org.apache.flink.table.planner.delegation.hive.copy

Class HiveParserRowResolver

java.lang.Object

org.apache.flink.table.planner.delegation.hive.copy.HiveParserRowResolver

All Implemented Interfaces:

Serializable

public class HiveParserRowResolver
extends Object
implements Serializable

Counterpart of hive's org.apache.hadoop.hive.ql.parse.RowResolver.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
HiveParserRowResolver()
HiveParserRowResolver(org.apache.hadoop.hive.ql.exec.RowSchema rowSchema, LinkedHashMap<String,LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo>> rslvMap, HashMap<String,String[]> invRslvMap, Map<String,String[]> altInvRslvMap, Map<String,HiveParserASTNode> expressionMap, boolean isExprResolver, LinkedHashMap<String,LinkedHashMap<String,String>> ambiguousColumns, boolean checkForAmbiguity)

Method Summary

Modifier and Type	Method and Description
static boolean	add(HiveParserRowResolver rrToAddTo, HiveParserRowResolver rrToAddFrom)
boolean	addMappingOnly(String tabAlias, String colAlias, org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo)
HiveParserRowResolver	duplicate()
org.apache.hadoop.hive.ql.exec.ColumnInfo	get(String tabAlias, String colAlias) Gets the column Info to tab_alias.col_alias type of a column reference.
String[]	getAlternateMappings(String internalName)
boolean	getCheckForAmbiguity()
ArrayList<org.apache.hadoop.hive.ql.exec.ColumnInfo>	getColumnInfos()
static HiveParserRowResolver	getCombinedRR(HiveParserRowResolver leftRR, HiveParserRowResolver rightRR) Return a new row resolver that is combination of left RR and right RR.
org.apache.hadoop.hive.ql.exec.ColumnInfo	getExpression(HiveParserASTNode node) Retrieves the ColumnInfo corresponding to a source expression which exactly matches the string rendering of the given HiveParserASTNode.
HiveParserASTNode	getExpressionSource(HiveParserASTNode node) Retrieves the source expression matching a given HiveParserASTNode's string rendering exactly.
LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo>	getFieldMap(String tabAlias)
boolean	getIsExprResolver()
HiveParserNamedJoinInfo	getNamedJoinInfo()
int	getPosition(String internalName)
List<String>	getReferenceableColumnAliases(String tableAlias, int max)
org.apache.hadoop.hive.ql.exec.RowSchema	getRowSchema()
LinkedHashMap<String,LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo>>	getRslvMap()
boolean	hasTableAlias(String tabAlias)
void	put(String tabAlias, String colAlias, org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo)
void	putExpression(HiveParserASTNode node, org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo) Puts a resolver entry corresponding to a source expression which is to be used for identical expression recognition (e.g. for matching expressions in the SELECT list with the GROUP BY clause).
boolean	putWithCheck(String tabAlias, String colAlias, String internalName, org.apache.hadoop.hive.ql.exec.ColumnInfo newCI) Adds column to RR, checking for duplicate columns.
String[]	reverseLookup(String internalName)
void	setCheckForAmbiguity(boolean check)
void	setIsExprResolver(boolean isExprResolver)
void	setNamedJoinInfo(HiveParserNamedJoinInfo namedJoinInfo)
String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

HiveParserRowResolver

public HiveParserRowResolver()

HiveParserRowResolver

public HiveParserRowResolver(org.apache.hadoop.hive.ql.exec.RowSchema rowSchema,
                             LinkedHashMap<String,LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo>> rslvMap,
                             HashMap<String,String[]> invRslvMap,
                             Map<String,String[]> altInvRslvMap,
                             Map<String,HiveParserASTNode> expressionMap,
                             boolean isExprResolver,
                             LinkedHashMap<String,LinkedHashMap<String,String>> ambiguousColumns,
                             boolean checkForAmbiguity)

Method Detail

putExpression

public void putExpression(HiveParserASTNode node,
                          org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo)

Puts a resolver entry corresponding to a source expression which is to be used for identical expression recognition (e.g. for matching expressions in the SELECT list with the GROUP BY clause). The convention for such entries is an empty-string ("") as the table alias together with the string rendering of the HiveParserASTNode as the column alias.

getExpression

public org.apache.hadoop.hive.ql.exec.ColumnInfo getExpression(HiveParserASTNode node)
                                                        throws org.apache.hadoop.hive.ql.parse.SemanticException

Retrieves the ColumnInfo corresponding to a source expression which exactly matches the string rendering of the given HiveParserASTNode.

Throws:

org.apache.hadoop.hive.ql.parse.SemanticException

getExpressionSource

public HiveParserASTNode getExpressionSource(HiveParserASTNode node)

Retrieves the source expression matching a given HiveParserASTNode's string rendering exactly.

put

public void put(String tabAlias,
                String colAlias,
                org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo)

addMappingOnly

public boolean addMappingOnly(String tabAlias,
                              String colAlias,
                              org.apache.hadoop.hive.ql.exec.ColumnInfo colInfo)

hasTableAlias

public boolean hasTableAlias(String tabAlias)

get

public org.apache.hadoop.hive.ql.exec.ColumnInfo get(String tabAlias,
                                                     String colAlias)
                                              throws org.apache.hadoop.hive.ql.parse.SemanticException

Gets the column Info to tab_alias.col_alias type of a column reference. I the tab_alias is not provided as can be the case with an non aliased column, this function looks up the column in all the table aliases in this row resolver and returns the match. It also throws an exception if the column is found in multiple table aliases. If no match is found a null values is returned. This allows us to interpret both select t.c1 type of references and select c1 kind of references. The later kind are what we call non aliased column references in the query.

Parameters:

tabAlias - The table alias to match (this is null if the column reference is non aliased)

colAlias - The column name that is being searched for

Returns:

ColumnInfo

Throws:

org.apache.hadoop.hive.ql.parse.SemanticException

getColumnInfos

public ArrayList<org.apache.hadoop.hive.ql.exec.ColumnInfo> getColumnInfos()

getReferenceableColumnAliases

public List<String> getReferenceableColumnAliases(String tableAlias,
                                                  int max)

getFieldMap

public LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo> getFieldMap(String tabAlias)

getPosition

public int getPosition(String internalName)

reverseLookup

public String[] reverseLookup(String internalName)

setIsExprResolver

public void setIsExprResolver(boolean isExprResolver)

getIsExprResolver

public boolean getIsExprResolver()

getAlternateMappings

public String[] getAlternateMappings(String internalName)

toString

public String toString()

Overrides:

toString in class Object

getRowSchema

public org.apache.hadoop.hive.ql.exec.RowSchema getRowSchema()

getRslvMap

public LinkedHashMap<String,LinkedHashMap<String,org.apache.hadoop.hive.ql.exec.ColumnInfo>> getRslvMap()

putWithCheck

public boolean putWithCheck(String tabAlias,
                            String colAlias,
                            String internalName,
                            org.apache.hadoop.hive.ql.exec.ColumnInfo newCI)
                     throws org.apache.hadoop.hive.ql.parse.SemanticException

Adds column to RR, checking for duplicate columns. Needed because CBO cannot handle the Hive behavior of blindly overwriting old mapping in RR and still somehow working after that.

Returns:

True if mapping was added without duplicates.

Throws:

org.apache.hadoop.hive.ql.parse.SemanticException

add

public static boolean add(HiveParserRowResolver rrToAddTo,
                          HiveParserRowResolver rrToAddFrom)
                   throws org.apache.hadoop.hive.ql.parse.SemanticException

Throws:

org.apache.hadoop.hive.ql.parse.SemanticException

getCombinedRR

public static HiveParserRowResolver getCombinedRR(HiveParserRowResolver leftRR,
                                                  HiveParserRowResolver rightRR)
                                           throws org.apache.hadoop.hive.ql.parse.SemanticException

Return a new row resolver that is combination of left RR and right RR. The schema will be schema of left, schema of right.

Throws:

org.apache.hadoop.hive.ql.parse.SemanticException

duplicate

public HiveParserRowResolver duplicate()

getNamedJoinInfo

public HiveParserNamedJoinInfo getNamedJoinInfo()

setNamedJoinInfo

public void setNamedJoinInfo(HiveParserNamedJoinInfo namedJoinInfo)

setCheckForAmbiguity

public void setCheckForAmbiguity(boolean check)

getCheckForAmbiguity

public boolean getCheckForAmbiguity()