public final class StreamXPath extends java.lang.Object
Represents an xpath for evaluation against an
XMLInputStream. Instances of this class are installed
which in turn can create
performs xpath matching.
An instance of
StreamXPath is a handle to a parsed
representation of a particular xpath expression - parsing occurs
when the StreamXPath is constructed. A single StreamXPath instance
can be installed any number of times in any number of factories so
that the expense of parsing a given xpath expression is paid only
The sequential nature of stream-based XML processing makes it impractical for StreamXPath to fully support the XPath specification. The StreamXPath constructor will throw an XPathUnsupportedException if the given XPath contains any of the unsupported features or constructs described below:
The root expression of a StreamXPath is always a location path which describes the node-set to be observed. It may be absolute or relative; in fact, it makes no difference, since the evaluation always occurs relative to the first node (Element event) encountered on the stream (which is assumed to be the doucment root element).
All of the XPath 1.0 core functions are supported with the exception of the following: last, size, id, lang, and count.
The string() function is not fully supported; it cannot convert a nodeset containing element nodes to a string. This is because the string value of an element is dependent on it's descendant elements, which will not be processed until later in the stream. This limitation also applies to implicit uses of 'string()' as described in the spec - any case where the string-value of a nodeset must be determined.
The following axes are supported in the location path which is the root expression of the xpath: self, child, descendant, descendant-or-self, following, following-sibling, attribute and namespace.
The root location path may contain predicates, and those predicate expressions may in turn contain location paths. Location paths appearing anywhere inside a predicate must be relative (i.e. no leading slash) and may only use the following axes: self, ancestor, and ancestor-or-self, attribute, and namespace.
XPath models an XML document as a tree of nodes. For purposes of streaming evaluation, a node is actually described by two XMLEvents - a START_ELEMENT and an END_ELEMENT. Whenever a START_ELEMENT is encountered that matches the stream, it's corresponding END_ELEMENT will eventually be observed as well.
START_DOCUMENT and END_DOCUMENT are just treated as special cases of START_ELEMENT and END_ELEMENT. It is expected that the first event on the stream will be a START_DOCUMENT event, and the last event will be END_DOCUMENT, although this is not strictly enforced. START_DOCUMENT (along with it's corresponding END_DOCUMENT event) is mapped to the 'Root Node' of the XPath Data model described in section 5.1 of the XPath specification. Thus, an observer on an XPath '.' will always see the START_DOCUMENT and END_DOCUMENT events.
Note: when a both a START-ELEMENT and its attribute and/or namespace declarations element are matches, the element notification is sent out first, followed by the attributes, and finally the namespaces.
|Modifier and Type||Method||Description|
This method allows an expression to be evaluated from the command line using the stream infrastructure.
Sets the variable bindings to be used when evaluating this XPath.
public void setVariableBindings(java.util.Map variableBindings)
variableBindings- Provides a mapping for resolving variable names which may appear in the XPath. Values in the map which are instances of
java.lang.Numberwill be used as values of the corresponding XPath types. Values which are instances of
java.util.Collectionare assumed to be node-lists. Values of other types are not currently recognized. Passing null has the effect of removing all bindings.
public java.lang.String toString()
public boolean equals(java.lang.Object o)
public int hashCode()
public static void main(java.lang.String args)