When reading CSV data using super-csv library, one can specify cell processors for columns. Some tests use the incorrect format:
ParseDate(YYYY-MM-dd), which should be
ParseDate('yyyy-MM-dd')

Affected files:
jberet-support/src/test/java/org/jberet/support/io/MovieTest.java
wildfly-jberet-samples/jberet-samples-common/src/main/resources/META-INF/batch-jobs/csvItemReader.xml

The symptom is, when reading items into java.util.Map types, the date objects only have year value, i.e., all movie open dates will be 2012-01-01. See attachment screenshot 06-psql.png (running intro-jberet app, csv2db.xml job, in OpenShift, screen showing PostgreSQL database content after job completion)

After the fix, the screenshot Screenshot-2017-10-1 OpenShift Web Console.png shows the correctly formated date in opn column.