org.apache.drill.exec.store

Class AbstractSchema

java.lang.Object

org.apache.drill.exec.store.AbstractSchema

All Implemented Interfaces:

AutoCloseable, org.apache.calcite.schema.Schema, SchemaPartitionExplorer

Direct Known Subclasses:

CapitalizingJdbcSchema, CassandraDrillSchema, CassandraRootDrillSchema, ClickhouseCatalogSchema, DruidSchemaFactory.DruidDataSources, DynamicRootSchema.RootSchema, ElasticsearchDrillSchema, FileSystemSchemaFactory.FileSystemSchema, HiveDatabaseSchema, HttpAPIConnectionSchema, HttpSchemaFactory.HttpSchema, KafkaMessageSchema, MongoDatabaseSchema, SubSchemaWrapper, WorkspaceSchemaFactory.WorkspaceSchema

public abstract class AbstractSchema
extends Object
implements org.apache.calcite.schema.Schema, SchemaPartitionExplorer, AutoCloseable

Nested Class Summary

Nested classes/interfaces inherited from interface org.apache.calcite.schema.Schema

org.apache.calcite.schema.Schema.TableType

Field Summary

Fields

Modifier and Type	Field and Description
protected String	name
static String	SCHEMA_SEPARATOR
protected List<String>	schemaPath

Constructor Summary

Constructors

Constructor and Description
AbstractSchema(List<String> parentSchemaPath, String name)

Method Summary

Modifier and Type	Method and Description
CreateTableEntry	appendToStatsTable(String tableName) Create an append statistics table entry for given tableName.
DrillTable	applyFunctionParameters(DrillTable drillTable, List<TableParamDef> paramDefs, List<Object> values) For the given list of parameters definitions executes action for the corresponding value.
boolean	areTableNamesCaseSensitive() Indicates if table names in schema are case sensitive.
void	close()
CreateTableEntry	createNewTable(String tableName, List<String> partitionColumns) Creates table entry using table name and list of partition columns if any.
CreateTableEntry	createNewTable(String tableName, List<String> partitionColumns, StorageStrategy storageStrategy) Creates table entry using table name, list of partition columns and storage strategy used to create table folder and files
CreateTableEntry	createStatsTable(String tableName) Create stats table entry for given tableName.
boolean	createView(View view) Create a new view given definition.
void	dropTable(String tableName)
void	dropView(String viewName) Drop the view with given name.
org.apache.calcite.schema.Schema	getDefaultSchema() The schema can be a top level schema which doesn't have its own tables, but refers to one of the default sub schemas for table look up.
org.apache.calcite.linq4j.tree.Expression	getExpression(org.apache.calcite.schema.SchemaPlus parentSchema, String name)
String	getFullSchemaName()
Set<String>	getFunctionNames()
List<TableParamDef>	getFunctionParameters() Returns of common table function parameters that can be used all Drill schema implementations.
Collection<org.apache.calcite.schema.Function>	getFunctions(String name) For the given table names returns list of acceptable table functions which are common for all Drill schemas.
String	getName()
List<String>	getSchemaPath()
org.apache.calcite.schema.Table	getStatsTable(String tableName) Get the statistics table for given tableName
Iterable<String>	getSubPartitions(String table, List<String> partitionColumns, List<String> partitionValues) Get a list of sub-partitions of a particular table and the partitions specified by partition columns and values.
org.apache.calcite.schema.Schema	getSubSchema(String name)
Set<String>	getSubSchemaNames()
org.apache.calcite.schema.Table	getTable(String name)
Set<String>	getTableNames()
Collection<Map.Entry<String,org.apache.calcite.schema.Schema.TableType>>	getTableNamesAndTypes() Used by InfoSchemaRecordGenerator.Tables for getting all table objects along with type for every requested schema.
List<org.apache.commons.lang3.tuple.Pair<String,? extends org.apache.calcite.schema.Table>>	getTablesByNames(Set<String> tableNames) Get the collection of Table tables specified in the tableNames.
org.apache.calcite.rel.type.RelProtoDataType	getType(String name)
protected Map<String,org.apache.calcite.rel.type.RelProtoDataType>	getTypeMap() Returns a map of types in this schema by name.
abstract String	getTypeName() Returns string describing schema type which shows where the schema came from.
Set<String>	getTypeNames()
String	getUser(String impersonated, String notImpersonated)
boolean	isMutable()
boolean	needToImpersonateReadingData() Does Drill needs to impersonate as user connected to Drill when reading data from DataSource?
boolean	showInInformationSchema() Reports whether to show items from this schema in INFORMATION_SCHEMA tables.
org.apache.calcite.schema.Schema	snapshot(org.apache.calcite.schema.SchemaVersion version)
String	toString()

Methods inherited from class java.lang.Object

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

Field Detail

SCHEMA_SEPARATOR

public static final String SCHEMA_SEPARATOR

See Also:

Constant Field Values

schemaPath

protected final List<String> schemaPath

name

protected final String name

Constructor Detail

AbstractSchema

public AbstractSchema(List<String> parentSchemaPath,
                      String name)

Method Detail

getSubPartitions

public Iterable<String> getSubPartitions(String table,
                                         List<String> partitionColumns,
                                         List<String> partitionValues)
                                  throws PartitionNotFoundException

Description copied from interface: SchemaPartitionExplorer

Get a list of sub-partitions of a particular table and the partitions specified by partition columns and values. Individual storage plugins will assign specific meaning to the parameters and return values.

For more info see docs in PartitionExplorer.

Specified by:

getSubPartitions in interface SchemaPartitionExplorer

partitionColumns - a list of partitions to match

partitionValues - list of values of each partition (corresponding to the partition column list)

Returns:

list of sub-partitions, will be empty if a there is no further level of sub-partitioning below, i.e. hit a leaf partition

Throws:

PartitionNotFoundException - when the partition does not exist in the given workspace

getName

public String getName()

getSchemaPath

public List<String> getSchemaPath()

getFullSchemaName

public String getFullSchemaName()

toString

public final String toString()

Overrides:

toString in class Object

getTypeName

public abstract String getTypeName()

Returns string describing schema type which shows where the schema came from. Good practice here is to return json type name of storage plugin's config.

Returns:

schema type name

getDefaultSchema

public org.apache.calcite.schema.Schema getDefaultSchema()

The schema can be a top level schema which doesn't have its own tables, but refers to one of the default sub schemas for table look up.

Default implementation returns itself.

Ex. "dfs" schema refers to the tables in "default" workspace when querying for tables in "dfs" schema.

Returns:

Return the default schema where tables are created or retrieved from.

createView

public boolean createView(View view)
                   throws IOException

Create a new view given definition.

Parameters:

view - View info including name, definition etc.

Returns:

Returns true if an existing view is replaced with the given view. False otherwise.

Throws:

IOException - in case of error creating a view

dropView

public void dropView(String viewName)
              throws IOException

Drop the view with given name.

Parameters:

viewName - view name

Throws:

IOException - in case of error dropping the view

createNewTable

public CreateTableEntry createNewTable(String tableName,
                                       List<String> partitionColumns,
                                       StorageStrategy storageStrategy)

Creates table entry using table name, list of partition columns and storage strategy used to create table folder and files

Parameters:

tableName - : new table name.

partitionColumns - : list of partition columns. Empty list if there is no partition columns.

storageStrategy - : storage strategy used to create table folder and files

Returns:

create table entry

createNewTable

public CreateTableEntry createNewTable(String tableName,
                                       List<String> partitionColumns)

Creates table entry using table name and list of partition columns if any. Table folder and files will be created using persistent storage strategy.

Parameters:

tableName - : new table name.

partitionColumns - : list of partition columns. Empty list if there is no partition columns.

Returns:

create table entry

createStatsTable

public CreateTableEntry createStatsTable(String tableName)

Create stats table entry for given tableName.

Parameters:

tableName - table name

Returns:

instance of create table entry

appendToStatsTable

public CreateTableEntry appendToStatsTable(String tableName)

Create an append statistics table entry for given tableName. If there is not existing statistics table, a new one is created.

Parameters:

tableName - table name

Returns:

instance of create table entry

getStatsTable

public org.apache.calcite.schema.Table getStatsTable(String tableName)

Get the statistics table for given tableName

Parameters:

tableName - table name

Returns:

instance of statistics table

showInInformationSchema

public boolean showInInformationSchema()

Reports whether to show items from this schema in INFORMATION_SCHEMA tables. (Controls ... TODO: Doc.: Mention what this typically controls or affects.)

This base implementation returns true.

getFunctions

public Collection<org.apache.calcite.schema.Function> getFunctions(String name)

For the given table names returns list of acceptable table functions which are common for all Drill schemas. When overriding this method, parent functions must be included first to be evaluated first. If not included, parent functions won't be taken into account when creating table instance.

Specified by:

getFunctions in interface org.apache.calcite.schema.Schema

Parameters:

name - table name

Returns:

list of table functions

getFunctionParameters

public List<TableParamDef> getFunctionParameters()

Returns of common table function parameters that can be used all Drill schema implementations.

Returns:

list of table function parameters

applyFunctionParameters

public DrillTable applyFunctionParameters(DrillTable drillTable,
                                          List<TableParamDef> paramDefs,
                                          List<Object> values)

For the given list of parameters definitions executes action for the corresponding value.

Parameters:

drillTable - Drill table instance

paramDefs - parameter definitions

values - parameter values

Returns:

updated Drill table instance

getTypeMap

protected Map<String,org.apache.calcite.rel.type.RelProtoDataType> getTypeMap()

Returns a map of types in this schema by name.

The implementations of getTypeNames() and getType(String) depend on this map. The default implementation of this method returns the empty map. Override this method to change their behavior.

Returns:

Map of types in this schema by name

getTypeNames

public Set<String> getTypeNames()

Specified by:

getTypeNames in interface org.apache.calcite.schema.Schema

getType

public org.apache.calcite.rel.type.RelProtoDataType getType(String name)

Specified by:

getType in interface org.apache.calcite.schema.Schema

getFunctionNames

public Set<String> getFunctionNames()

Specified by:

getFunctionNames in interface org.apache.calcite.schema.Schema

getSubSchema

public org.apache.calcite.schema.Schema getSubSchema(String name)

Specified by:

getSubSchema in interface org.apache.calcite.schema.Schema

getSubSchemaNames

public Set<String> getSubSchemaNames()

Specified by:

getSubSchemaNames in interface org.apache.calcite.schema.Schema

isMutable

public boolean isMutable()

Specified by:

isMutable in interface org.apache.calcite.schema.Schema

getTable

public org.apache.calcite.schema.Table getTable(String name)

Specified by:

getTable in interface org.apache.calcite.schema.Schema

getTableNames

public Set<String> getTableNames()

Specified by:

getTableNames in interface org.apache.calcite.schema.Schema

getExpression

public org.apache.calcite.linq4j.tree.Expression getExpression(org.apache.calcite.schema.SchemaPlus parentSchema,
                                                               String name)

Specified by:

getExpression in interface org.apache.calcite.schema.Schema

close

public void close()
           throws Exception

Specified by:

close in interface AutoCloseable

Throws:

Exception

snapshot

public org.apache.calcite.schema.Schema snapshot(org.apache.calcite.schema.SchemaVersion version)

Specified by:

snapshot in interface org.apache.calcite.schema.Schema

dropTable

public void dropTable(String tableName)

getTablesByNames

public List<org.apache.commons.lang3.tuple.Pair<String,? extends org.apache.calcite.schema.Table>> getTablesByNames(Set<String> tableNames)

Get the collection of Table tables specified in the tableNames.

Parameters:

tableNames - the requested tables, specified by the table names

Returns:

the collection of requested tables

getTableNamesAndTypes

public Collection<Map.Entry<String,org.apache.calcite.schema.Schema.TableType>> getTableNamesAndTypes()

Used by InfoSchemaRecordGenerator.Tables for getting all table objects along with type for every requested schema. It's desired for this method to work fast because it impacts SHOW TABLES query.

Returns:

collection of table names and types

areTableNamesCaseSensitive

public boolean areTableNamesCaseSensitive()

Indicates if table names in schema are case sensitive. By default they are. If schema implementation claims its table names are case insensitive, it is responsible for making case insensitive look up by table name.

Returns:

true if table names are case sensitive

getUser

public String getUser(String impersonated,
                      String notImpersonated)

Parameters:

impersonated - User whom to impersonate. Usually SchemaConfig.getUserName()

notImpersonated - Regular user (table owner for Views or Drillbit process user for Tables)

Returns:

endUser

needToImpersonateReadingData

public boolean needToImpersonateReadingData()

Does Drill needs to impersonate as user connected to Drill when reading data from DataSource?

Returns:

True when both Drill impersonation and DataSource impersonation are enabled.