/*
* Parse the user-data supplied by the HiveFragmenter from InputData. Based on the
* user-data construct the partition fields and the InputFormat for current split
*/
private InputFormat<?, ?> createInputFormat(InputData input) throws Exception {
String userData = new String(input.getFragmentUserData());
String[] toks = userData.split(HiveDataFragmenter.HIVE_UD_DELIM);
initPartitionFields(toks[3]);
filterInFragmenter = new Boolean(toks[4]);
return HiveDataFragmenter.makeInputFormat(toks[0] /* inputFormat name */, jobConf);
}
/*
* We are testing one filter against all the partition fields.
* The filter has the form "fieldA = valueA".
* The partitions have the form partitionOne=valueOne/partitionTwo=ValueTwo/partitionThree=valueThree
* 1. For a filter to match one of the partitions, lets say partitionA for example, we need:
* fieldA = partittionOne and valueA = valueOne. If this condition occurs, we return true.
* 2. If fieldA does not match any one of the partition fields we also return true, it means we ignore this filter
* because it is not on a partition field.
* 3. If fieldA = partittionOne and valueA != valueOne, then we return false.
*/
private boolean testOneFilter(
List<HivePartition> partitionFields, Object filter, InputData input) {
// Let's look first at the filter
FilterParser.BasicFilter bFilter = (FilterParser.BasicFilter) filter;
boolean isFilterOperationEqual = (bFilter.getOperation() == FilterParser.Operation.HDOP_EQ);
if (!isFilterOperationEqual) /* in case this is not an "equality filter" we ignore it here - in partition filtering */ {
return true;
}
int filterColumnIndex = bFilter.getColumn().index();
String filterValue = bFilter.getConstant().constant().toString();
ColumnDescriptor filterColumn = input.getColumn(filterColumnIndex);
String filterColumnName = filterColumn.columnName();
for (HivePartition partition : partitionFields) {
if (filterColumnName.equals(partition.name)) {
/* the filter field matches a partition field, but the values do not match */
return filterValue.equals(partition.val);
}
}
/* filter field did not match any partition field, so we ignore this filter and hence return true */
return true;
}
