org.cloudgraph.hbase.service

Class GraphQuery

java.lang.Object

org.cloudgraph.hbase.service.GraphQuery

All Implemented Interfaces:

org.plasma.sdo.access.QueryDispatcher

Direct Known Subclasses:

GraphStreamQuery

public class GraphQuery
extends Object
implements org.plasma.sdo.access.QueryDispatcher

Assembles and returns one or more data graphs from HBase given a PlasmaQuery™ based XPath or DSL query. First an HBase row filter is assembled using PredicateRowFilterAssembler which uses query literal values and logic found in the 'where' clause or predicate(s).

Any "slice" of a graph or set of sub-graphs can be selected using the PlasmaQuery™ API by specifying paths through the graph. Paths may include any number of predicates along the path. Based on this selection criteria an GraphFetchColumnFilterAssembler is used to precisely restrict the HBase columns returned for each result row.

Then for each resulting HBase row, a data graph assembler is used to reconstruct and return the original graph structure from the resulting HBase row.

The PlasmaQuery™ API provides a flexible mechanism to fully describe any arbitrary SDO results Data Graph, independent of any persistence framework or type of data store. PlasmaQuery™ supports XPath expressions as a free-text "surface language", parsed by the API implementation and used to construct an underlying query object model representation. As an alternative to free-text, PlasmaQuery™ contains a query Domain Specific Language (DSL) generator and API facilitating (IDE) code-completion, 100% compile-time checking and resulting in code with an almost "fluent" English appearance based on your business model.

Since:

0.5

Author:

Scott Cinnamond

See Also:

Query, PredicateRowFilterAssembler, SimpleGraphAssembler, GraphFetchColumnFilterAssembler

Field Summary

Fields

Modifier and Type	Field and Description
protected ServiceContext	context

Constructor Summary

Constructors

Constructor and Description
GraphQuery(ServiceContext context)

Method Summary

Modifier and Type	Method and Description
protected boolean	canAbortScan(boolean hasOrdering, Integer startRange, Integer endRange, Set<org.plasma.sdo.PlasmaDataGraph> result)
void	close()
protected void	collectRowKeyProperties(org.plasma.query.collector.SelectionCollector collector, org.plasma.sdo.PlasmaType type, ServiceContext context) Recursively collects row key properties adding them to the given collector for the given type.
int	count(org.plasma.query.model.Query query) Returns a count of the given query.
protected ResultsAssembler	createResultsAssembler(org.plasma.query.model.Query query, org.plasma.query.collector.SelectionCollector selection, Expr whereSyntaxTree, ResultsComparator orderingComparator, ResultsComparator groupingComparator, Expr havingSyntaxTree, TableReader rootTableReader, GraphAssemblerFactory assemblerFactory)
protected HBaseFilterAssembler	createRootColumnFilterAssembler(org.plasma.sdo.PlasmaType type, org.plasma.query.collector.SelectionCollector collector, ServiceContext context) Create and add a column filter for the initial column set based on existence of path predicates in the Select.
protected org.apache.hadoop.hbase.client.Scan	createScan(FuzzyRowKey fuzzyScan, org.apache.hadoop.hbase.filter.Filter columnFilter)
protected org.apache.hadoop.hbase.client.Scan	createScan(PartialRowKey partialRowKey, org.apache.hadoop.hbase.filter.Filter columnFilter)
protected void	execute(CompleteRowKey rowKey, TableReader rootTableReader, org.apache.hadoop.hbase.filter.Filter columnFilter, ResultsAssembler collector)
protected void	execute(FuzzyRowKey fuzzyScan, TableReader rootTableReader, org.apache.hadoop.hbase.filter.Filter columnFilter, ResultsAssembler collector)
protected void	execute(org.apache.hadoop.hbase.client.Get get, TableReader rootTableReader, ResultsAssembler collector)
protected void	execute(List<CompleteRowKey> rowKeys, TableReader rootTableReader, org.apache.hadoop.hbase.filter.Filter columnFilter, ResultsAssembler collector)
protected void	execute(List<org.apache.hadoop.hbase.client.Get> gets, TableReader rootTableReader, ResultsAssembler collector)
protected void	execute(PartialRowKey partialRowKey, TableReader rootTableReader, org.apache.hadoop.hbase.filter.Filter columnFilter, ResultsAssembler collector)
protected org.plasma.sdo.PlasmaDataGraph[]	execute(org.plasma.query.model.Query query, org.plasma.sdo.PlasmaType type, org.plasma.query.collector.SelectionCollector selection, org.apache.hadoop.hbase.filter.Filter columnFilter, Expr whereSyntaxTree, ResultsComparator orderingComparator, ResultsComparator groupingComparator, Expr havingSyntaxTree, List<PartialRowKey> partialScans, List<FuzzyRowKey> fuzzyScans, List<CompleteRowKey> completeKeys, Timestamp snapshotDate)
protected void	execute(org.apache.hadoop.hbase.client.Scan scan, TableReader rootTableReader, ResultsAssembler collector)
org.plasma.sdo.PlasmaDataGraph[]	find(org.plasma.query.model.Query query, int requestMax, Timestamp snapshotDate)
org.plasma.sdo.PlasmaDataGraph[]	find(org.plasma.query.model.Query query, Timestamp snapshotDate)
protected List	getVariables(org.plasma.query.model.Where where)
protected void	log(org.plasma.query.model.Query query)
protected String	serializeGraph(commonj.sdo.DataGraph graph)

Methods inherited from class java.lang.Object

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

Field Detail

context

protected ServiceContext context

Constructor Detail

GraphQuery

public GraphQuery(ServiceContext context)

Method Detail

close

public void close()

Specified by:

close in interface org.plasma.sdo.access.QueryDispatcher

find

public org.plasma.sdo.PlasmaDataGraph[] find(org.plasma.query.model.Query query,
                                             Timestamp snapshotDate)

Specified by:

find in interface org.plasma.sdo.access.QueryDispatcher

find

public org.plasma.sdo.PlasmaDataGraph[] find(org.plasma.query.model.Query query,
                                             int requestMax,
                                             Timestamp snapshotDate)

Specified by:

find in interface org.plasma.sdo.access.QueryDispatcher

count

public int count(org.plasma.query.model.Query query)

Returns a count of the given query.

Specified by:

count in interface org.plasma.sdo.access.QueryDispatcher

Parameters:

query - the query

Returns:

the query results size

execute

protected org.plasma.sdo.PlasmaDataGraph[] execute(org.plasma.query.model.Query query,
                                                   org.plasma.sdo.PlasmaType type,
                                                   org.plasma.query.collector.SelectionCollector selection,
                                                   org.apache.hadoop.hbase.filter.Filter columnFilter,
                                                   Expr whereSyntaxTree,
                                                   ResultsComparator orderingComparator,
                                                   ResultsComparator groupingComparator,
                                                   Expr havingSyntaxTree,
                                                   List<PartialRowKey> partialScans,
                                                   List<FuzzyRowKey> fuzzyScans,
                                                   List<CompleteRowKey> completeKeys,
                                                   Timestamp snapshotDate)

createResultsAssembler

protected ResultsAssembler createResultsAssembler(org.plasma.query.model.Query query,
                                                  org.plasma.query.collector.SelectionCollector selection,
                                                  Expr whereSyntaxTree,
                                                  ResultsComparator orderingComparator,
                                                  ResultsComparator groupingComparator,
                                                  Expr havingSyntaxTree,
                                                  TableReader rootTableReader,
                                                  GraphAssemblerFactory assemblerFactory)

canAbortScan

protected boolean canAbortScan(boolean hasOrdering,
                               Integer startRange,
                               Integer endRange,
                               Set<org.plasma.sdo.PlasmaDataGraph> result)

execute

protected void execute(PartialRowKey partialRowKey,
                       TableReader rootTableReader,
                       org.apache.hadoop.hbase.filter.Filter columnFilter,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

createScan

protected org.apache.hadoop.hbase.client.Scan createScan(PartialRowKey partialRowKey,
                                                         org.apache.hadoop.hbase.filter.Filter columnFilter)

execute

protected void execute(CompleteRowKey rowKey,
                       TableReader rootTableReader,
                       org.apache.hadoop.hbase.filter.Filter columnFilter,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

execute

protected void execute(List<CompleteRowKey> rowKeys,
                       TableReader rootTableReader,
                       org.apache.hadoop.hbase.filter.Filter columnFilter,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

execute

protected void execute(FuzzyRowKey fuzzyScan,
                       TableReader rootTableReader,
                       org.apache.hadoop.hbase.filter.Filter columnFilter,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

createScan

protected org.apache.hadoop.hbase.client.Scan createScan(FuzzyRowKey fuzzyScan,
                                                         org.apache.hadoop.hbase.filter.Filter columnFilter)
                                                  throws IOException

Throws:

IOException

execute

protected void execute(org.apache.hadoop.hbase.client.Get get,
                       TableReader rootTableReader,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

execute

protected void execute(List<org.apache.hadoop.hbase.client.Get> gets,
                       TableReader rootTableReader,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

execute

protected void execute(org.apache.hadoop.hbase.client.Scan scan,
                       TableReader rootTableReader,
                       ResultsAssembler collector)
                throws IOException

Throws:

IOException

createRootColumnFilterAssembler

protected HBaseFilterAssembler createRootColumnFilterAssembler(org.plasma.sdo.PlasmaType type,
                                                               org.plasma.query.collector.SelectionCollector collector,
                                                               ServiceContext context)

Create and add a column filter for the initial column set based on existence of path predicates in the Select. If no path predicates exist in the selection, we know the entire selection graph can be fetched in one round trip, so GraphFetchColumnFilterAssembler is create. Otherwise a (@link InitialFetchColumnFilterAssembler) is used to populate the root(s), and then subsequent fetches are used as directed within the graph assembler facilitated by the path predicates.

Parameters:

type - the root type

collector - the selection collector

Returns:

the new filter assembler

See Also:

GraphFetchColumnFilterAssembler, InitialFetchColumnFilterAssembler

collectRowKeyProperties

protected void collectRowKeyProperties(org.plasma.query.collector.SelectionCollector collector,
                                       org.plasma.sdo.PlasmaType type,
                                       ServiceContext context)

Recursively collects row key properties adding them to the given collector for the given type. Any additional properties associated with types discovered during traversal of user defined row key field paths are added recursively.

Parameters:

collector - the property collector

type - the current type

getVariables

protected List getVariables(org.plasma.query.model.Where where)

log

protected void log(org.plasma.query.model.Query query)

serializeGraph

protected String serializeGraph(commonj.sdo.DataGraph graph)
                         throws IOException

Throws:

IOException