Bug 44650 - CSV Dataset now handles quoted column values

git-svn-id: https://svn.apache.org/repos/asf/jakarta/jmeter/trunk@640095 13f79535-47bb-0310-9956-ffa450edef68

Former-commit-id: 6b24c1c6a3

src/components/org/apache/jmeter/config/CSVDataSet.java

@@ -19,10 +19,12 @@
 package org.apache.jmeter.config;
 
 import java.io.IOException;
+import java.io.StringReader;
 
 import org.apache.jmeter.config.ConfigTestElement;
 import org.apache.jmeter.engine.event.LoopIterationEvent;
 import org.apache.jmeter.engine.event.LoopIterationListener;
+import org.apache.jmeter.save.CSVSaveService;
 import org.apache.jmeter.services.FileServer;
 import org.apache.jmeter.testbeans.TestBean;
 import org.apache.jmeter.threads.JMeterVariables;
@@ -32,10 +34,6 @@ import org.apache.jorphan.util.JMeterStopThreadException;
 import org.apache.jorphan.util.JOrphanUtils;
 import org.apache.log.Logger;
 
-/**
- * @author mstover
- * 
- */
 public class CSVDataSet extends ConfigTestElement implements TestBean, LoopIterationListener {
 	private static final Logger log = LoggingManager.getLoggerForClass();
 
@@ -52,6 +50,8 @@ public class CSVDataSet extends ConfigTestElement implements TestBean, LoopItera
 
     private transient String delimiter;
 
+    private transient boolean quoted = false;
+    
     private transient boolean recycle = true;
     
     private transient boolean stopThread = false;
@@ -81,7 +81,9 @@ public class CSVDataSet extends ConfigTestElement implements TestBean, LoopItera
             JMeterVariables threadVars = this.getThreadContext().getVariables();
 			String line = server.readLine(_fileName,getRecycle());
             if (line!=null) {// i.e. not EOF
-                String[] lineValues = JOrphanUtils.split(line, delim,false);
+                String[] lineValues = getQuotedData() ? 
+                        CSVSaveService.csvReadFile(new StringReader(line), delim.charAt(0))
+                        : JOrphanUtils.split(line, delim, false);
     			for (int a = 0; a < vars.length && a < lineValues.length; a++) {
     				threadVars.put(vars[a], lineValues[a]);
     			}
@@ -152,6 +154,14 @@ public class CSVDataSet extends ConfigTestElement implements TestBean, LoopItera
 		this.delimiter = delimiter;
 	}
 
+    public boolean getQuotedData() {
+        return quoted;
+    }
+
+    public void setQuotedData(boolean quoted) {
+        this.quoted = quoted;
+    }
+
     public boolean getRecycle() {
         return recycle;
     }

src/components/org/apache/jmeter/config/CSVDataSetBeanInfo.java

@@ -35,11 +35,12 @@ public class CSVDataSetBeanInfo extends BeanInfoSupport {
     private static final String DELIMITER = "delimiter";             //$NON-NLS-1$
     private static final String RECYCLE = "recycle";                 //$NON-NLS-1$
     private static final String STOPTHREAD = "stopThread";           //$NON-NLS-1$
+    private static final String QUOTED_DATA = "quotedData";          //$NON-NLS-1$
 
 	public CSVDataSetBeanInfo() {
 		super(CSVDataSet.class);
 		createPropertyGroup("csv_data",             //$NON-NLS-1$
-                new String[] { FILENAME, FILE_ENCODING, VARIABLE_NAMES, DELIMITER, RECYCLE, STOPTHREAD });
+                new String[] { FILENAME, FILE_ENCODING, VARIABLE_NAMES, DELIMITER, QUOTED_DATA, RECYCLE, STOPTHREAD });
         
 		PropertyDescriptor p = property(FILENAME);
 		p.setValue(NOT_UNDEFINED, Boolean.TRUE);
@@ -61,6 +62,11 @@ public class CSVDataSetBeanInfo extends BeanInfoSupport {
 		p.setValue(DEFAULT, ",");        //$NON-NLS-1$
 		p.setValue(NOT_EXPRESSION, Boolean.TRUE);
         
+        p = property(QUOTED_DATA);
+        p.setValue(NOT_UNDEFINED, Boolean.TRUE);
+        p.setValue(DEFAULT, Boolean.FALSE);
+        p.setValue(NOT_EXPRESSION, Boolean.TRUE);
+        
         p = property(RECYCLE);
         p.setValue(NOT_UNDEFINED, Boolean.TRUE);
         p.setValue(DEFAULT, Boolean.TRUE);

src/components/org/apache/jmeter/config/CSVDataSetResources.properties

@@ -8,6 +8,8 @@ variableNames.displayName=Variable Names (comma-delimited)
 variableNames.shortDescription=List your variable names in order to match the order of columns in your csv data.  Separate by commas.
 delimiter.displayName=Delimiter (use '\\t' for tab)
 delimiter.shortDescription=Enter the delimiter ('\\t' for tab)
+quotedData.displayName=Allow quoted data?
+quotedData.shortDescription=Allow CSV data values to be quoted?
 recycle.displayName=Recycle on EOF ?
 recycle.shortDescription=Should the file be re-read from the start on reaching EOF ?
 stopThread.displayName=Stop thread on EOF ?

xdocs/changes.xml

@@ -121,6 +121,7 @@ Note __split() and __regex() cannot be used on the Test Plan.</li>
 As a special case, if the HTTP Sampler path starts with "http://" or "https://" then this is used as the full URL.
 </li>
 <li>Bug 44575 - Result Saver can now save only successful results</li>
+<li>Bug 44650 - CSV Dataset now handles quoted column values</li>
 </ul>
 
 <h4>Non-functional changes</h4>

xdocs/usermanual/component_reference.xml

@@ -2084,11 +2084,13 @@ the Report.
 	<br></br>
 </description>
 
-<component name="CSV Data Set Config" index="&sect-num;.4.1"  width="363" height="251" screenshot="csvdatasetconfig.png">
+<component name="CSV Data Set Config" index="&sect-num;.4.1"  width="308" height="278" screenshot="csvdatasetconfig.png">
 <description>
-        <p>
+    <p>
 	CSV Data Set Config is used to read lines from a file, and split them into variables.
 	It is easier to use than the __CSVRead() and _StringFromFile() functions.
+	Versions of JMeter after 2.3.1 allow variables to be quoted; this allows the value to contain a delimiter.
+	Previously it was necessary to choose a delimiter that was not used in any values.
 	</p>
 	<p>
 	The file is only opened once, and each thread will use a different line from the file. 
@@ -2100,7 +2102,7 @@ the Report.
 	However the variables do work in the HTTP Auth Manager, as the username etc are processed at run-time.
 	</note>
 	<p>
-	As a special case, the string "\t" (without quotes) is treated as a Tab.
+	As a special case, the string "\t" (without quotes) in the delimiter field is treated as a Tab.
 	</p>
 	<p>
 	When the end of file (EOF) is reached, and the recycle option is true, reading starts again with the first line of the file.
@@ -2125,6 +2127,7 @@ the Report.
   <property name="Delimiter" required="Yes">Delimiter to be used to split the records in the file.
   If there are fewer values on the line than there are variables the remaining variables are not updated -
   so they will retain their previous value (if any).</property>
+  <property name="Allow quoted data?" required="Yes">Should the CSV file allow values to be quoted?</property>
   <property name="Recycle on EOF?" required="Yes">Should the file be re-read from the beginning on reaching EOF? (default is true)</property>
   <property name="Stop thread on EOF?" required="Yes">Should the thread be stopped on EOF, if Recycle is false? (default is false)</property>
 </properties>