org.apache.metamodel.elasticsearch

Class ElasticSearchDataContext

java.lang.Object

org.apache.metamodel.AbstractDataContext

org.apache.metamodel.QueryPostprocessDataContext

org.apache.metamodel.elasticsearch.ElasticSearchDataContext

All Implemented Interfaces:

HasReadTypeConverters, DataContext, UpdateableDataContext

public class ElasticSearchDataContext
extends QueryPostprocessDataContext
implements DataContext, UpdateableDataContext

DataContext implementation for ElasticSearch analytics engine. ElasticSearch has a data storage structure hierarchy that briefly goes like this:

• Index
• Document type (short: Type) (within an index)
• Documents (of a particular type)

When instantiating this DataContext, an index name is provided. Within this index, each document type is represented as a table. This implementation supports either automatic discovery of a schema or manual specification of a schema, through the SimpleTableDef class.

Field Summary

Fields

Modifier and Type	Field and Description
static String	FIELD_ID
static org.elasticsearch.common.unit.TimeValue	TIMEOUT_SCROLL

Fields inherited from class org.apache.metamodel.QueryPostprocessDataContext

INFORMATION_SCHEMA_NAME

Constructor Summary

Constructors

Constructor and Description
ElasticSearchDataContext(org.elasticsearch.client.Client client, String indexName) Constructs a ElasticSearchDataContext and automatically detects the schema structure/view on all indexes (see detectSchema(Client, String)).
ElasticSearchDataContext(org.elasticsearch.client.Client client, String indexName, SimpleTableDef... tableDefs) Constructs a ElasticSearchDataContext.

Method Summary

Methods

Modifier and Type	Method and Description
protected org.elasticsearch.index.query.QueryBuilder	createQueryBuilderForSimpleWhere(Table table, List<FilterItem> whereItems, LogicalOperator logicalOperator) Creates, if possible, a QueryBuilder object which can be used to push down one or more FilterItems to ElasticSearch's backend.
static SimpleTableDef[]	detectSchema(org.elasticsearch.client.Client client, String indexName) Performs an analysis of the available indexes in an ElasticSearch cluster Client instance and detects the elasticsearch types structure based on the metadata provided by the ElasticSearch java client.
static SimpleTableDef	detectTable(org.elasticsearch.cluster.ClusterState cs, String indexName, String documentType) Performs an analysis of an available index type in an ElasticSearch Client client and tries to detect the index structure based on the metadata provided by the java client.
protected Number	executeCountQuery(Table table, List<FilterItem> whereItems, boolean functionApproximationAllowed) Executes a simple count query, if possible.
protected Row	executePrimaryKeyLookupQuery(Table table, List<SelectItem> selectItems, Column primaryKeyColumn, Object keyValue) Executes a query which obtains a row by primary key (as defined by Column.isPrimaryKey()).
void	executeUpdate(UpdateScript update) Submits an UpdateScript for execution on the DataContext.
org.elasticsearch.client.Client	getElasticSearchClient() Gets the Client that this DataContext is wrapping.
String	getIndexName() Gets the name of the index that this DataContext is working on.
protected Schema	getMainSchema()
protected String	getMainSchemaName()
protected DataSet	materializeMainSchemaTable(Table table, Column[] columns, int maxRows) Executes a simple one-table query against a table in the main schema of the subclasses of this class.
protected DataSet	materializeMainSchemaTable(Table table, List<SelectItem> selectItems, List<FilterItem> whereItems, int firstRow, int maxRows) Execute a simple one-table query against a table in the main schema of the subclasses of this class.

Methods inherited from class org.apache.metamodel.QueryPostprocessDataContext

addConverter, executeQuery, getDefaultSchemaName, getMainSchemaInternal, getSchemaByNameInternal, getSchemaNamesInternal, isMainSchemaTable, isScalarFunctionMaterialized, materializeFromItem, materializeMainSchemaTable, materializeMainSchemaTable, materializeTable, materializeTable

Methods inherited from class org.apache.metamodel.AbstractDataContext

compileQuery, executeQuery, executeQuery, getColumnByQualifiedLabel, getDefaultSchema, getSchemaByName, getSchemaNames, getSchemas, getTableByQualifiedLabel, isQualifiedPathDelim, onSchemaCacheRefreshed, parseQuery, query, refreshSchemas

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.metamodel.DataContext

compileQuery, executeQuery, executeQuery, executeQuery, getColumnByQualifiedLabel, getDefaultSchema, getSchemaByName, getSchemaNames, getSchemas, getTableByQualifiedLabel, parseQuery, query, refreshSchemas

Field Detail

FIELD_ID

public static final String FIELD_ID

See Also:

Constant Field Values

TIMEOUT_SCROLL

public static final org.elasticsearch.common.unit.TimeValue TIMEOUT_SCROLL

Constructor Detail

ElasticSearchDataContext

public ElasticSearchDataContext(org.elasticsearch.client.Client client,
                        String indexName,
                        SimpleTableDef... tableDefs)

Constructs a ElasticSearchDataContext. This constructor accepts a custom array of SimpleTableDefs which allows the user to define his own view on the indexes in the engine.

Parameters:

client - the ElasticSearch client

indexName - the name of the ElasticSearch index to represent

tableDefs - an array of SimpleTableDefs, which define the table and column model of the ElasticSearch index.

ElasticSearchDataContext

public ElasticSearchDataContext(org.elasticsearch.client.Client client,
                        String indexName)

Constructs a ElasticSearchDataContext and automatically detects the schema structure/view on all indexes (see detectSchema(Client, String)).

Parameters:

client - the ElasticSearch client

indexName - the name of the ElasticSearch index to represent

Method Detail

detectSchema

public static SimpleTableDef[] detectSchema(org.elasticsearch.client.Client client,
                            String indexName)

Performs an analysis of the available indexes in an ElasticSearch cluster Client instance and detects the elasticsearch types structure based on the metadata provided by the ElasticSearch java client.

Parameters:

client - the client to inspect

indexName -

Returns:

a mutable schema instance, useful for further fine tuning by the user.

See Also:

detectTable(ClusterState, String, String)

detectTable

public static SimpleTableDef detectTable(org.elasticsearch.cluster.ClusterState cs,
                         String indexName,
                         String documentType)
                                  throws Exception

Performs an analysis of an available index type in an ElasticSearch Client client and tries to detect the index structure based on the metadata provided by the java client.

Parameters:

cs - the ElasticSearch cluster

indexName - the name of the index

documentType - the name of the index type

Returns:

a table definition for ElasticSearch.

Throws:

Exception

getMainSchema

protected Schema getMainSchema()
                        throws MetaModelException

Specified by:

getMainSchema in class QueryPostprocessDataContext

Returns:

the main schema that subclasses of this class produce

Throws:

MetaModelException

getMainSchemaName

protected String getMainSchemaName()
                            throws MetaModelException

Specified by:

getMainSchemaName in class QueryPostprocessDataContext

Returns:

the name of the main schema that subclasses of this class produce

Throws:

MetaModelException

materializeMainSchemaTable

protected DataSet materializeMainSchemaTable(Table table,
                                 List<SelectItem> selectItems,
                                 List<FilterItem> whereItems,
                                 int firstRow,
                                 int maxRows)

Description copied from class: QueryPostprocessDataContext

Execute a simple one-table query against a table in the main schema of the subclasses of this class. This default implementation will delegate to QueryPostprocessDataContext.materializeMainSchemaTable(Table, List, int, int) and apply WHERE item filtering afterwards.

Overrides:

materializeMainSchemaTable in class QueryPostprocessDataContext

Returns:

materializeMainSchemaTable

protected DataSet materializeMainSchemaTable(Table table,
                                 Column[] columns,
                                 int maxRows)

Description copied from class: QueryPostprocessDataContext

Executes a simple one-table query against a table in the main schema of the subclasses of this class.

Specified by:

materializeMainSchemaTable in class QueryPostprocessDataContext

Parameters:

table - the table to query

columns - the columns of the table to query

maxRows - the maximum amount of rows needed or -1 if all rows are wanted.

Returns:

a dataset with the raw table/column content.

createQueryBuilderForSimpleWhere

protected org.elasticsearch.index.query.QueryBuilder createQueryBuilderForSimpleWhere(Table table,
                                                                          List<FilterItem> whereItems,
                                                                          LogicalOperator logicalOperator)

Creates, if possible, a QueryBuilder object which can be used to push down one or more FilterItems to ElasticSearch's backend.

Parameters:

table -

whereItems -

logicalOperator -

Returns:

a QueryBuilder if one was produced, or null if the items could not be pushed down to an ElasticSearch query

executePrimaryKeyLookupQuery

protected Row executePrimaryKeyLookupQuery(Table table,
                               List<SelectItem> selectItems,
                               Column primaryKeyColumn,
                               Object keyValue)

Description copied from class: QueryPostprocessDataContext

Executes a query which obtains a row by primary key (as defined by Column.isPrimaryKey()). This method is provided to allow subclasses to optimize lookup queries since they are quite common and often a datastore can retrieve the row using some specialized means which is much more performant than scanning all records manually.

Overrides:

executePrimaryKeyLookupQuery in class QueryPostprocessDataContext

Parameters:

table - the table on which the lookup is requested.

selectItems - the items to select from the lookup query.

primaryKeyColumn - the column that is the primary key

keyValue - the primary key value that is specified in the lookup query.

Returns:

the row if the particular table, or null if not available.

executeCountQuery

protected Number executeCountQuery(Table table,
                       List<FilterItem> whereItems,
                       boolean functionApproximationAllowed)

Description copied from class: QueryPostprocessDataContext

Executes a simple count query, if possible. This method is provided to allow subclasses to optimize count queries since they are quite common and often a datastore can retrieve the count using some specialized means which is much more performant than counting all records manually.

Overrides:

executeCountQuery in class QueryPostprocessDataContext

Parameters:

table - the table on which the count is requested.

whereItems - a (sometimes empty) list of WHERE items.

functionApproximationAllowed - whether approximation is allowed or not.

Returns:

the count of the particular table, or null if not available.

executeUpdate

public void executeUpdate(UpdateScript update)

Description copied from interface: UpdateableDataContext

Submits an UpdateScript for execution on the DataContext. Since implementations of the DataContext vary quite a lot, there is no golden rule as to how an update script will be executed. But the implementors should strive towards handling an UpdateScript as a single transactional change to the data store.

Specified by:

executeUpdate in interface UpdateableDataContext

Parameters:

update - the update script to execute

getElasticSearchClient

public org.elasticsearch.client.Client getElasticSearchClient()

Gets the Client that this DataContext is wrapping.

Returns:

getIndexName

public String getIndexName()

Gets the name of the index that this DataContext is working on.

Returns: