Using Jackson and Immutables to create a Map from certain CSV columns

Question

I have a CSV that I'm parsing through Jackson to create list of immutable type for each row. The CSV has some columns to represent a collection like:

Name,Age,Height,Item1,Item2,Item3,Item4,Item5
AB,1,1.123,ABC,DEF,GHI,JKL,MNO

At present I'm parsing this into the following type:

@Value.Immutable
@JsonSerialize
@JsonDeserialize
@JsonPropertyOrder({"name","age","height","item1","item2","item3","item4","item5"})

public abstract class Schema {

    abstract String name();
    abstract int age();
    abstract double height();

    abstract String item1();
    abstract String item2();
    abstract String item3();
    abstract String item4();
    abstract String item5();

}

I'm using the csv schema from the immutable class:

        CsvMapper csvMapper = new CsvMapper()
                .enable(CsvGenerator.Feature.ALWAYS_QUOTE_STRINGS)
                .enable(CsvParser.Feature.EMPTY_STRING_AS_NULL);
        CsvSchema schema = csvMapper.typedSchemaFor(Schema.class).withHeader();

This works fine, however I'd like to make the items into a map, either using their column name or the item number as a key.

    abstract Map<String, String> items();
//or
    abstract Map<int, String> items();

I can do this by creating a second object that makes a collection, but I was wondering if there's a way for Jackson/Immutables to do this automatically? Unfortunately I can't alter the CSV file as its from an external source.

Edit: As a slightly clunky solution, I can create a map using:

    @Value.Redacted abstract String item1();
    @Value.Redacted abstract String item2();
    @Value.Redacted abstract String item3();
    @Value.Redacted abstract String item4();
    @Value.Redacted abstract String item5();

    @Value.Derived
    public Map<Integer, String> items() {
        Map<Integer, String> items = new HashMap<>();
        items.put(1, item1());
        items.put(2, item2());
        items.put(3, item3());
        items.put(4, item4());
        items.put(5, item5());
        return items;
    }

This hides the individual items, but I wonder if there is a more efficient way?

Answer 1

You can use a MappingIterator to deserialize your csv to a Map<String, String>. If you use your csv:

Name,Age,Height,Item1,Item2,Item3,Item4,Item5
AB,1,1.123,ABC,DEF,GHI,JKL,MNO

You can define a CsvSchema schema that uses first row as a header and instantiates a new MappingIterator iterator from your CsvMapper mapper:

CsvMapper mapper = new CsvMapper();
// use first row as header
CsvSchema schema = CsvSchema.emptySchema().withHeader();
MappingIterator<Map<String, String>> it = mapper.readerFor(Map.class)
        .with(schema)
        .readValues(csv);

So you can read your data row and convert it to a Map<String, String> map removing the unwanted columns:

CsvMapper mapper = new CsvMapper();
CsvSchema schema = CsvSchema.emptySchema().withHeader();
MappingIterator<Map<String, String>> it = mapper.readerFor(Map.class)
        .with(schema)
        .readValues(csv);
        
Map<String, String> items = it.next();
items.remove("Name");
items.remove("Age");
items.remove("Height");
//ok, it prints {Item1=ABC, Item2=DEF, Item3=GHI, Item4=JKL, Item5=MNO}
System.out.println(items);