Background
Repository queries taking collection parameters are limited to 32K collection items. (This is caused by Hibernate PostgreSQL driver having limit of 32767 bind variables per statement.)
For example, org.onap.cps.spi.repository.FragmentRepository#deleteByAnchorIn:
Collection<AnchorEntity> findAllByDataspaceAndNameIn(DataspaceEntity dataspaceEntity, Collection<String> anchorNames);
This would throw an exception if supplied with a collection of 32,768 items.
CPS-1574 is to remove those limits (and write tests to verify).
Proposal #1: List partitioning
This solution partitions the input collection into lists of 32,000 and calls the original repository method, collecting the results:
Collection<AnchorEntity> findAllByDataspaceAndNameIn(DataspaceEntity dataspaceEntity, Collection<String> anchorNames);
default Collection<AnchorEntity> findAllByDataspaceAndNameInBatch(final DataspaceEntity dataspaceEntity,
final List<String> anchorNames) {
Collection<AnchorEntity> results = new ArrayList<>(anchorNames.size());
for (final Collection<String> anchorNamesBatch : Lists.partition(anchorNames, 32000)) {
results.addAll(findAllByDataspaceAndNameIn(dataspaceEntity, anchorNamesBatch));
}
return results;
}
In addition, if we wish to avoid the overhead of adding the results to a collection when not needed, we may directly call the original when the collection size is less than 32,000:
Collection<AnchorEntity> findAllByDataspaceAndNameIn(DataspaceEntity dataspaceEntity, Collection<String> anchorNames);
default Collection<AnchorEntity> findAllByDataspaceAndNameInBatch(final DataspaceEntity dataspaceEntity,
final List<String> anchorNames) {
if (anchorNames.size() < 32000) {
return findAllByDataspaceAndNameIn(dataspaceEntity, anchorNames);
}
Collection<AnchorEntity> results = new ArrayList<>(anchorNames.size());
for (final Collection<String> anchorNamesBatch : Lists.partition(anchorNames, 32000)) {
results.addAll(findAllByDataspaceAndNameIn(dataspaceEntity, anchorNamesBatch));
}
return results;
}