After converting to dto using mapstruct, child id order jumbles up. If i return just the entity without converting to dto the child is properly ordered
"data":[ { "personName":"TestUser", "cars":[ { "id":"2021-000001" }, { "id":"2021-000003" }, { "id":"2021-000002" }, { "id":"2021-000004" } } ]
The reason why the order is different is due to the fact that MapStruct used a HashSet by default when mapping into a Set.
The current implementation types for MapStruct are the following:
| Interface type | Implementation type |
|---|---|
Iterable |
ArrayList |
Collection |
ArrayList |
List |
ArrayList |
Set |
HashSet |
SortedSet |
TreeSet |
NavigableSet |
TreeSet |
Map |
HashMap |
SortedMap |
TreeMap |
NavigableMap |
TreeMap |
ConcurrentMap |
ConcurrentHashMap |
ConcurrentNavigableMap |
ConcurrentSkipListMap |
This can be seen in Implementation types used for collection mappings
There is currently a feature request to change this into an order preserving set / map and it will be included in 1.5.
At the moment you can fix this by using an @ObjectFactory that will return a LinkedHashSet
