gudusoft.gsqlparser.resolver2.namespace

Class PivotNamespace

Object

gudusoft.gsqlparser.resolver2.namespace.AbstractNamespace

gudusoft.gsqlparser.resolver2.namespace.PivotNamespace

All Implemented Interfaces:

INamespace

public class PivotNamespace
extends AbstractNamespace

Namespace representing a PIVOT table.

A PIVOT transforms rows into columns using an aggregate function:

 SELECT ...
 FROM source_table
 PIVOT (aggregate_function(value_column) FOR pivot_column IN (val1, val2, ...)) AS alias
 

The PIVOT produces:

• Columns from the IN clause (val1, val2, ...) - these are the pivoted columns
• Pass-through columns from the source table (all columns NOT used in aggregate or FOR clause)

For column resolution purposes:

• Pivot columns (from IN clause) resolve to the PivotNamespace
• Other columns may come from the source table (pass-through columns)

Field Summary

Fields

Modifier and Type	Field and Description
static String	PIVOT_TABLE_SUFFIX Display name format for PIVOT table
static String	UNPIVOT_TABLE_SUFFIX Display name format for UNPIVOT table

Fields inherited from class gudusoft.gsqlparser.resolver2.namespace.AbstractNamespace

columnSources, columnSourcesWithRefs, guessColumnStrategy, nameMatcher, node, referenceTraceabilityEnabled, validated

Constructor Summary

Constructors

Constructor and Description
PivotNamespace(TTable pivotTable, TPivotClause pivotClause, TTable sourceTable, String alias, INameMatcher nameMatcher)

Method Summary

Modifier and Type	Method and Description
protected void	doValidate() Subclasses override this to perform actual validation logic.
Map<String,ColumnSource>	getAllColumnSources() Get all column sources for this PIVOT namespace.
List<TTable>	getAllFinalTables() Get all final physical tables this namespace depends on.
String	getDisplayName() Get a display name for this namespace (for debugging/error messages)
TTable	getFinalTable() Get the final physical table this namespace represents.
TPivotClause	getPivotClause() Get the PIVOT clause.
Set<String>	getPivotColumns() Get the set of pivot column names (from IN clause).
TTable	getPivotTable() Get the pivot table.
INamespace	getSourceNamespace() Get the namespace representing the source table.
TTable	getSourceTable() Get the TTable that this namespace is associated with.
TTable	getUnderlyingSourceTable() Get the underlying table that PIVOT operates on.
ColumnLevel	hasColumn(String columnName) Check if a column exists in this namespace.
static String	inListItemAlias(TResultColumn resultColumn) The explicit alias of one PIVOT IN-list item, or null when it has none.
static String	inListValueColumnName(EDbVendor vendor, String rawValue) The name a PIVOT gives to the column generated by one unaliased IN-list value.
boolean	isPivotColumn(String columnName) Check if a column is a pivot column (from IN clause).
boolean	isUnpivot() Check if this is an UNPIVOT namespace.
ColumnSource	resolveColumn(String columnName) Resolve a column name to its source.
void	setSourceNamespace(INamespace sourceNamespace) Set the namespace representing the source table.
static String	stripDelimitersFrom(String name) Strip a leading+trailing quote pair.
String	toString()

Methods inherited from class gudusoft.gsqlparser.resolver2.namespace.AbstractNamespace

addColumnSource, addColumnSource, containsColumnByMatcher, containsColumnNameByMatcher, enableReferenceTraceability, ensureValidated, findColumnKeyByMatcher, getAllUniqueColumns, getColumnReferences, getColumnSourceWithReferences, getGuessColumnStrategy, getNameMatcher, getNode, isReferenceTraceabilityEnabled, isValidated, noteColumnSourceAdded, setGuessColumnStrategy, validate

Methods inherited from class Object

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

Methods inherited from interface gudusoft.gsqlparser.resolver2.namespace.INamespace

addInferredColumn, getInferredColumns, getMetadataState, getSelectStatement, getStarColumn, hasAuthoritativeOutputColumn, hasStarColumn, resolveColumnPath, supportsDynamicInference

Field Detail

PIVOT_TABLE_SUFFIX

public static final String PIVOT_TABLE_SUFFIX

Display name format for PIVOT table

See Also:

Constant Field Values

UNPIVOT_TABLE_SUFFIX

public static final String UNPIVOT_TABLE_SUFFIX

Display name format for UNPIVOT table

See Also:

Constant Field Values

Constructor Detail

PivotNamespace

public PivotNamespace(TTable pivotTable,
                      TPivotClause pivotClause,
                      TTable sourceTable,
                      String alias,
                      INameMatcher nameMatcher)

Method Detail

isUnpivot

public boolean isUnpivot()

Check if this is an UNPIVOT namespace.

setSourceNamespace

public void setSourceNamespace(INamespace sourceNamespace)

Set the namespace representing the source table. This is used for resolving pass-through columns.

getSourceNamespace

public INamespace getSourceNamespace()

Get the namespace representing the source table.

getPivotTable

public TTable getPivotTable()

Get the pivot table.

getUnderlyingSourceTable

public TTable getUnderlyingSourceTable()

Get the underlying table that PIVOT operates on. This is the table whose columns are being pivoted.

getSourceTable

public TTable getSourceTable()

Get the TTable that this namespace is associated with. Used for legacy sync to set TObjectName.sourceTable. For TableNamespace: returns the table itself For SubqueryNamespace: returns the TTable wrapping the subquery For CTENamespace: returns the CTE referencing table For others: returns null For PivotNamespace, returns the PIVOT table (the result of the PIVOT operation). This is the immediate source table for columns resolved through this PIVOT.

Returns:

The source TTable, or null if not applicable

getPivotClause

public TPivotClause getPivotClause()

Get the PIVOT clause.

getPivotColumns

public Set<String> getPivotColumns()

Get the set of pivot column names (from IN clause).

isPivotColumn

public boolean isPivotColumn(String columnName)

Check if a column is a pivot column (from IN clause).

getDisplayName

public String getDisplayName()

Description copied from interface: INamespace

Get a display name for this namespace (for debugging/error messages)

getFinalTable

public TTable getFinalTable()

Description copied from interface: INamespace

Get the final physical table this namespace represents. For tables: returns the table itself For subqueries/CTEs: recursively traces back to physical table(s)

Returns:

Physical table, or null if not determinable

getAllFinalTables

public List<TTable> getAllFinalTables()

Description copied from interface: INamespace

Get all final physical tables this namespace depends on. For simple table namespace: returns single table For subquery/CTE: returns all tables in the query tree

Specified by:

getAllFinalTables in interface INamespace

Overrides:

getAllFinalTables in class AbstractNamespace

Returns:

List of all physical tables

doValidate

protected void doValidate()

Description copied from class: AbstractNamespace

Subclasses override this to perform actual validation logic.

Specified by:

doValidate in class AbstractNamespace

inListItemAlias

public static String inListItemAlias(TResultColumn resultColumn)

The explicit alias of one PIVOT IN-list item, or null when it has none.

IN ('direct' AS Store) names the generated column Store. Oracle, Snowflake, BigQuery and Teradata all accept this form and in every one of them the alias — not the value — is the output column name, so it takes precedence over any value-derived naming.

Shared with ScopeBuilder.addPivotInClauseColumns and matched by TPivotInClause.linkColumnToTable on purpose. The name a pivot column is published under and the name it can be resolved by have to be produced by the same rule, or a reference to the real column reports NOT_FOUND while a stale name still matches.

The alias is returned verbatim, quote state included. Stripping it would be wrong in both directions for IN ('x' AS "Store"): bare SELECT Store would become a false exact match, and the only valid reference, SELECT "Store", would report NOT_FOUND. Retaining quote state is what the identifier normalization guide requires; comparison is the matcher's job, not this method's.

inListValueColumnName

public static String inListValueColumnName(EDbVendor vendor,
                                           String rawValue)

The name a PIVOT gives to the column generated by one unaliased IN-list value.

Vendor-specific, and the difference is not cosmetic. Oracle and Snowflake name the column after the literal as written, quotes included, so IN ('direct') yields a column really named 'direct' that must be referenced as "'direct'" (Mantis 4662). BigQuery requires its IN-list values to form valid identifiers and names the column direct, so the quotes are still stripped there and for every other vendor.

Shared with ScopeBuilder for the same reason as inListItemAlias(gudusoft.gsqlparser.nodes.TResultColumn).

stripDelimitersFrom

public static String stripDelimitersFrom(String name)

Strip a leading+trailing quote pair. SQL Server brackets are part of the identity.

hasColumn

public ColumnLevel hasColumn(String columnName)

Description copied from interface: INamespace

Check if a column exists in this namespace. Returns a three-level answer: EXISTS, MAYBE, NOT_EXISTS

Specified by:

hasColumn in interface INamespace

Overrides:

hasColumn in class AbstractNamespace

Parameters:

columnName - Column name to check

Returns:

Column existence level

resolveColumn

public ColumnSource resolveColumn(String columnName)

Description copied from interface: INamespace

Resolve a column name to its source.

Specified by:

resolveColumn in interface INamespace

Overrides:

resolveColumn in class AbstractNamespace

Parameters:

columnName - Column name to resolve

Returns:

ColumnSource if found, null otherwise

getAllColumnSources

public Map<String,ColumnSource> getAllColumnSources()

Get all column sources for this PIVOT namespace. This includes both pivot columns (from IN clause) and pass-through columns from the source.

For PIVOT, the output columns are:

• Pivot columns: generated from IN clause values (e.g., 'AA', 'KH', 'DL', '9E')
• Pass-through columns: source columns NOT consumed by FOR or aggregate

This method is critical for star column expansion: when outer query uses SELECT * from a PIVOT table, we need to return ALL columns the pivot table exposes.

Specified by:

getAllColumnSources in interface INamespace

Overrides:

getAllColumnSources in class AbstractNamespace

Returns:

Map of column name -> ColumnSource

toString

public String toString()

Overrides:

toString in class AbstractNamespace