Parsing enums with SuperCSV ICsvBeanReader

Question

I parse CSV file and create a domain objects using supercsv. My domain object has one enum field, e.g.:

public class TypeWithEnum {

private Type type;

public TypeWithEnum(Type type) {
    this.type = type;
}

public Type getType() {
    return type;
}

public void setType(Type type) {
    this.type = type;
}
}

My enum looks like this:

public enum Type {

    CANCEL, REFUND
}

Trying to create beans out of this CSV file:

final String[] header = new String[]{ "type"  };
ICsvBeanReader inFile = new CsvBeanReader(new FileReader(
    getFilePath(this.getClass(), "learning/enums.csv")), CsvPreference.STANDARD_PREFERENCE);

final CellProcessor[] processors = 
    new CellProcessor[]{ TODO WHAT TO PUT HERE? };
TypeWithEnum myEnum = inFile.read(
    TypeWithEnum.class, header, processors);

this fails with Error while filling an object context: null offending processor: null at org.supercsv.io.CsvBeanReader.fillObject(Unknown Source) at org.supercsv.io.CsvBeanReader.read(Unknown Source)

Any hint on parsing enums? Should I write my own processor for this?

I already tried to write my own processor, something like this:

class MyCellProcessor extends CellProcessorAdaptor {
    public Object execute(Object value, CSVContext context) {
        Type type = Type.valueOf(value.toString());
        return next.execute(type, context);
    }
}

but it dies with the same exception.

The content of my enums.csv file is simple:

CANCEL
REFUND

Answer 1

I tried to reproduce your Error but everything works for me. I use SuperCSV 1.52:

  private enum ENUMS_VALUES{TEST1, TEST2, TEST3};
  @Test
  public void testEnum3() throws IOException
  {
    String testInput = new String("TEST1\nTEST2\nTEST3");
    ICsvBeanReader  reader = new CsvBeanReader(new StringReader(testInput), CsvPreference.EXCEL_NORTH_EUROPE_PREFERENCE);
    final String[] header = new String[] {"header"};
    reader.read(this.getClass(), header, new CellProcessor[] {new CellProcessorAdaptor() {

      @Override
      public Object execute(Object pValue, CSVContext pContext)
      {
        return next.execute(ENUMS_VALUES.valueOf((String)pValue), pContext);
      }}});

  }

  @Test
  public void testEnum4() throws IOException
  {
    String testInput = new String("TEST1\nTEST2\nTEST3");
    ICsvBeanReader reader = new CsvBeanReader(new StringReader(testInput), CsvPreference.EXCEL_NORTH_EUROPE_PREFERENCE);
    final String[] header = new String[] {"header"};
    reader.read(this.getClass(), header, new CellProcessor[] {new CellProcessorAdaptor()
    {

      @Override
      public Object execute(Object pValue, CSVContext pContext)
      {
        return ENUMS_VALUES.valueOf((String)pValue);
      }}});
  }

  public void setHeader(ENUMS_VALUES value)
  {
    System.out.println(value);
  }

Answer 2

Here is a generic cell processor for enums

/** A cell processor to convert strings to enums. */
public class EnumCellProcessor<T extends Enum<T>> implements CellProcessor {

    private Class<T> enumClass;
    private boolean ignoreCase;

    /**
     * @param enumClass the enum class used for conversion
     */
    public EnumCellProcessor(Class<T> enumClass) {
        this.enumClass = enumClass;
    }

    /**
     * @param enumClass the enum class used for conversion
     * @param ignoreCase if true, the conversion is made case insensitive
     */
    public EnumCellProcessor(Class<T> enumClass, boolean ignoreCase) {
        this.enumClass = enumClass;
        this.ignoreCase = ignoreCase;
    }

    @Override
    public Object execute(Object value, CsvContext context) {
        if (value == null)
            return null;

        String valueAsStr = value.toString();

        for (T s : enumClass.getEnumConstants()) {
            if (ignoreCase ? s.name().equalsIgnoreCase(valueAsStr) : s.name().equals(valueAsStr)) {
                return s;
            }
        }

        throw new SuperCsvCellProcessorException(valueAsStr + " cannot be converted to enum " + enumClass.getName(), context, this);
    }

}

and you will use it

new EnumCellProcessor<Type>(Type.class);

Answer 3

The exception you're getting is because CsvBeanReader cannot instantiate your TypeWithEnum class, as it doesn't have a default (no arguments) constructor. It's probably a good idea to print the stack trace so you can see the full details of what went wrong.

Super CSV relies on the fact that you should have supplied a valid Java bean, i.e. a class with a default constructor and public getters/setters for each of its fields.

So you can fix the exception by adding the following to TypeWithEnum:

public TypeWithEnum(){
}

As for hints on parsing enums the two easiest options are:

1. Using the HashMapper processor

@Test
public void hashMapperTest() throws Exception {

    // two lines of input
    String input = "CANCEL\nREFUND";

    // you could also put the header in the CSV file
    // and use inFile.getCSVHeader(true)
    final String[] header = new String[] { "type" };

    // map from enum name to enum
    final Map<Object, Object> typeMap = new HashMap<Object, Object>();
    for( Type t : Type.values() ) {
        typeMap.put(t.name(), t);
    }

    // HashMapper will convert from the enum name to the enum
    final CellProcessor[] processors = 
        new CellProcessor[] { new HashMapper(typeMap) };

    ICsvBeanReader inFile = 
        new CsvBeanReader(new StringReader(input),
        CsvPreference.STANDARD_PREFERENCE);

    TypeWithEnum myEnum;
    while((myEnum = inFile.read(TypeWithEnum.class, header, processors)) !=null){
        System.out.println(myEnum.getType());
    }

}

2. Creating a custom CellProcessor

Create your processor

package org.supercsv;

import org.supercsv.cellprocessor.CellProcessorAdaptor;
import org.supercsv.cellprocessor.ift.CellProcessor;
import org.supercsv.exception.SuperCSVException;
import org.supercsv.util.CSVContext;

public class TypeProcessor extends CellProcessorAdaptor {

    public TypeProcessor() {
        super();
    }

    public TypeProcessor(CellProcessor next) {
        super(next);
    }

    public Object execute(Object value, CSVContext context) {

        if (!(value instanceof String)){
            throw new SuperCSVException("input should be a String!");
        }

        // parse the String to a Type
        Type type = Type.valueOf((String) value);

        // execute the next processor in the chain
        return next.execute(type, context);
    }

}

Use it!

@Test
public void customProcessorTest() throws Exception {

    // two lines of input
    String input = "CANCEL\nREFUND";

    final String[] header = new String[] { "type" };

    // HashMapper will convert from the enum name to the enum
    final CellProcessor[] processors = 
        new CellProcessor[] { new TypeProcessor() };

    ICsvBeanReader inFile = 
        new CsvBeanReader(new StringReader(input),
        CsvPreference.STANDARD_PREFERENCE);
    TypeWithEnum myEnum;
    while((myEnum = inFile.read(TypeWithEnum.class, header, processors)) !=null){
        System.out.println(myEnum.getType());
    }

}

I'm working on an upcoming release of Super CSV. I'll be sure to update the website to make it clear that you have to have a valid Java bean - and maybe a description of the available processors, for those not inclined to read Javadoc.