
org.simpleframework.xml.stream.InputAttribute Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of simple-xml Show documentation
Show all versions of simple-xml Show documentation
Simple is a high performance XML serialization and configuration framework for Java
/*
* InputAttribute.java July 2006
*
* Copyright (C) 2006, Niall Gallagher
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
* implied. See the License for the specific language governing
* permissions and limitations under the License.
*/
package org.simpleframework.xml.stream;
/**
* The InputAttribute
is used to represent an attribute
* within an element. Rather than representing an attribute as a
* name value pair of strings, an attribute is instead represented
* as an input node, in the same manner as an element. The reason
* for representing an attribute in this way is such that a uniform
* means of extracting and parsing values can be used for inputs.
*
* @author Niall Gallagher
*/
class InputAttribute implements InputNode {
/**
* This is the parent node to this attribute instance.
*/
private InputNode parent;
/**
* This is the reference associated with this attribute node.
*/
private String reference;
/**
* This is the prefix associated with this attribute node.
*/
private String prefix;
/**
* Represents the name of this input attribute instance.
*/
private String name;
/**
* Represents the value for this input attribute instance.
*/
private String value;
/**
* This is the source associated with this input attribute.
*/
private Object source;
/**
* Constructor for the InputAttribute
object. This
* is used to create an input attribute using the provided name
* and value, all other values for this input node will be null.
*
* @param parent this is the parent node to this attribute
* @param name this is the name for this attribute object
* @param value this is the value for this attribute object
*/
public InputAttribute(InputNode parent, String name, String value) {
this.parent = parent;
this.value = value;
this.name = name;
}
/**
* Constructor for the InputAttribute
object. This
* is used to create an input attribute using the provided name
* and value, all other values for this input node will be null.
*
* @param parent this is the parent node to this attribute
* @param attribute this is the attribute containing the details
*/
public InputAttribute(InputNode parent, Attribute attribute) {
this.reference = attribute.getReference();
this.prefix = attribute.getPrefix();
this.source = attribute.getSource();
this.value = attribute.getValue();
this.name = attribute.getName();
this.parent = parent;
}
/**
* This is used to return the source object for this node. This
* is used primarily as a means to determine which XML provider
* is parsing the source document and producing the nodes. It
* is useful to be able to determine the XML provider like this.
*
* @return this returns the source of this input node
*/
public Object getSource() {
return source;
}
/**
* This is used to acquire the Node
that is the
* parent of this node. This will return the node that is
* the direct parent of this node and allows for siblings to
* make use of nodes with their parents if required.
*
* @return this returns the parent node for this node
*/
public InputNode getParent() {
return parent;
}
/**
* This provides the position of this node within the document.
* This allows the user of this node to report problems with
* the location within the document, allowing the XML to be
* debugged if it does not match the class schema.
*
* @return this returns the position of the XML read cursor
*/
public Position getPosition() {
return parent.getPosition();
}
/**
* Returns the name of the node that this represents. This is
* an immutable property and will not change for this node.
*
* @return returns the name of the node that this represents
*/
public String getName() {
return name;
}
/**
* This is used to acquire the namespace prefix for the node.
* If there is no namespace prefix for the node then this will
* return null. Acquiring the prefix enables the qualification
* of the node to be determined. It also allows nodes to be
* grouped by its prefix and allows group operations.
*
* @return this returns the prefix associated with this node
*/
public String getPrefix() {
return prefix;
}
/**
* This allows the namespace reference URI to be determined.
* A reference is a globally unique string that allows the
* node to be identified. Typically the reference will be a URI
* but it can be any unique string used to identify the node.
* This allows the node to be identified within the namespace.
*
* @return this returns the associated namespace reference URI
*/
public String getReference() {
return reference;
}
/**
* Returns the value for the node that this represents. This
* is an immutable value for the node and cannot be changed.
*
* @return the name of the value for this node instance
*/
public String getValue() {
return value;
}
/**
* This method is used to determine if this node is the root
* node for the XML document. This will return false as this
* node can never be the root node because it is an attribute.
*
* @return this will always return false for attribute nodes
*/
public boolean isRoot() {
return false;
}
/**
* This is used to determine if this node is an element. This
* node instance can not be an element so this method returns
* false. Returning null tells the users of this node that any
* attributes added to the node map will be permenantly lost.
*
* @return this returns false as this is an attribute node
*/
public boolean isElement() {
return false;
}
/**
* Because the InputAttribute
object represents an
* attribute this method will return null. If nodes are added
* to the node map the values will not be available here.
*
* @return this always returns null for a requested attribute
*/
public InputNode getAttribute(String name) {
return null;
}
/**
* Because the InputAttribute
object represents an
* attribute this method will return an empty map. If nodes are
* added to the node map the values will not be maintained.
*
* @return this always returns an empty node map of attributes
*/
public NodeMap getAttributes() {
return new InputNodeMap(this);
}
/**
* Because the InputAttribute
object represents an
* attribute this method will return null. An attribute is a
* simple name value pair an so can not contain any child nodes.
*
* @return this always returns null for a requested child node
*/
public InputNode getNext() {
return null;
}
/**
* Because the InputAttribute
object represents an
* attribute this method will return null. An attribute is a
* simple name value pair an so can not contain any child nodes.
*
* @param name this is the name of the next expected element
*
* @return this always returns null for a requested child node
*/
public InputNode getNext(String name) {
return null;
}
/**
* This method is used to skip all child elements from this
* element. This allows elements to be effectively skipped such
* that when parsing a document if an element is not required
* then that element can be completely removed from the XML.
*/
public void skip() {
return;
}
/**
* This is used to determine if this input node is empty. An
* empty node is one with no attributes or children. This can
* be used to determine if a given node represents an empty
* entity, with which no extra data can be extracted.
*
* @return this will always return false as it has a value
*/
public boolean isEmpty() {
return false;
}
/**
* This is the string representation of the attribute. It is
* used for debugging purposes. When evaluating the attribute
* the to string can be used to print out the attribute name.
*
* @return this returns a text description of the attribute
*/
public String toString() {
return String.format("attribute %s='%s'", name, value);
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy