rovats/java-utils

A collection of useful Java utilities

java-utils's Issues

Batch collector: Return number of records processed at the end

License?

What is the license on this?

Add a strict mode for parallel streams

Currently elements collected by each thread are passed to the batch processor without checking the batch size (to save merging the two lists), but if strict mode is enabled, the batch processor should only be called with elements exactly equal to the batch size.

Add distinctByKeyCollector

When dealing with a stream of objects that don't override equals, distinct() and Collectors.toSet() won't work.

But if they have a unique key, a distinctByKeyCollector can be used.

Do not call batchprocessor with empty list

If the stream size is a multiple of the batch size (e.g. stream size = 20, batch size = 10), the finishers calls the batchProcessor with empty list. I'm thinking this may not be necessary? (it's actually causing issue with my batchprocessor as it does not expect empty list)

Adding the empty check may help, but I'm not sure there are other implications of this.

public Function<List<T>, List<T>> finisher() {
        return ts -> {
            if (!ts.isEmpty()) { 
                batchProcessor.accept(ts);
                recordsProcessed += ts.size();
            }
            return Collections.emptyList();
        };
    }

Thanks
Michal

Add more tests and documentation

Test the main functionality

Hi, your tests do not test what is most important - chunking. I your implementation would return the input list, tests would pass.