OpenCSV: How to create CSV file from POJO with custom column headers and custom column positions?

Question

I have created a MappingsBean class where all the columns of the CSV file are specified. Next I parse XML files and create a list of mappingbeans. Then I write that data int

Answer 1

There is another version for 5.2 version because I have a problem with @CsvCustomBindByName annotation when I tried answers above.

I defined custom annotation :

@Target(ElementType.FIELD)
@Inherited
@Retention(RetentionPolicy.RUNTIME)
public @interface CsvPosition {

  int position();
}

and custom mapping strategy

public class CustomMappingStrategy extends HeaderColumnNameMappingStrategy {

  private final Field[] fields;

  public CustomMappingStrategy(Class clazz) {
    fields = clazz.getDeclaredFields();
    Arrays.sort(fields, (f1, f2) -> {
      CsvPosition position1 = f1.getAnnotation(CsvPosition.class);
      CsvPosition position2 = f2.getAnnotation(CsvPosition.class);
      return Integer.compare(position1.position(), position2.position());
    });
  }

  @Override
  public String[] generateHeader(T bean) throws CsvRequiredFieldEmptyException {
    String[] header = new String[fields.length];
    for (Field f : fields) {
      CsvPosition position = f.getAnnotation(CsvPosition.class);
      header[position.position() - 1] = getName(f);
    }
    headerIndex.initializeHeaderIndex(header);
    return header;
  }

  private String getName(Field f) {
    CsvBindByName csvBindByName = f.getAnnotation(CsvBindByName.class);
    CsvCustomBindByName csvCustomBindByName = f.getAnnotation(CsvCustomBindByName.class);
    return csvCustomBindByName != null
      ? csvCustomBindByName.column() == null || csvCustomBindByName.column().isEmpty() ? f.getName() : csvCustomBindByName.column()
      : csvBindByName.column() == null || csvBindByName.column().isEmpty() ? f.getName() : csvBindByName.column();
  }

}

My POJO beans are annotated like this

public class Record {

  @CsvBindByName(required = true)
  @CsvPosition(position = 1)
  Long id;
  @CsvCustomBindByName(required = true, converter = BoolanCSVField.class)
  @CsvPosition(position = 2)
  Boolean deleted;
  ...
}

and final code for writer :

CustomMappingStrategy mappingStrategy = new CustomMappingStrategy<>(Record.class);
mappingStrategy.setType(Record.class);
StatefulBeanToCsv beanToCsv = new StatefulBeanToCsvBuilder(writer)
.withApplyQuotesToAll(false)
.withOrderedResults(true)
.withMappingStrategy(mappingStrategy)
.build();

I hope it will helpful for someone