Github user dawidwys commented on a diff in the pull request:

    https://github.com/apache/flink/pull/1848#discussion_r60392270
  
    --- Diff: 
flink-tests/src/test/java/org/apache/flink/test/javaApiOperators/PartitionITCase.java
 ---
    @@ -546,43 +549,264 @@ public void testRangePartitionInIteration() throws 
Exception {
                result.collect(); // should fail
        }
     
    +
    +
    +   @Test
    +   public void testRangePartitionerOnSequenceDataWithOrders() throws 
Exception {
    +           final ExecutionEnvironment env = 
ExecutionEnvironment.getExecutionEnvironment();
    +           DataSet<Tuple2<Long, Long>> dataSet = env.generateSequence(0, 
10000)
    +                           .map(new MapFunction<Long, Tuple2<Long, 
Long>>() {
    +                   @Override
    +                   public Tuple2<Long, Long> map(Long value) throws 
Exception {
    +                           return new Tuple2<>(value / 5000, value % 5000);
    +                   }
    +           });
    +
    +           final Tuple2Comparator<Long> tuple2Comparator = new 
Tuple2Comparator<>(new LongComparator(true),
    +                                                                           
                                                                           new 
LongComparator(false));
    +
    +           MinMaxSelector<Tuple2<Long, Long>> minMaxSelector = new 
MinMaxSelector<>(tuple2Comparator);
    +
    +           final List<Tuple2<Tuple2<Long, Long>, Tuple2<Long, Long>>> 
collected = dataSet.partitionByRange(0, 1)
    +                           .withOrders(Order.ASCENDING, Order.DESCENDING)
    +                           .mapPartition(minMaxSelector)
    +                           .collect();
    +
    +           Collections.sort(collected, new 
Tuple2Comparator<>(tuple2Comparator));
    +
    +           Tuple2<Long, Long> previousMax = null;
    +           for (Tuple2<Tuple2<Long, Long>, Tuple2<Long, Long>> tuple2 : 
collected) {
    --- End diff --
    
    In fact I tried to do similar as what you described with some differences.
    Ad. 1  I performed this step in two different places - I simplified it a 
bit and added a comment.
    Ad.2  I sort on both fields but don't think it is a problem, as the min 
value is of greater importance.
    Ad.3 In fact I forgot about this step which I added.
    4. I also check for cases like described in my response to your comment on 
line 809 regarding the Tuple2Comparator. Ensuring that partitioning on the 
second part of composite key is done properly.


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

Reply via email to