Java Reference
In-Depth Information
<beans:property name="lineAggregator">
<beans:property name="delimiter" value=","/>
<beans:property name="fieldExtractor">
<beans:property name="names"
<beans:property name="resource" ref="summaryFile" />
<step id="generateAccountSummaryStep">
<chunk reader="accountSummaryReader" writer="accountSummaryWriter"
With all the steps configured, you can finally configure the job itself. In the job, you configure the
three steps as discussed. However, for step1 , if the step returns STOPPED , you stop the job. If the job is
restarted, it reexecutes step1 . If step1 returns any other successful value, the job continues with step2
and finally step3 . Listing 6-26 shows the configuration for this logic using the <stop> tag.
Listing 6-26. Configuring a Job Using the <stop> Tag
<job id="transactionJob">
<step id="step1" parent="importTransactionFileStep">
<stop on="STOPPED" restart="step1"/>
<next on="*" to="step2"/>
<step id="step2" parent="applyTransactionsStep" next="step3"/>
<step id="step3" parent="generateAccountSummaryStep"/>
How do you perform the check to be sure you read in the correct number of records? In this case,
you develop a custom reader that also serves as a step listener. The code in Listing 6-27 shows how the
reader reads in all the records and keeps a count of how many are read in. When the step is complete,
the listener validates that the number of records read in matches the number of records expected. If they
match, the ExitStatus that was determined by the regular processing is returned. If they don't match,
you override the ExitStatus by returning your own ExitStatus.STOPPED value.
Search WWH ::

Custom Search