org.apache.iceberg.expressions

Class ExpressionUtil

java.lang.Object

org.apache.iceberg.expressions.ExpressionUtil

public class ExpressionUtil
extends java.lang.Object

Expression utility methods.

Method Summary

Modifier and Type	Method and Description
static java.lang.String	describe(Term term)
static boolean	equivalent(Expression left, Expression right, Types.StructType struct, boolean caseSensitive) Returns whether two unbound expressions will accept the same inputs.
static Expression	extractByIdInclusive(Expression expression, Schema schema, boolean caseSensitive, int... ids) Extracts an expression that references only the given column IDs from the given expression.
static Expression	sanitize(Expression expr) Produces an unbound Expression with the same structure, but with data values replaced by descriptions.
static Expression	sanitize(Types.StructType struct, Expression expr, boolean caseSensitive) Produces an unbound Expression with the same structure, but with data values replaced by descriptions.
static boolean	selectsPartitions(Expression expr, PartitionSpec spec, boolean caseSensitive) Returns whether an expression selects whole partitions for a partition spec.
static boolean	selectsPartitions(Expression expr, Table table, boolean caseSensitive) Returns whether an expression selects whole partitions for all partition specs in a table.
static java.lang.String	toSanitizedString(Expression expr) Produces a sanitized expression string with the same structure, but with data values replaced by descriptions.
static java.lang.String	toSanitizedString(Types.StructType struct, Expression expr, boolean caseSensitive) Produces a sanitized expression string with the same structure, but with data values replaced by descriptions.
static <T> UnboundTerm<T>	unbind(BoundTerm<T> term)
static <T> UnboundTerm<T>	unbind(Term term)

Methods inherited from class java.lang.Object

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

Method Detail

sanitize

public static Expression sanitize(Expression expr)

Produces an unbound Expression with the same structure, but with data values replaced by descriptions.

Numbers are replaced with magnitude and type, string-like values are replaced by hashes, and date/time values are replaced by the type.

Parameters:

expr - an Expression to sanitize

Returns:

a sanitized Expression

sanitize

public static Expression sanitize(Types.StructType struct,
                                  Expression expr,
                                  boolean caseSensitive)

Produces an unbound Expression with the same structure, but with data values replaced by descriptions.

Numbers are replaced with magnitude and type, string-like values are replaced by hashes, and date/time values are replaced by the type.

Parameters:

struct - a StructType to bind the expression

expr - an Expression to sanitize

caseSensitive - whether to bind case sensitively

Returns:

a sanitized Expression

toSanitizedString

public static java.lang.String toSanitizedString(Expression expr)

Produces a sanitized expression string with the same structure, but with data values replaced by descriptions.

Numbers are replaced with magnitude and type, string-like values are replaced by hashes, and date/time values are replaced by the type.

Parameters:

expr - an Expression to sanitize

Returns:

a sanitized expression string

toSanitizedString

public static java.lang.String toSanitizedString(Types.StructType struct,
                                                 Expression expr,
                                                 boolean caseSensitive)

Produces a sanitized expression string with the same structure, but with data values replaced by descriptions.

Numbers are replaced with magnitude and type, string-like values are replaced by hashes, and date/time values are replaced by the type.

Parameters:

struct - a StructType to bind the expression

expr - an Expression to sanitize

caseSensitive - whether to bind case sensitively

Returns:

a sanitized expression string

extractByIdInclusive

public static Expression extractByIdInclusive(Expression expression,
                                              Schema schema,
                                              boolean caseSensitive,
                                              int... ids)

Extracts an expression that references only the given column IDs from the given expression.

The result is inclusive. If a row would match the original filter, it must match the result filter.

Parameters:

expression - a filter Expression

schema - a Schema

caseSensitive - whether binding is case sensitive

ids - field IDs used to match predicates to extract from the expression

Returns:

an Expression that selects at least the same rows as the original using only the IDs

equivalent

public static boolean equivalent(Expression left,
                                 Expression right,
                                 Types.StructType struct,
                                 boolean caseSensitive)

Returns whether two unbound expressions will accept the same inputs.

If this returns true, the expressions are guaranteed to return the same evaluation for the same input. However, if this returns false the expressions may return the same evaluation for the same input. That is, expressions may be equivalent even if this returns false.

Parameters:

left - an unbound expression

right - an unbound expression

struct - a struct type for binding

caseSensitive - whether to bind expressions using case-sensitive matching

Returns:

true if the expressions are equivalent

selectsPartitions

public static boolean selectsPartitions(Expression expr,
                                        Table table,
                                        boolean caseSensitive)

Returns whether an expression selects whole partitions for all partition specs in a table.

For example, ts < '2021-03-09T10:00:00.000' selects whole partitions in an hourly spec, [hours(ts)], but does not select whole partitions in a daily spec, [days(ts)].

Parameters:

expr - an unbound expression

table - a table

caseSensitive - whether expression binding should be case sensitive

Returns:

true if the expression will select whole partitions in all table specs

selectsPartitions

public static boolean selectsPartitions(Expression expr,
                                        PartitionSpec spec,
                                        boolean caseSensitive)

Returns whether an expression selects whole partitions for a partition spec.

For example, ts < '2021-03-09T10:00:00.000' selects whole partitions in an hourly spec, [hours(ts)], but does not select whole partitions in a daily spec, [days(ts)].

Parameters:

expr - an unbound expression

spec - a partition spec

Returns:

true if the expression will select whole partitions in the given spec

describe

public static java.lang.String describe(Term term)

unbind

public static <T> UnboundTerm<T> unbind(BoundTerm<T> term)

unbind

public static <T> UnboundTerm<T> unbind(Term term)