Testing a FieldSetMapper which reads by column names throws exception - spring-batch

Question

The line mapping logic in the job config xml is as follows:
<property name="lineTokenizer">
<bean
class="org.springframework.batch.item.file.transform.DelimitedLineTokenizer">
<property name="names" value="ID,NAME"/>
<property name="strict" value="false"/>
</bean>
</property>
<property name="fieldSetMapper">
<bean class="com.company.batch.mappers.EmpFieldSetMapper"/>
</property>
The field set mapper read logic is as follows:
fieldSet.readString("ID");
But while writing UT for EmpFieldSetMapper, I am getting below error:
java.lang.IllegalArgumentException: Cannot access columns by name without meta data
at org.springframework.batch.item.file.transform.DefaultFieldSet.indexOf(DefaultFieldSet.java:675)
at org.springframework.batch.item.file.transform.DefaultFieldSet.readString(DefaultFieldSet.java:169)
The UT is as follows:
#Before
public void setUp() throws Exception {
mapper = new EmpFieldSetMapper();
DefaultFieldSetFactory fieldSetFactory = new DefaultFieldSetFactory();
String[] inputValues = { "123", "RAJ" };
fieldSet = fieldSetFactory.create(inputValues);
}
#Test
public void testMapFieldSet() {
try {
Model model = mapper.mapFieldSet(fieldSet);
assertEquals("ID field mapping is wrong", "123", model.getId());
assertEquals("NAME field mapping is wrong", "RAJ", model.getName());
} catch (BindException e) {
fail("Exception during field set mapping");
}
}
I think I need to change the DefaultFieldSetMapper to something else but unsure about it. This issue can be resolved by replacing the column names with index but I want to retain the column names in EmpFieldSetMapper. So need suggestions.

Answer 1

I found the answer. Need to pass in column names as well.
String[] inputValues = { "123", "RAJ" };
String[] inputColumns = { "ID", "NAME" };
fieldSet = fieldSetFactory.create(inputValues, inputColumns);