mappers

Easy Batch development is POJO-centric and has been designed with the idea that input records should be projected in the object oriented world and not used in their original raw format.

It is the RecordMapper that lets you map input record payloads to domain objects. You can register an implementation of the RecordMapper interface as follows:

Job job = new JobBuilder()
    .mapper(new MyRecordMapper())
    .build();

There are several built-in record mappers to map records from a variety of data formats:

Record mapper	Expected record type	Module	Description
DelimitedRecordMapper	StringRecord	easybatch-flatfile	Maps delimited-values records to Java objects
FixedLengthRecordMapper	StringRecord	easybatch-flatfile	Maps fixed-length field records to Java objects
UnivocityFixedWidthRecordMapper	StringRecord	easybatch-univocity	Maps fixed-width records to Java objects using Univocity
ApacheCommonCsvRecordMapper	StringRecord	easybatch-apache-common-csv	Maps delimited-values records to Java objects
OpenCsvRecordMapper	StringRecord	easybatch-opencsv	Maps delimited-values records to Java objects using OpenCSV
UnivocityCsvRecordMapper	StringRecord	easybatch-univocity	Maps comma separated-values records to Java objects using Univocity
UnivocityTsvRecordMapper	StringRecord	easybatch-univocity	Maps tab separated-values records to Java objects using Univocity
XmlRecordMapper	XmlRecord	easybatch-xml	Maps xml records to Java objects annotated with JAXB annotations
XstreamRecordMapper	XmlRecord	easybatch-xstream	Maps xml records to Java objects using XStream
JacksonRecordMapper	JsonRecord	easybatch-jackson	Maps json records to Java objects using Jackson
GsonRecordMapper	JsonRecord	easybatch-gson	Maps json records to Java objects using Gson
YamlRecordMapper	YamlRecord	easybatch-yaml	Maps yaml records to Java objects using yamlbeans
JdbcRecordMapper	JdbcRecord	easybatch-jdbc	Maps JDBC records to Java objects
SpringJdbcRecordMapper	JdbcRecord	easybatch-spring	Maps json records to Java objects using Spring JDBC
MongoDBRecordMapper	MongoDBRecord	easybatch-mongodb	Maps mongodb records to Java objects using Morphia
MsExcelRecordMapper	MsExcelRecord	easybatch-msexcel	Maps MsExcel records to Java objects using Apache POI

Record mappers take a Record, map it's payload to a domain object and return another Record with the same header but with the domain object as payload.

Custom type conversion

Some record mappers ( DelimitedRecordMapper, FixedLengthRecordMapper, ApacheCommonCsvRecordMapper and JdbcRecordMapper) convert raw textual data to typed data in Java objects. Easy Batch supports all Java primitive and wrapper types. If you want to provide a custom type converter, you can implement the org.easybatch.core.api.TypeConverter interface and register your implementation within the record mapper in use.

Delimited-values records considerations

If a delimited record is not well formed, the DelimitedRecordMapper throws an exception that causes the record to be rejected in the following cases:

• Fields number is not equal to the expected fields number (missing field, extra field, etc) as specified in the CSV RFC (section 2.4)](http://www.ietf.org/rfc/rfc4180.txt[).
• A field is not qualified as expected with the data qualifier, which means that the DelimitedRecordMapper expects all fields to be qualified when a qualifier is specified.

DelimitedRecordMapper limitations

The DelimitedRecordMapper is intended to cover basic requirements of delimited-values mapping. It does not support detecting delimiters and line breaks in a qualified field. If you need these features, you can use the ApacheCommonCsvRecordMapper or the OpenCsvRecordMapper.

Fixed-length records considerations

If a fixed length record is not well formed, the FixedLengthRecordMapper throws an exception that causes the record to be rejected. A fixed length record is not well formed if its length is not equal to expected record length.

Xml record validation

When you create a XmlRecordMapper, you should specify your target domain object type. If you need to validate Xml records against a Xsd schema, you can specify the schema at creation time as follows:

Job job = new JobBuilder()
        .mapper(new XmlRecordMapper(MyPojoType.class, myXsdFile))
        .build();

Json to object mapping

Easy Batch does not provide support for JSON mapping yet. The goal is to use the reference implementation of JSON-B. As of version 5.0.0, the reference implementation of JSON-B is not available yet. Built-in support for JSON mapping will be provided as soon as the reference implementation of JSON-B will become available. You can still use easybatch-jackson and easybatch-gson modules which provide JSON mapping using third party libraries.