Using the XML Parser for C++

28.1 Introduction to Oracle XML Parser for C++

Oracle XML parser for C++ determines whether an XML document is well-formed and optionally validates it against a document type definition (DTD) or Extensible Markup Language (XML) schema. The parser constructs an object tree that can be accessed through one of these two XML APIs:

Document Object Model (DOM): Tree-based APIs. A tree-based API compiles an XML document into an internal tree structure, then allows an application to navigate that tree using the DOM, a standard tree-based API for XML and HTML documents.
Simple API for XML (SAX): Event-based APIs. An event-based API reports parsing events (such as the start and end of elements) directly to the application through a user defined SAX even handler, and does not usually build an internal tree. The application implements handlers to deal with the different events, much like handling events in a graphical user interface.

Tree-based APIs are useful for a wide range of applications, but they often put a great strain on system resources, especially if the document is large (under very controlled circumstances, it is possible to construct the tree in a lazy fashion to avoid some of this problem). Furthermore, some applications must build their own, different data trees, and it is very inefficient to build a tree of parse nodes only to map it onto a new tree.

28.2 DOM Namespace

The DOM namespace is the namespace for DOM-related types and interfaces.

DOM interfaces are represented as generic references to different implementations of the DOM specification. They are parameterized by Node that supports various specializations and instantiations. Of them, the most important is xmlnode which corresponds to the current C implementation

These generic references do not have a NULL-like value. Any implementation must never create a reference with no state (like NULL). If it is necessary to signal that something has no state, the implementation must throw an exception.

Many methods might throw the SYNTAX_ERR exception, if the DOM tree is incorrectly formed, or they might throw UNDEFINED_ERR, when encountering incorrect parameters or unexpected NULL pointers. If these are the only errors that a particular method might throw, it is not reflected in the method signature.

Actual DOM trees do not depend on the context, TCtx. However, manipulations on DOM trees in the current, xmlctx-based implementation require access to the current context, TCtx. This is accomplished by passing the context pointer to the constructor of DOMImplRef. In multithreaded environment DOMImplRef is always created in the thread context and, so, has the pointer to the right context.

DOMImplRef provides a way to create DOM trees. DomImplRef is a reference to the actual DOMImplementation object that is created when a regular, noncopy constructor of DomImplRef is invoked. This works well in a multithreaded environment where DOM trees must be shared, and each thread has a separate TCtx associated with it. This works equally well in a single threaded environment.

DOMString is one encoding supported by Oracle implementations. The support of other encodings is an Oracle extension. The oratext* data type is used for all encodings. Interfaces represent DOM level 2 Core interfaces according to Document Object Model Core. These C++ interfaces support the DOM specification as closely as possible. However, Oracle cannot guarantee that the specification is fully supported by our implementation because the World Wide Web Consortium (W3C) specification does not cover C++ binding.

28.2.1 DOM Data Types

DomNodeType defines types of DOM nodes. DomExceptionCode defines exception codes returned by the DOM API.

28.2.2 DOM Interfaces

The DOM interfaces are described.

DOMException Interface—See exception DOMException in the W3C DOM documentation. DOM operations raise exceptions only in "exceptional" circumstances: when an operation is impossible to perform (either for logical reasons, because data is lost, or because the implementation has become unstable). The functionality of XMLException can be used for a wider range of exceptions.

NodeRef Interface—See interface Node in the W3C documentation.

DocumentRef Interface—See interface Document in the W3C documentation.

DocumentFragmentRef Interface—See interface DocumentFragment in the W3C documentation.

ElementRef Interface—See interface Element in the W3C documentation.

AttrRef Interface—See interface Attr in the W3C documentation.

CharacterDataRef Interface—See interface CharacterData in the W3C documentation.

TextRef Interface—See Text nodes in the W3C documentation.

CDATASectionRef Interface—See CDATASection nodes in the W3C documentation.

CommentRef Interface—See Comment nodes in the W3C documentation.

ProcessingInstructionRef Interface—See PI nodes in the W3C documentation.

EntityRef Interface—See Entity nodes in the W3C documentation.

EntityReferenceRef Interface—See EntityReference nodes in the W3C documentation.

NotationRef Interface—See Notation nodes in the W3C documentation.

DocumentTypeRef Interface—See DTD nodes in the W3C documentation.

DOMImplRef Interface—See interface DOMImplementation in the W3C DOM documentation. DOMImplementation is fundamental for manipulating DOM trees. Every DOM tree is attached to a particular DOM implementation object. Several DOM trees can be attached to the same DOM implementation object. Each DOM tree can be deleted and deallocated by deleting the document object. All DOM trees attached to a particular DOM implementation object are deleted when this object is deleted. The DOMImplementation object is not visible to the user directly. It is visible through the class DOMImplRef. This functionality is needed because of requirements for multithreaded environments.

NodeListRef Interface—Abstract implementation of node list. See interface NodeList in the W3C documentation.

NamedNodeMapRef Interface—Abstract implementation of a node map. See interface NamedNodeMap in the W3C documentation.

28.2.3 DOM Traversal and Range Data Types

AcceptNodeCode is the data type for values returned by node filters for iterators and tree walkers. WhatToShowCode is the data type for codes to filter nodes. RangeExceptionCode is the data type for exceptions that can be thrown by interface Range. CompareHowCode is the data type for range comparisons.

28.2.4 DOM Traversal and Range Interfaces

The DOM 2 traversal and range interfaces are NodeFilter, NodeIterator, TreeWalker, DocumentTraversal, RangeException, Range, and DocumentRange.

28.3 Parser Namespace

Interfaces associated with the parser namespace are described.

DOMParser Interface—DOM parser root class.

GParser Interface—Root class for XML parsers.

ParserException Interface—Exception class for parser and validator.

SAXHandler Interface—Root class for current SAX handler implementations.

SAXHandlerRoot Interface—Root class for all SAX handlers.

SAXParser Interface—Root class for all SAX parsers.

SchemaValidator Interface—XML schema-aware validator.

28.3.1 GParser Interface

Interface GParser is the root class for all XML parser interfaces and implementations. It is not an abstract class; that is, it is not an interface. It is a real class that you can use to set and check parser parameters.

28.3.2 DOMParser Interface

Interface DOMParser is the DOM parser root abstract class or interface. In addition to parsing and checking that a document is well formed, it provides ways to validate a document against a document type definition (DTD) or an XML schema.

28.3.3 SAXParser Interface

Interface SAXParser is the root abstract class for all SAX parsers.

28.3.3.1 SAX Event Handlers

To use SAX, a SAX event handler class must be provided by the user and passed to the SAXParser in a parse() invocation or set before such invocation.

SAXHandlerRoot Interface—root class for all SAX handlers.

SAXHandler Interface—root class for current SAX handler implementations.

28.4 Thread Safety for the XML Parser for C++

If threads are forked in the midst of the init–parse–term sequence of invocations, unpredictable behavior or results can occur.

28.5 XML Parser for C++ Usage

Invoke Tools::Factory to create a parser and initialize the parsing process. The XML input can be kind of InputSource (see IO namespace). DOMParser invocation produces the DOM tree. SAXParser invocation produces SAX events. Invoking the parser destructor terminates the process.

28.6 XML Parser for C++ Default Behavior

The default behavior for the XML parser for C++ is described.

Character set encoding is 8-bit encoding of Unicode (UTF-8). If all your documents are ASCII, you are encouraged to set the encoding to US-ASCII for better performance.
Messages are printed to stderr unless msghdlr is specified.
XML parser for C++ determines whether an XML document is well-formed and optionally validates it against a DTD. The parser constructs an object tree that can be accessed through a DOM interface or operates serially through a SAX interface.
A parse tree which can be accessed by DOM APIs is built unless saxcb is set to use the SAX callback APIs. You can set any of the SAX callback functions to NULL if not needed.
The default behavior for the parser is to check that the input is well-formed but not to check whether it is valid. The flag XML_FLAG_VALIDATE can be set to validate the input. The default behavior for white space processing is to be fully conformant with the XML 1.0 spec, that is, all white space is reported back to the application but it is indicated which white space is ignorable. However, some applications may prefer to set the XML_FLAG_DISCARD_WHITESPACE which discards all white space between an end-element tag and this start-element tag.

Note:

Oracle recommends that you set the default encoding explicitly if using only single-byte character sets (such as US-ASCII or any of the ISO-8859 character sets) for performance up to 25% faster than with multibyte character sets, such as UTF-8.
In both of these cases, an event-based API provides a simpler, lower-level access to an XML document: you can parse documents much larger than your available system memory, and you can construct your own data structures using your callback event handlers.

28.7 C++ Sample Files

Directory xdk/demo/cpp/parser/ contains several XML applications that show how to use the XML parser for C++ with the DOM and SAX interfaces.

Change directories to the sample directory ($ORACLE_HOME/xdk/demo/cpp on Solaris, for example) and read the README file. This document explains how to build the sample programs.

Table 28-1 lists the sample files in the directory. Each file *Main.cpp has a corresponding *Gen.cpp and *Gen.hpp.

Table 28-1 XML Parser for C++ Sample Files

Sample File Name	Description
DOMSampleMain.cpp	Sample usage of C++ interfaces of XML parser and DOM.
FullDOMSampleMain.cpp	Manually build DOM and then exercise.
SAXSampleMain.cpp	Source for SAXSample program.