com.yahoo.bard.webservice.data

Class DruidQueryBuilder

java.lang.Object

com.yahoo.bard.webservice.data.DruidQueryBuilder

@Singleton
public class DruidQueryBuilder
extends Object

Druid Query builder class.

Field Summary

Fields

Modifier and Type	Field and Description
protected PhysicalTableResolver	resolver
protected LogicalTableDictionary	tableDictionary

Constructor Summary

Constructors

Constructor and Description
DruidQueryBuilder(LogicalTableDictionary tableDictionary, PhysicalTableResolver resolver) Constructor.

Method Summary

Modifier and Type	Method and Description
protected GroupByQuery	buildGroupByQuery(TemplateDruidQuery template, ConstrainedTable table, Granularity granularity, org.joda.time.DateTimeZone timeZone, Set<Dimension> groupByDimensions, Filter filter, Having having, Set<org.joda.time.Interval> intervals, LimitSpec druidOrderBy) Builds a druid groupBy query recursively nesting dataSource based on the TemplateDruidQuery.
DruidAggregationQuery<?>	buildQuery(DataApiRequest request, TemplateDruidQuery template) Build a druid query object from an API request and it's templateDruidQuery.
protected TimeSeriesQuery	buildTimeSeriesQuery(TemplateDruidQuery template, ConstrainedTable table, Granularity granularity, org.joda.time.DateTimeZone timeZone, Filter filter, Set<org.joda.time.Interval> intervals) Builds a druid TimeSeries query.
protected TopNQuery	buildTopNQuery(TemplateDruidQuery template, ConstrainedTable table, Granularity granularity, org.joda.time.DateTimeZone timeZone, Set<Dimension> groupByDimension, Filter filter, Set<org.joda.time.Interval> intervals, TopNMetric metricSpec, int topN) Builds a druid topN query.
protected boolean	canOptimizeTimeSeries(DataApiRequest apiRequest, TemplateDruidQuery templateDruidQuery) Determine if the optimization to a Timeseries query can be done.
protected boolean	canOptimizeTopN(DataApiRequest apiRequest, TemplateDruidQuery templateDruidQuery) Determine if the optimization to a TopN query can be done.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

tableDictionary

protected final LogicalTableDictionary tableDictionary

resolver

protected final PhysicalTableResolver resolver

Constructor Detail

DruidQueryBuilder

@Inject
public DruidQueryBuilder(LogicalTableDictionary tableDictionary,
                                 PhysicalTableResolver resolver)

Constructor.

Parameters:

tableDictionary - Dictionary of logical tables used to look up table groups

resolver - Strategy for resolving the physical table

Method Detail

buildQuery

public DruidAggregationQuery<?> buildQuery(DataApiRequest request,
                                           TemplateDruidQuery template)
                                    throws DimensionRowNotFoundException,
                                           NoMatchFoundException

Build a druid query object from an API request and it's templateDruidQuery.

Parameters:

request - DataApiRequest to use in building the query

template - TemplateDruidQuery to build out the query with

Returns:

a DruidAggregationQuery

Throws:

DimensionRowNotFoundException - if the filters filter out all dimension rows

NoMatchFoundException - if no PhysicalTable satisfies this request

buildGroupByQuery

protected GroupByQuery buildGroupByQuery(TemplateDruidQuery template,
                                         ConstrainedTable table,
                                         Granularity granularity,
                                         org.joda.time.DateTimeZone timeZone,
                                         Set<Dimension> groupByDimensions,
                                         Filter filter,
                                         Having having,
                                         Set<org.joda.time.Interval> intervals,
                                         LimitSpec druidOrderBy)

Builds a druid groupBy query recursively nesting dataSource based on the TemplateDruidQuery.

Parameters:

template - The query template, possibly nested

table - The physical table that underlies the lowest-level datasource

granularity - The granularity from the request

timeZone - The time zone from the request

groupByDimensions - The grouping dimensions from the request

filter - The filters specified in the request (only applied at the lowest level)

having - The having clause specified in the request.

intervals - The intervals specified from the request

druidOrderBy - The order by

Returns:

a GroupByQuery

buildTopNQuery

protected TopNQuery buildTopNQuery(TemplateDruidQuery template,
                                   ConstrainedTable table,
                                   Granularity granularity,
                                   org.joda.time.DateTimeZone timeZone,
                                   Set<Dimension> groupByDimension,
                                   Filter filter,
                                   Set<org.joda.time.Interval> intervals,
                                   TopNMetric metricSpec,
                                   int topN)

Builds a druid topN query.

Parameters:

template - The query template. Not nested since nesting is not supported in druid topN queries

table - The physical table that underlies the lowest-level datasource

granularity - The grain from the request

timeZone - The time zone from the request

groupByDimension - The grouping dimension from the request

filter - The filters specified in the request

intervals - The intervals specified from the request

metricSpec - The topn metric spec

topN - The number of requested top entries per time bucket

Returns:

a TopNQuery

buildTimeSeriesQuery

protected TimeSeriesQuery buildTimeSeriesQuery(TemplateDruidQuery template,
                                               ConstrainedTable table,
                                               Granularity granularity,
                                               org.joda.time.DateTimeZone timeZone,
                                               Filter filter,
                                               Set<org.joda.time.Interval> intervals)

Builds a druid TimeSeries query.

Parameters:

template - The query template. Not nested since nesting is not supported in druid timeseries queries

table - The physical table that underlies the lowest-level datasource

granularity - The grain from the request

timeZone - The time zone from the request

filter - The filters specified in the request

intervals - The intervals specified from the request

Returns:

a TimeSeriesQuery

canOptimizeTopN

protected boolean canOptimizeTopN(DataApiRequest apiRequest,
                                  TemplateDruidQuery templateDruidQuery)

Determine if the optimization to a TopN query can be done.

Parameters:

apiRequest - The request data

templateDruidQuery - The template query

Returns:

true if the optimization can be done, false if it can't

canOptimizeTimeSeries

protected boolean canOptimizeTimeSeries(DataApiRequest apiRequest,
                                        TemplateDruidQuery templateDruidQuery)

Determine if the optimization to a Timeseries query can be done.

Parameters:

apiRequest - The request data

templateDruidQuery - The template query

Returns:

true if the optimization can be done, false if it can't