org.apache.drill.common.expression

Class PathSegment

java.lang.Object

org.apache.drill.common.expression.PathSegment

Direct Known Subclasses:

PathSegment.ArraySegment, PathSegment.NameSegment

public abstract class PathSegment
extends Object

Used to represent path to nested field within schema as a chain of path segments. Where each segment in the chain may be either named or array segment.

See Also:

PathSegment.NameSegment, PathSegment.ArraySegment, SchemaPath

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	PathSegment.ArraySegment
static class	PathSegment.NameSegment

Field Summary

Fields

Modifier and Type	Field and Description
protected Object	originalValue Holds original value associated with the path segment.
protected TypeProtos.MajorType	originalValueType Indicates the type of original value.

Constructor Summary

Constructors

Constructor and Description
PathSegment(PathSegment child, Object originalValue, TypeProtos.MajorType originalValueType)

Method Summary

Modifier and Type	Method and Description
abstract PathSegment	clone()
abstract PathSegment	cloneWithNewChild(PathSegment newChild) Makes copy of segment chain with newChild added at the end.
boolean	contains(PathSegment otherSeg) Check if another path is contained in this one.
boolean	equals(Object obj)
PathSegment.ArraySegment	getArraySegment()
PathSegment	getChild()
PathSegment.NameSegment	getNameSegment()
Object	getOriginalValue()
TypeProtos.MajorType	getOriginalValueType()
int	hashCode()
abstract boolean	isArray()
boolean	isLastPath()
abstract boolean	isNamed()
protected abstract boolean	segmentEquals(PathSegment other)
protected abstract int	segmentHashCode()

Methods inherited from class java.lang.Object

finalize, getClass, notify, notifyAll, toString, wait, wait, wait

Field Detail

originalValue

protected final Object originalValue

Holds original value associated with the path segment. Used when reading data from DICT.

originalValueType

protected final TypeProtos.MajorType originalValueType

Indicates the type of original value.

See Also:

originalValue

Constructor Detail

PathSegment

public PathSegment(PathSegment child,
                   Object originalValue,
                   TypeProtos.MajorType originalValueType)

Method Detail

cloneWithNewChild

public abstract PathSegment cloneWithNewChild(PathSegment newChild)

Makes copy of segment chain with newChild added at the end.

Parameters:

newChild - new child to add

Returns:

full copy of segment chain with new child added at the end of chain

clone

public abstract PathSegment clone()

Overrides:

clone in class Object

getOriginalValue

public Object getOriginalValue()

getOriginalValueType

public TypeProtos.MajorType getOriginalValueType()

getNameSegment

public PathSegment.NameSegment getNameSegment()

getArraySegment

public PathSegment.ArraySegment getArraySegment()

isArray

public abstract boolean isArray()

isNamed

public abstract boolean isNamed()

isLastPath

public boolean isLastPath()

getChild

public PathSegment getChild()

segmentHashCode

protected abstract int segmentHashCode()

segmentEquals

protected abstract boolean segmentEquals(PathSegment other)

hashCode

public int hashCode()

Overrides:

hashCode in class Object

equals

public boolean equals(Object obj)

Overrides:

equals in class Object

contains

public boolean contains(PathSegment otherSeg)

Check if another path is contained in this one. This is useful for 2 cases. The first is checking if the other is lower down in the tree, below this path. The other is if a path is actually contained above the current one. Examples: [a] . contains( [a.b.c] ) returns true [a.b.c] . contains( [a] ) returns true This behavior is used for cases like scanning json in an event based fashion, when we arrive at a node in a complex type, we will know the complete path back to the root. This method can be used to determine if we need the data below. This is true in both the cases where the column requested from the user is below the current node (in which case we may ignore other nodes further down the tree, while keeping others). This is also the case if the requested path is further up the tree, if we know we are at position a.b.c and a.b was a requested column, we need to scan all of the data at and below the current a.b.c node.

Parameters:

otherSeg - - path segment to check if it is contained below this one.

Returns:

- is this a match