org.apache.calcite.sql

Class SqlDataTypeSpec

java.lang.Object

org.apache.calcite.sql.SqlNode

org.apache.calcite.sql.SqlDataTypeSpec

All Implemented Interfaces:

Cloneable

public class SqlDataTypeSpec
extends SqlNode

Represents a SQL data type specification in a parse tree.

A SqlDataTypeSpec is immutable; once created, you cannot change any of the fields.

todo: This should really be a subtype of SqlCall.

In its full glory, we will have to support complex type expressions like:

ROW(
NUMBER(5, 2) NOT NULL AS foo,
ROW(BOOLEAN AS b, MyUDT NOT NULL AS i) AS rec)

Currently it only supports simple datatypes like CHAR, VARCHAR and DOUBLE, with optional precision and scale.

Field Summary

Fields inherited from class org.apache.calcite.sql.SqlNode

EMPTY_ARRAY, pos

Constructor Summary

Constructors

Constructor and Description
SqlDataTypeSpec(SqlIdentifier typeName, int precision, int scale, String charSetName, TimeZone timeZone, SqlParserPos pos) Creates a type specification representing a regular, non-collection type.
SqlDataTypeSpec(SqlIdentifier collectionsTypeName, SqlIdentifier typeName, int precision, int scale, String charSetName, SqlParserPos pos) Creates a type specification representing a collection type.
SqlDataTypeSpec(SqlIdentifier collectionsTypeName, SqlIdentifier typeName, int precision, int scale, String charSetName, TimeZone timeZone, Boolean nullable, SqlParserPos pos) Creates a type specification.

Method Summary

Methods

Modifier and Type	Method and Description
<R> R	accept(SqlVisitor<R> visitor) Accepts a generic visitor.
SqlNode	clone(SqlParserPos pos) Clones a SqlNode with a different position.
RelDataType	deriveType(RelDataTypeFactory typeFactory) Does not throw an error if the type is not built-in.
RelDataType	deriveType(SqlValidator validator) Throws an error if the type is not built-in.
boolean	equalsDeep(SqlNode node, boolean fail) Returns whether this node is structurally equivalent to another node.
String	getCharSetName()
SqlIdentifier	getCollectionsTypeName()
SqlDataTypeSpec	getComponentTypeSpec() Returns a new SqlDataTypeSpec corresponding to the component type if the type spec is a collections type spec. Collection types are ARRAY and MULTISET.
SqlMonotonicity	getMonotonicity(SqlValidatorScope scope) Returns whether expression is always ascending, descending or constant.
int	getPrecision()
int	getScale()
TimeZone	getTimeZone()
SqlIdentifier	getTypeName()
void	unparse(SqlWriter writer, int leftPrec, int rightPrec) Writes a SQL representation of this node to a writer.
void	validate(SqlValidator validator, SqlValidatorScope scope) Validates this node.
SqlDataTypeSpec	withNullable(Boolean nullable) Returns a copy of this data type specification with a given nullability.

Methods inherited from class org.apache.calcite.sql.SqlNode

clone, cloneArray, equalDeep, equalDeep, findValidOptions, getKind, getParserPosition, isA, toSqlString, toSqlString, toString, validateExpr

Methods inherited from class java.lang.Object

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

Constructor Detail

SqlDataTypeSpec

public SqlDataTypeSpec(SqlIdentifier typeName,
               int precision,
               int scale,
               String charSetName,
               TimeZone timeZone,
               SqlParserPos pos)

Creates a type specification representing a regular, non-collection type.

SqlDataTypeSpec

public SqlDataTypeSpec(SqlIdentifier collectionsTypeName,
               SqlIdentifier typeName,
               int precision,
               int scale,
               String charSetName,
               SqlParserPos pos)

Creates a type specification representing a collection type.

SqlDataTypeSpec

public SqlDataTypeSpec(SqlIdentifier collectionsTypeName,
               SqlIdentifier typeName,
               int precision,
               int scale,
               String charSetName,
               TimeZone timeZone,
               Boolean nullable,
               SqlParserPos pos)

Creates a type specification.

Method Detail

clone

public SqlNode clone(SqlParserPos pos)

Description copied from class: SqlNode

Clones a SqlNode with a different position.

Overrides:

clone in class SqlNode

getMonotonicity

public SqlMonotonicity getMonotonicity(SqlValidatorScope scope)

Description copied from class: SqlNode

Returns whether expression is always ascending, descending or constant. This property is useful because it allows to safely aggregte infinite streams of values.

The default implementation returns SqlMonotonicity.NOT_MONOTONIC.

Overrides:

getMonotonicity in class SqlNode

getCollectionsTypeName

public SqlIdentifier getCollectionsTypeName()

getTypeName

public SqlIdentifier getTypeName()

getScale

public int getScale()

getPrecision

public int getPrecision()

getCharSetName

public String getCharSetName()

getTimeZone

public TimeZone getTimeZone()

withNullable

public SqlDataTypeSpec withNullable(Boolean nullable)

Returns a copy of this data type specification with a given nullability.

getComponentTypeSpec

public SqlDataTypeSpec getComponentTypeSpec()

Returns a new SqlDataTypeSpec corresponding to the component type if the type spec is a collections type spec.
Collection types are ARRAY and MULTISET.

unparse

public void unparse(SqlWriter writer,
           int leftPrec,
           int rightPrec)

Description copied from class: SqlNode

Writes a SQL representation of this node to a writer.

The leftPrec and rightPrec parameters give us enough context to decide whether we need to enclose the expression in parentheses. For example, we need parentheses around "2 + 3" if preceded by "5 *". This is because the precedence of the "*" operator is greater than the precedence of the "+" operator.

The algorithm handles left- and right-associative operators by giving them slightly different left- and right-precedence.

If SqlWriter.isAlwaysUseParentheses() is true, we use parentheses even when they are not required by the precedence rules.

For the details of this algorithm, see SqlCall.unparse(org.apache.calcite.sql.SqlWriter, int, int).

Specified by:

unparse in class SqlNode

Parameters:

writer - Target writer

leftPrec - The precedence of the SqlNode immediately preceding this node in a depth-first scan of the parse tree

rightPrec - The precedence of the SqlNode immediately

validate

public void validate(SqlValidator validator,
            SqlValidatorScope scope)

Description copied from class: SqlNode

Validates this node.

The typical implementation of this method will make a callback to the validator appropriate to the node type and context. The validator has methods such as SqlValidator.validateLiteral(org.apache.calcite.sql.SqlLiteral) for these purposes.

Specified by:

validate in class SqlNode

scope - Validator

accept

public <R> R accept(SqlVisitor<R> visitor)

Description copied from class: SqlNode

Accepts a generic visitor.

Implementations of this method in subtypes simply call the appropriate visit method on the visitor object.

The type parameter R must be consistent with the type parameter of the visitor.

Specified by:

accept in class SqlNode

equalsDeep

public boolean equalsDeep(SqlNode node,
                 boolean fail)

Description copied from class: SqlNode

Returns whether this node is structurally equivalent to another node. Some examples:

• 1 + 2 is structurally equivalent to 1 + 2
• 1 + 2 + 3 is structurally equivalent to (1 + 2) + 3, but not to 1 + (2 + 3), because the '+' operator is left-associative

Specified by:

equalsDeep in class SqlNode

deriveType

public RelDataType deriveType(SqlValidator validator)

Throws an error if the type is not built-in.

deriveType

public RelDataType deriveType(RelDataTypeFactory typeFactory)

Does not throw an error if the type is not built-in.