Apache Commons CSV doesn't ignore missing column

Question

Using Apache Commons CSV for parsing, but doesn't ignore missing column and throws exception.

with this sample data:

name age
Ali 35
John 25
Vahid 75

Below code record.get(DataColumns.surname) throws java.lang.IllegalArgumentException: Mapping for surname not found, expected one of [name, surname, age]. I need it returns null, optional or default value. Is there any option? I know it is possible with record.toMap().get(DataColumns.surname.name()) but its performance will not be good:

...
enum DataColumns { name, surname, age }
...
Reader in = new BufferedReader(new FileReader(fileName));

try (CSVParser records = CSVFormat.TDF
                .withDelimiter(' ')
                .withIgnoreSurroundingSpaces()
                .withAllowDuplicateHeaderNames(false)
                .withIgnoreHeaderCase()
                .withTrim()
                .withHeader(DataColumns.class)
                .withFirstRecordAsHeader()
                .withSkipHeaderRecord()
                .withAllowMissingColumnNames(false)
                .withIgnoreEmptyLines()
                .parse(in)) {

   for (CSVRecord record : records) {
       String name = record.get(DataColumns.name);
       String surname = record.get(DataColumns.surname);
       Short age = Short.valueOf(record.get(DataColumns.age)); 
   }
}

...

Answer 1

You might try using record.isMapped(columnName) to check if the column exists, recording into a variable so you don't have to check again every line.

Another option would be to use records.getHeaderNames() and store it into a variable once, before the loop, maybe even using a Set<String> for an extra kick of existance checking performance: Set<String> headerNames = new HashSet<>(records.getHeaderNames()).

Then, you can use the resulting variable inside the loop by calling headerNames.contains(columnName) to check whether the column exists or not.

Plese, see: https://javadoc.io/doc/org.apache.commons/commons-csv/latest/org/apache/commons/csv/CSVRecord.html