org.apache.druid.segment.indexing.granularity

Interface GranularitySpec

All Known Implementing Classes:

ArbitraryGranularitySpec, BaseGranularitySpec, UniformGranularitySpec

public interface GranularitySpec

Tells the indexer how to group events based on timestamp. The events may then be further partitioned based on anything, using a ShardSpec.

Method Summary

Modifier and Type	Method and Description
Map<String,Object>	asMap(com.fasterxml.jackson.databind.ObjectMapper objectMapper)
com.google.common.base.Optional<org.joda.time.Interval>	bucketInterval(org.joda.time.DateTime dt) Time-grouping interval corresponding to some instant, if any.
Granularity	getQueryGranularity()
Granularity	getSegmentGranularity()
List<org.joda.time.Interval>	inputIntervals() Returns user provided intervals as-is state.
boolean	isRollup()
TreeSet<org.joda.time.Interval>	materializedBucketIntervals() This is a helper method for areas of the code, not in the overlord, were for performance reasons might need the materialized set of bucket intervals
Iterable<org.joda.time.Interval>	sortedBucketIntervals() Iterable all time groups, broken up on segment boundaries.
GranularitySpec	withIntervals(List<org.joda.time.Interval> inputIntervals)

Method Detail

sortedBucketIntervals

Iterable<org.joda.time.Interval> sortedBucketIntervals()

Iterable all time groups, broken up on segment boundaries. Should be sorted by interval start and non-overlapping.

Returns:

Iterable of all time groups

inputIntervals

List<org.joda.time.Interval> inputIntervals()

Returns user provided intervals as-is state. used for configuring granular path spec

Returns:

bucketInterval

com.google.common.base.Optional<org.joda.time.Interval> bucketInterval(org.joda.time.DateTime dt)

Time-grouping interval corresponding to some instant, if any.

Parameters:

dt - instant to return time interval for

Returns:

optional time interval

materializedBucketIntervals

TreeSet<org.joda.time.Interval> materializedBucketIntervals()

This is a helper method for areas of the code, not in the overlord, were for performance reasons might need the materialized set of bucket intervals

Returns:

A fast lookup, ordered set, of the materialized bucket interval

getSegmentGranularity

Granularity getSegmentGranularity()

isRollup

boolean isRollup()

getQueryGranularity

Granularity getQueryGranularity()

withIntervals

GranularitySpec withIntervals(List<org.joda.time.Interval> inputIntervals)

asMap

Map<String,Object> asMap(com.fasterxml.jackson.databind.ObjectMapper objectMapper)