotations;
	import com.google.gwt.core.ext.typeinfo.HasMetaData;
	import com.google.gwt.core.ext.typeinfo.JClassType;
	import com.google.gwt.core.ext.typeinfo.JMethod;
	import com.google.gwt.core.ext.typeinfo.TypeOracle;
	import com.google.gwt.dev.util.Util;
	import com.google.gwt.util.tools.Utility;

	import junit.framework.TestCase;

	import org.w3c.dom.Document;
	import org.w3c.dom.Element;

	import java.io.File;
	import java.io.FileOutputStream;
	import java.io.IOException;
	import java.text.BreakIterator;
	import java.text.DateFormat;
	import java.util.ArrayList;
	import java.util.Date;
	import java.util.HashMap;
	import java.util.List;
	import java.util.Map;
	import java.util.regex.Matcher;
	import java.util.regex.Pattern;

	import javax.xml.parsers.DocumentBuilder;
	import javax.xml.parsers.DocumentBuilderFactory;
	import javax.xml.parsers.ParserConfigurationException;

	/**
	* Generates a detailed report that contains the results of all of the
	* benchmark-related unit tests executed during a unit test session. The primary
	* user of this class is {@link com.google.gwt.junit.JUnitShell}.
	*
	* The report is in XML format. To view the XML reports, use benchmarkViewer.
	*/
	public class BenchmarkReport {

	/**
	* Converts a set of test results for a single benchmark method into XML.
	*/
	private class BenchmarkXml {

	private MetaData metaData;

	private List<BenchmarkResults> results;

	private TestCase test;

	BenchmarkXml(TestCase test, List<BenchmarkResults> results) {
	this.test = test;
	this.results = results;
	Map<String, MetaData> methodMetaData = testMetaData.get(test.getClass().toString());
	metaData = methodMetaData.get(test.getName());
	}

	Element toElement(Document doc) {
	Element benchmark = doc.createElement("benchmark");
	benchmark.setAttribute("class", test.getClass().getName());
	benchmark.setAttribute("name", metaData.getTestName());
	benchmark.setAttribute("description", metaData.getTestDescription());

	String sourceCode = metaData.getSourceCode();
	if (sourceCode != null) {
	Element sourceCodeElement = doc.createElement("source_code");
	sourceCodeElement.appendChild(doc.createTextNode(sourceCode));
	benchmark.appendChild(sourceCodeElement);
	}

	// TODO(tobyr): create target_code element

	for (BenchmarkResults result : results) {
	benchmark.appendChild(toElement(doc, result));
	}

	return benchmark;
	}

	private Element toElement(Document doc, BenchmarkResults result) {
	Element resultElement = doc.createElement("result");
	resultElement.setAttribute("host", result.getHost());
	resultElement.setAttribute("agent", result.getAgent());

	Throwable exception = result.getException();

	if (exception != null) {
	Element exceptionElement = doc.createElement("exception");
	exceptionElement.appendChild(doc.createTextNode(exception.toString()));
	resultElement.appendChild(exceptionElement);
	}

	List<Trial> trials = result.getTrials();

	for (Trial trial : trials) {
	Element trialElement = toElement(doc, trial);
	resultElement.appendChild(trialElement);
	}

	return resultElement;
	}

	private Element toElement(Document doc, Trial trial) {
	Element trialElement = doc.createElement("trial");

	Map<String, String> variables = trial.getVariables();

	for (Map.Entry<String, String> entry : variables.entrySet()) {
	Object name = entry.getKey();
	Object value = entry.getValue();
	Element variableElement = doc.createElement("variable");
	variableElement.setAttribute("name", name.toString());
	variableElement.setAttribute("value", value.toString());
	trialElement.appendChild(variableElement);
	}

	trialElement.setAttribute("timing",
	String.valueOf(trial.getRunTimeMillis()));

	return trialElement;
	}
	}

	/**
	* Parses .java source files to get source code for methods.
	*/
	private class Parser {

	/**
	* Maps classes to the contents of their source files.
	*/
	private Map<JClassType, String> classSources = new HashMap<JClassType, String>();

	/**
	* Returns the source code for the method of the given name.
	*
	* @param logger to log the process
	* @param method a not <code>null</code> method
	* @return <code>null</code> if the source code for the method can not be
	* located
	*/
	public String getMethod(TreeLogger logger, JMethod method) {
	JClassType clazz = method.getEnclosingType();

	if (!classSources.containsKey(clazz)) {
	char[] sourceContents = null;
	File sourceFile = findSourceFile(clazz);
	if (sourceFile != null) {
	sourceContents = Util.readFileAsChars(sourceFile);
	classSources.put(clazz, new String(sourceContents));
	}

	if (sourceContents == null) {
	classSources.put(clazz, null);
	String msg = "An unknown I/O exception occured while trying to read "
	+ sourceFile.getAbsolutePath();
	logger.log(TreeLogger.WARN, msg, null);
	} else {
	classSources.put(clazz, new String(sourceContents));
	String msg = "BenchmarkReport read the contents of "
	+ sourceFile.getAbsolutePath();
	TreeLogger branch = logger.branch(TreeLogger.DEBUG, msg, null);
	if (logger.isLoggable(TreeLogger.SPAM)) {
	branch.log(TreeLogger.SPAM, new String(sourceContents), null);
	}
	}
	}

	String source = classSources.get(clazz);

	if (source == null) {
	return source;
	}

	try {
	return source.substring(method.getDeclStart(), method.getDeclEnd() + 1);
	} catch (IndexOutOfBoundsException e) {
	logger.log(TreeLogger.WARN, "Unable to parse " + method.getName(), e);
	// Have seen this happen when the compiler read the source using one
	// character encoding, and then this Parser read it in a different
	// encoding. I don't know if there are other cases in which this can
	// occur.
	return null;
	}
	}
	}

	/**
	* Converts an entire report into XML.
	*/
	private class ReportXml {

	private Map<String, Element> categoryElementMap = new HashMap<String, Element>();

	private Date date = new Date();

	private String version = "unknown";

	Element toElement(Document doc) {
	Element report = doc.createElement("gwt_benchmark_report");
	String dateString = DateFormat.getDateTimeInstance().format(date);
	report.setAttribute("date", dateString);
	report.setAttribute("gwt_version", version);

	// Add each test result into the report.
	// Add the category for the test result, if necessary.
	for (Map.Entry<TestCase, List<BenchmarkResults>> entry : testResults.entrySet()) {
	TestCase test = entry.getKey();
	List<BenchmarkResults> results = entry.getValue();
	BenchmarkXml xml = new BenchmarkXml(test, results);
	Element categoryElement = getCategoryElement(doc, report,
	xml.metaData.getCategory().getClassName());
	categoryElement.appendChild(xml.toElement(doc));
	}

	return report;
	}

	/**
	* Locates or creates the category element by the specified name.
	*
	* @param doc The document to search
	* @param report The report to which the category belongs
	* @param name The name of the category
	*
	* @return The matching category element
	*/
	private Element getCategoryElement(Document doc, Element report, String name) {
	Element e = categoryElementMap.get(name);

	if (e != null) {
	return e;
	}

	Element categoryElement = doc.createElement("category");
	categoryElementMap.put(name, categoryElement);
	CategoryImpl category = testCategories.get(name);
	categoryElement.setAttribute("name", category.getName());
	categoryElement.setAttribute("description", category.getDescription());

	report.appendChild(categoryElement);

	return categoryElement;
	}
	}

	private static final String GWT_BENCHMARK_CATEGORY = "gwt.benchmark.category";

	private static final String GWT_BENCHMARK_DESCRIPTION = "gwt.benchmark.description";

	private static final String GWT_BENCHMARK_NAME = "gwt.benchmark.name";

	private static File findSourceFile(JClassType clazz) {
	final char separator = File.separator.charAt(0);
	String filePath = clazz.getPackage().getName().replace('.', separator)
	+ separator + clazz.getSimpleSourceName() + ".java";
	String[] paths = getClassPath();

	for (int i = 0; i < paths.length; ++i) {
	File maybeSourceFile = new File(paths[i] + separator + filePath);

	if (maybeSourceFile.exists()) {
	return maybeSourceFile;
	}
	}

	return null;
	}

	private static String[] getClassPath() {
	String path = System.getProperty("java.class.path");
	return path.split(File.pathSeparator);
	}

	private static String getSimpleMetaData(HasMetaData hasMetaData, String name) {
	String[][] allValues = hasMetaData.getMetaData(name);

	if (allValues == null) {
	return null;
	}

	StringBuffer result = new StringBuffer();

	for (int i = 0; i < allValues.length; ++i) {
	String[] values = allValues[i];
	for (int j = 0; j < values.length; ++j) {
	result.append(values[j]);
	result.append(" ");
	}
	}

	String resultString = result.toString().trim();
	return resultString.equals("") ? null : resultString;
	}

	private Parser parser = new Parser();

	private Map<String, CategoryImpl> testCategories = new HashMap<String, CategoryImpl>();

	private Map<String, Map<String, MetaData>> testMetaData = new HashMap<String, Map<String, MetaData>>();

	private Map<TestCase, List<BenchmarkResults>> testResults = new HashMap<TestCase, List<BenchmarkResults>>();

	private TypeOracle typeOracle;

	public BenchmarkReport() {
	}

	/**
	* Adds the Benchmark to the report. All of the metadata about the benchmark
	* (category, name, description, etc...) is recorded from the TypeOracle.
	*
	* @param logger to log the process
	* @param deprecationBranch to log usages of old-style metadata
	* @param benchmarkClass The benchmark class to record. Must not be
	* <code>null</code>.
	* @param typeOracle The <code>TypeOracle</code> for the compilation session
	* must not be <code>null</code>.
	*/
	public void addBenchmark(TreeLogger logger, TreeLogger deprecationBranch,
	JClassType benchmarkClass, TypeOracle typeOracle) {

	this.typeOracle = typeOracle;
	String categoryType = getBenchmarkCategory(deprecationBranch,
	benchmarkClass);

	Map<String, JMethod> zeroArgMethods = BenchmarkGenerator.getNotOverloadedTestMethods(benchmarkClass);
	Map<String, JMethod> parameterizedMethods = BenchmarkGenerator.getParameterizedTestMethods(
	benchmarkClass, TreeLogger.NULL);
	List<JMethod> testMethods = new ArrayList<JMethod>(zeroArgMethods.size()
	+ parameterizedMethods.size());
	testMethods.addAll(zeroArgMethods.values());
	testMethods.addAll(parameterizedMethods.values());

	Map<String, MetaData> metaDataMap = testMetaData.get(benchmarkClass.toString());
	if (metaDataMap == null) {
	metaDataMap = new HashMap<String, MetaData>();
	testMetaData.put(benchmarkClass.toString(), metaDataMap);
	}

	// Add all of the benchmark methods
	for (JMethod method : testMethods) {
	String methodName = method.getName();
	String methodCategoryType = getBenchmarkCategory(deprecationBranch,
	method);
	if (methodCategoryType == null) {
	methodCategoryType = categoryType;
	}
	CategoryImpl methodCategory = getCategory(deprecationBranch,
	methodCategoryType);

	String methodSource = parser.getMethod(logger, method);
	StringBuffer sourceBuffer = (methodSource == null) ? null
	: new StringBuffer(methodSource);
	StringBuffer summary = new StringBuffer();
	StringBuffer comment = new StringBuffer();
	getComment(sourceBuffer, summary, comment);

	MetaData metaData = new MetaData(benchmarkClass.toString(), methodName,
	(sourceBuffer != null) ? sourceBuffer.toString() : null,
	methodCategory, methodName, summary.toString());
	metaDataMap.put(methodName, metaData);
	}
	}

	public void addBenchmarkResults(TestCase test, BenchmarkResults results) {
	List<BenchmarkResults> currentResults = testResults.get(test);
	if (currentResults == null) {
	currentResults = new ArrayList<BenchmarkResults>();
	testResults.put(test, currentResults);
	}
	currentResults.add(results);
	}

	/**
	* Generates reports for all of the benchmarks which were added to the
	* generator.
	*
	* @param outputPath The path to write the reports to.
	* @throws ParserConfigurationException If an error occurs during xml parsing
	* @throws IOException If anything goes wrong writing to outputPath
	*/
	public void generate(String outputPath) throws ParserConfigurationException,
	IOException {

	// Don't generate a new report if no tests were actually run.
	if (testResults.size() == 0) {
	return;
	}

	DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
	DocumentBuilder builder = factory.newDocumentBuilder();
	Document doc = builder.newDocument();
	doc.appendChild(new ReportXml().toElement(doc));
	byte[] xmlBytes = Util.toXmlUtf8(doc);
	FileOutputStream fos = null;
	try {
	fos = new FileOutputStream(outputPath);
	fos.write(xmlBytes);
	} finally {
	Utility.close(fos);
	}

	// TODO(bruce) The code below is commented out because of GWT Issue 958.

	// // TODO(tobyr) Looks like indenting is busted
	// // http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6296446
	// // Not a big deal, since we don't intend to read the XML by hand anyway
	// TransformerFactory transformerFactory = TransformerFactory.newInstance();
	// // Think this can be used with JDK 1.5
	// // transformerFactory.setAttribute( "indent-number", new Integer(2) );
	// Transformer serializer = transformerFactory.newTransformer();
	// serializer.setOutputProperty(OutputKeys.METHOD, "xml");
	// serializer.setOutputProperty(OutputKeys.INDENT, "yes");
	// serializer
	// .setOutputProperty("{ http://xml.apache.org/xslt }indent-amount", "2");
	// BufferedOutputStream docOut = new BufferedOutputStream(
	// new FileOutputStream(outputPath));
	// serializer.transform(new DOMSource(doc), new StreamResult(docOut));
	// docOut.close();
	}

	private <T extends HasMetaData & HasAnnotations> String getBenchmarkCategory(
	TreeLogger deprecationBranch, T element) {
	String category = getSimpleMetaData(element, GWT_BENCHMARK_CATEGORY);
	if (category != null) {
	deprecationBranch.log(TreeLogger.WARN, GWT_BENCHMARK_CATEGORY + " has "
	+ "been deprecated with no replacement.", null);
	}
	return category;
	}

	private CategoryImpl getCategory(TreeLogger deprecationBranch, String name) {
	CategoryImpl c = testCategories.get(name);

	if (c != null) {
	return c;
	}

	c = getCategoryMetaData(deprecationBranch, name);
	testCategories.put(name, c);
	return c;
	}

	private CategoryImpl getCategoryMetaData(TreeLogger deprecationBranch,
	String typeName) {
	if (typeName == null) {
	return new CategoryImpl(null, "", "");
	}

	JClassType categoryType = typeOracle.findType(typeName);

	if (categoryType == null) {
	return new CategoryImpl(typeName, "", "");
	}

	String categoryName = getSimpleMetaData(categoryType, GWT_BENCHMARK_NAME);
	String categoryDescription = getSimpleMetaData(categoryType,
	GWT_BENCHMARK_DESCRIPTION);

	if (categoryName != null \|\| categoryDescription != null) {
	deprecationBranch.log(TreeLogger.WARN, GWT_BENCHMARK_NAME + " and "
	+ GWT_BENCHMARK_DESCRIPTION + " have been deprecated with no "
	+ "replacement", null);
	}

	categoryName = categoryName == null ? "" : categoryName;
	categoryDescription = categoryDescription == null ? ""
	: categoryDescription;

	return new CategoryImpl(typeName, categoryName, categoryDescription);
	}

	/**
	* Parses out the JavaDoc comment from a string of source code. Returns the
	* first sentence summary in <code>summary</code> and the body of the entire
	* comment (including the summary) in <code>comment</code>.
	*
	* @param sourceCode The source code of a function, including its comment.
	* Modified to remove leading whitespace.
	* @param summary Modified to contain the first sentence of the comment.
	* @param comment Modified to contain the entire comment.
	*/
	private void getComment(StringBuffer sourceCode, StringBuffer summary,
	StringBuffer comment) {

	if (sourceCode == null) {
	return;
	}

	summary.setLength(0);
	comment.setLength(0);

	String regex = "/\\\\(.(?!}-\\/))\\*/";

	Pattern p = Pattern.compile(regex, Pattern.DOTALL);
	Matcher m = p.matcher(sourceCode);

	// Early out if there is no javadoc comment.
	if (!m.find()) {
	return;
	}

	String commentStr = m.group();

	p = Pattern.compile("(/\\\\\\s*)" + // The comment header
	"(((\\s\\\\s)([^\n\r])[\n\r]+))" // The comment body
	);

	m = p.matcher(commentStr);

	if (!m.find()) {
	return;
	}

	String stripped = m.group(2);

	p = Pattern.compile("^\\p{Blank}\\\\p{Blank}", Pattern.MULTILINE);
	String bareComment = p.matcher(stripped).replaceAll("");

	BreakIterator iterator = BreakIterator.getSentenceInstance();
	iterator.setText(bareComment);
	int firstSentenceEnd = iterator.next();
	if (firstSentenceEnd == BreakIterator.DONE) {
	summary.append(bareComment);
	} else {
	summary.append(bareComment.substring(0, firstSentenceEnd));
	}

	comment.append(bareComment);

	// Measure the indentation width on the second line to infer what the
	// first line indent should be.
	p = Pattern.compile("[^\\r\\n]+[\\r\\n]+(\\s+)\\*", Pattern.MULTILINE);
	m = p.matcher(sourceCode);
	int indentLen = 0;
	if (m.find()) {
	String indent = m.group(1);
	indentLen = indent.length() - 1;
	}
	StringBuffer leadingIndent = new StringBuffer();
	for (int i = 0; i < indentLen; ++i) {
	leadingIndent.append(' ');
	}

	// By inserting at 0 here, we are assuming that sourceCode begins with
	// /**, which is actually a function of how JDT sees a declaration start.
	// If in the future, you see bogus indentation here, it means that this
	// assumption is bad.
	sourceCode.insert(0, leadingIndent);
	}
	}