Preface

About This Document

The goal of this document is to provide you with an understanding of the C++/Parser programming model and allow you to efficiently evaluate XSD/e against your project's technical requirements. As such, this document is intended for embedded C++ developers and software architects who are looking for an embedded XML processing solution. Prior experience with XML and C++ is required to understand this document. Basic understanding of XML Schema is advantageous but not expected or required.

More Information

Beyond this guide, you may also find the following sources of information useful:

• XSD/e Compiler Command Line Manual
• The INSTALL file in the XSD/e distribution provides build instructions for various platforms.
• The examples/cxx/parser/ directory in the XSD/e distribution contains a collection of examples and a README file with an overview of each example.
• The xsde-users mailing list is the place to ask technical questions about XSD/e and the Embedded C++/Parser mapping. Furthermore, the archives may already have answers to some of your questions.

1 Introduction

Welcome to CodeSynthesis XSD/e and the Embedded C++/Parser mapping. XSD/e is a dependency-free W3C XML Schema to C++ compiler for mobile, embedded, and light-weight applications. Embedded C++/Parser is an XML Schema to C++ mapping that represents an XML vocabulary as a set of parser skeletons which you can implement to perform XML processing as required by your application logic.

1.1 Mapping Overview

The Embedded C++/Parser mapping provides event-driven, stream-oriented XML parsing, XML Schema validation, and C++ data binding. It was specifically designed and optimized for mobile and embedded systems where hardware constraints require high efficiency and economical use of resources. As a result, the generated parsers are 2-10 times faster than general-purpose validating XML parsers while at the same time maintaining extremely low static and dynamic memory footprints. For example, a validating parser executable can be as small as 120KB in size. The size can be further reduced by disabling support for XML Schema validation.

The generated code and the runtime library are also highly-portable and, in their minimal configuration, can be used without STL, RTTI, iostream, C++ exceptions, and C++ templates.

To speed up application development, the C++/Parser mapping can be instructed to generate sample parser implementations and a test driver which can then be filled with the application logic code. The mapping also provides a wide range of mechanisms for controlling and customizing the generated code.

The next chapter shows how to create a simple application that uses the Embedded C++/Parser mapping to parse, validate, and extract data from a simple XML instance document. The following chapters describe the Embedded C++/Parser mapping in more detail.

1.2 Benefits

Traditional XML access APIs such as Document Object Model (DOM) or Simple API for XML (SAX) as well as general-purpose XML Schema validators have a number of drawbacks that make them less suitable for creating mobile and embedded XML processing applications. These drawbacks include:

• Text-based representation results in inefficient use of resources.
• Extra validation code that is not used by the application.
• Generic representation of XML in terms of elements, attributes, and text forces an application developer to write a substantial amount of bridging code that identifies and transforms pieces of information encoded in XML to a representation more suitable for consumption by the application logic.
• String-based flow control defers error detection to runtime. It also reduces code readability and maintainability.
• Lack of type safety because all information is represented as text.
• Resulting applications are hard to debug, change, and maintain.

In contrast, statically-typed, vocabulary-specific parser skeletons produced by the Embedded C++/Parser mapping use native data representations (for example, integers are passed as integers, not as text) and include validation code only for XML Schema constructs that are used in the application. This results in efficient use of resources and compact object code.

Furthermore, the parser skeletons allow you to operate in your domain terms instead of the generic elements, attributes, and text. Static typing helps catch errors at compile-time rather than at run-time. Automatic code generation frees you for more interesting tasks (such as doing something useful with the information stored in the XML documents) and minimizes the effort needed to adapt your applications to changes in the document structure. To summarize, the C++/Parser mapping has the following key advantages over generic XML access APIs:

• Ease of use. The generated code hides all the complexity associated with recreating the document structure, maintaining the dispatch state, and converting the data from the text representation to data types suitable for manipulation by the application logic. Parser skeletons also provide a convenient mechanism for building custom in-memory representations.
• Natural representation. The generated parser skeletons implement parser callbacks as virtual functions with names corresponding to elements and attributes in XML. As a result, you process the XML data using your domain vocabulary instead of generic elements, attributes, and text.
• Concise code. With a separate parser skeleton for each XML Schema type, the application implementation is simpler and thus easier to read and understand.
• Safety. The XML data is delivered to parser callbacks as statically typed objects. The parser callbacks themselves are virtual functions. This helps catch programming errors at compile-time rather than at runtime.
• Maintainability. Automatic code generation minimizes the effort needed to adapt the application to changes in the document structure. With static typing, the C++ compiler can pin-point the places in the application code that need to be changed.
• Efficiency. The generated parser skeletons use native data representations and combine data extraction, validation, and even dispatching in a single step. This makes them much more efficient than traditional architectures with separate stages for validation and data extraction/dispatch.

2 Hello World Example

In this chapter we will examine how to parse a very simple XML document using the XSD/e-generated C++/Parser skeletons. All the code presented in this chapter is based on the hello example which can be found in the examples/cxx/parser/ directory of the XSD/e distribution.

2.1 Writing XML Document and Schema

First, we need to get an idea about the structure of the XML documents we are going to process. Our hello.xml, for example, could look like this:

<?xml version="1.0"?>
<hello>

  <greeting>Hello</greeting>

  <name>sun</name>
  <name>moon</name>
  <name>world</name>

</hello>
      

Then we can write a description of the above XML in the XML Schema language and save it into hello.xsd:

<?xml version="1.0"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">

  <xs:complexType name="hello">
    <xs:sequence>
      <xs:element name="greeting" type="xs:string"/>
      <xs:element name="name" type="xs:string" maxOccurs="unbounded"/>
    </xs:sequence>
  </xs:complexType>

  <xs:element name="hello" type="hello"/>

</xs:schema>
      

Even if you are not familiar with the XML Schema language, it should be easy to connect declarations in hello.xsd to elements in hello.xml. The hello type is defined as a sequence of the nested greeting and name elements. Note that the term sequence in XML Schema means that elements should appear in a particular order as opposed to appearing multiple times. The name element has its maxOccurs property set to unbounde which means it can appear multiple times in an XML document. Finally, the globally-defined hello element prescribes the root element for our vocabulary. For an easily-approachable introduction to XML Schema refer to XML Schema Part 0: Primer.

The above schema is a specification of our vocabulary; it tells everybody what valid XML instances of our vocabulary should look like. The next step is to compile this schema to generate C++ parser skeletons.

2.2 Translating Schema to C++

Now we are ready to translate our hello.xsd to C++ parser skeletons. To do this we invoke the XSD/e compiler from a terminal (UNIX) or a command prompt (Windows):

$ xsde cxx-parser hello.xsd
  

The XSD/e compiler produces two C++ files: hello-pskel.hxx and hello-pskel.cxx. The following code fragment is taken from hello-pskel.hxx; it should give you an idea about what gets generated:

class hello_pskel
{
public:
  // Parser callbacks. Override them in your implementation.
  //
  virtual void
  pre ();

  virtual void
  greeting (const std::string&);

  virtual void
  name (const std::string&);

  virtual void
  post_hello ();

  // Parser construction API.
  //
  void
  greeting_parser (xml_schema::string_pskel&);

  void
  name_parser (xml_schema::string_pskel&);

  void
  parsers (xml_schema::string_pskel& /* greeting */,
           xml_schema::string_pskel& /* name */);

private:
  ...
};
  

The first four member functions shown above are called parser callbacks. You would normally override them in your implementation of the parser to do something useful. Let's go through all of them one by one.

The pre() function is an initialization callback. It is called when a new element of type hello is about to be parsed. You would normally use this function to allocate a new instance of the resulting type or clear accumulators that are used to gather information during parsing. The default implementation of this function does nothing.

The post_hello() function is a finalization callback. Its name is constructed by adding the parser skeleton name to the post_ prefix. The finalization callback is called when parsing of the element is complete and the result, if any, should be returned. Note that in our case the return type of post_hello() is void which means there is nothing to return. More on parser return types later.

You may be wondering why the finalization callback is called post_hello() instead of post() just like pre(). The reason for this is that finalization callbacks can have different return types and result in function signature clashes across inheritance hierarchies. To prevent this, the signatures of finalization callbacks are made unique by adding the type name to their names.

The greeting() and name() functions are called when the greeting and name elements have been parsed, respectively. Their arguments are of type std::string and contain the data extracted from XML.

The last three functions are for connecting parsers to each other. For example, there is a predefined parser for built-in XML Schema type string in the XSD/e runtime. We will be using it to parse the contents of greeting and name elements, as shown in the next section.

2.3 Implementing Application Logic

At this point we have all the parts we need to do something useful with the information stored in XML documents. The first step is to implement the parser:

#include <iostream>
#include "hello-pskel.hxx"

class hello_pimpl: public hello_pskel
{
public:
  virtual void
  greeting (const std::string& g)
  {
    greeting_ = g;
  }

  virtual void
  name (const std::string& n)
  {
    std::cout << greeting_ << ", " << n << "!" << std::endl;
  }

private:
  std::string greeting_;
};
  

We left both pre() and post_hello() with the default implementations; we don't have anything to initialize or return. The rest is pretty straightforward: we store the greeting in a member variable and later, when parsing names, use it to say hello.

An observant reader my ask what happens if the name element comes before greeting? Don't we need to make sure greeting_ was initialized and report an error otherwise? The answer is no, we don't have to do any of this. The hello_pskel parser skeleton performs validation of XML according to the schema from which it was generated. As a result, it will check the order of the greeting and name elements and report an error if it is violated.

Now it is time to put this parser implementation to work:

using namespace std;

int
main (int argc, char* argv[])
{
  try
  {
    // Construct the parser.
    //
    xml_schema::string_pimpl string_p;
    hello_pimpl hello_p;

    hello_p.greeting_parser (string_p);
    hello_p.name_parser (string_p);

    // Parse the XML instance.
    //
    xml_schema::document_pimpl doc_p (hello_p, "hello");

    hello_p.pre ();
    doc_p.parse (argv[1]);
    hello_p.post_hello ();
  }
  catch (const xml_schema::parser_exception& e)
  {
    cerr << argv[1] << ":" << e.line () << ":" << e.column ()
         << ": " << e.text () << endl;
    return 1;
  }
}
  

The first part of this code snippet instantiates individual parsers and assembles them into a complete vocabulary parser. xml_schema::string_pimpl is an implementation of a parser for built-in XML Schema type string. It is provided by the XSD/e runtime along with parsers for other built-in types (for more information on the built-in parsers see Chapter 6, "Built-In XML Schema Type Parsers"). We use string_pimpl to parse the greeting and name elements as indicated by the calls to greeting_parser() and name_parser().

Then we instantiate a document parser (doc_p). The first argument to its constructor is the parser for the root element (hello_p in our case). The second argument is the root element name.

The final piece is the calls to pre(), parse(), and post_hello(). The call to parse() perform the actual XML parsing while the calls to pre() and post_hello() make sure that the parser for the root element can perform proper initialization and cleanup.

While our parser implementation and test driver are pretty small and easy to write by hand, for bigger XML vocabularies it can be a substantial effort. To help with this task XSD/e can automatically generate sample parser implementations and a test driver from your schemas. You can request the generation of a sample implementation with empty function bodies by specifying the --generate-noop-impl option. Or you can generate a sample implementation that prints the data store in XML by using the --generate-print-impl option. To request the generation of a test driver you can use the --generate-test-driver option. For more information on these options refer to the XSD/e Compiler Command Line Manual. The 'generated' example in the XSD/e distribution shows the sample implementation generation feature in action.

2.4 Compiling and Running

After saving all the parts from the previous section in driver.cxx, we are ready to compile our first application and run it on the test XML document. On UNIX this can be done with the following commands:

$ c++ -I.../libxsde -c driver.cxx hello-pskel.cxx
$ c++ -o driver driver.o hello-pskel.o .../libxsde/xsde/libxsde.a
$ ./driver hello.xml
Hello, sun!
Hello, moon!
Hello, world!
  

Here .../libxsde represents the path to the libxsde directory in the XSD/e distribution. We can also test the error handling. To test XML well-formedness checking, we can try to parse hello-pskel.hxx:

$ ./driver hello-pskel.hxx
hello-pskel.hxx:1:0: not well-formed (invalid token)
  

We can also try to parse a valid XML but not from our vocabulary, for example hello.xsd:

$ ./driver hello.xsd
hello.xsd:2:57: unexpected element encountered
  

3 Parser Skeletons

As we have seen in the previous chapter, the XSD/e compiler generates a parser skeleton class for each type defined in XML Schema. In this chapter we will take a closer look at different functions that comprise a parser skeleton as well as the way to connect our implementations of these parser skeletons to create a complete parser.

In this and subsequent chapters we will use the following schema that describes a collection of person records. We save it in people.xsd:

<?xml version="1.0"?>
<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">

  <xs:simpleType name="gender">
    <xs:restriction base="xs:string">
      <xs:enumeration value="male"/>
      <xs:enumeration value="female"/>
    </xs:restriction>
  </xs:simpleType>

  <xs:complexType name="person">
    <xs:sequence>
      <xs:element name="first-name" type="xs:string"/>
      <xs:element name="last-name" type="xs:string"/>
      <xs:element name="gender" type="gender"/>
      <xs:element name="age" type="xs:short"/>
    </xs:sequence>
  </xs:complexType>

  <xs:complexType name="people">
    <xs:sequence>
      <xs:element name="person" type="person" maxOccurs="unbounded"/>
    </xs:sequence>
  </xs:complexType>

  <xs:element name="people" type="people"/>

</xs:schema>
  

A sample XML instance to go along with this schema is saved in people.xml:

<?xml version="1.0"?>
<people>
  <person>
    <first-name>John</first-name>
    <last-name>Doe</last-name>
    <gender>male</gender>
    <age>32</age>
  </person>
  <person>
    <first-name>Jane</first-name>
    <last-name>Doe</last-name>
    <gender>female</gender>
    <age>28</age>
  </person>
</people>
  

Compiling people.xsd with the XSD/e compiler results in three parser skeletons being generated: gender_pskel, person_pskel, and people_pskel. We are going to examine and implement each of them in the subsequent sections.

3.1 Implementing the Gender Parser

The generated gender_pskel parser skeleton looks like this:

class gender_pskel: public xml_schema::string_pskel
{
public:
  gender_pskel (xml_schema::string_pskel* base_impl);

  // Parser callbacks. Override them in your implementation.
  //
  virtual void
  pre ();

  virtual void
  post_gender ();
};
  

Notice that gender_pskel inherits from xml_schema::string_pskel which is a parser skeleton for built-in XML Schema type string and is predefined in the XSD/e runtime library. This is an example of the general rule that parser skeletons follow: if a type in XML Schema inherits from another then there will be an equivalent inheritance between the corresponding parser skeleton classes. The gender_pskel class also declares a constructor which expects a pointer to the base parser skeleton. We will discuss the purpose of this constructor shortly.

The pre() and post_gender() callbacks should look familiar from the previous chapter. Let's now implement the parser. Our implementation will simply print the gender to cout:

class gender_pimpl: public gender_pskel
{
public:
  gender_pimpl ()
    : gender_pskel (&base_impl_)
  {
  }

  virtual void
  post_gender ()
  {
    std::string s = post_string ();
    cout << "gender: " << s << endl;
  }

private:
  xml_schema::string_pimpl base_impl_;
};
  

While the code is quite short, there is a lot going on. First, notice that we define a member variable base_impl_ of type xml_schema::string_pimpl and then pass it to the gender_pskel's constructor. We have encountered xml_schema::string_pimpl already; it is an implementation of the xml_schema::string_pskel parser skeleton for built-in XML Schema type string. By passing base_impl_ to the gender_pskel's constructor we provide an implementation for the part of the parser skeleton that is inherited from string_pskel.

This is another common theme in the C++/Parser programming model: reusing implementations of the base parsers in the derived ones. In our case, string_pimpl will do all the dirty work of extracting the data and we can just get it at the end with the call to post_string(). For more information on parser implementation reuse refer to Section 5.6, "Parser Reuse".

In case you are curious, here are the definitions for xml_schema::string_pskel and xml_schema::string_pimpl:

namespace xml_schema
{
  class string_pskel: public parser_simple_content
  {
  public:
    virtual std::string
    post_string () = 0;
  };

  class string_pimpl: public string_pskel
  {
  public:
    virtual void
    _pre ();

    virtual void
    _characters (const xml_schema::ro_string&);

    virtual std::string
    post_string ();

  protected:
    std::string str_;
  };
}
  

There are three new pieces in this code that we haven't seen yet. Those are the parser_simple_content class and the _pre() and _characters() functions. The parser_simple_content class is defined in the XSD/e runtime and is a base class for all parser skeletons that conform to the simple content model in XML Schema. Types with the simple content model cannot have nested elements—only text and attributes. There is also the parser_complex_content class which corresponds to the complex content mode (types with nested elements, for example, person from people.xsd).

The _pre() function is a parser callback. Remember we talked about the pre() and post_*() callbacks in the previous chapter? There are actually two more callbacks with similar roles: _pre() and _post (). As a result, each parser skeleton has four special callbacks:

  virtual void
  pre ();

  virtual void
  _pre ();

  virtual void
  _post ();

  virtual void
  post_name ();
  

pre() and _pre() are initialization callbacks. They get called in that order before a new instance of the type is about to be parsed. The difference between pre() and _pre() is conventional: pre() can be completely overridden by a derived parser. The derived parser can also override _pre() but has to always call the original version. This allows you to partition initialization into customizable and required parts.

Similarly, _post() and post_name() are finalization callbacks with exactly the same semantics: post_name() can be completely overridden by the derived parser while the original _post() should always be called.

The final bit we need to discuss in this section is the _characters() function. As you might have guessed, it is also a callback. A low-level one that delivers raw character content for the type being parsed. You will seldom need to use this callback directly. Using implementations for the built-in parsers provided by the XSD/e runtime is usually a simpler and more convenient alternative.

At this point you might be wondering why some post_*() callbacks, for example post_string(), return some data while others, for example post_gender(), have void as a return type. This is a valid concern and it will be addressed in the next chapter.

3.2 Implementing the Person Parser

The generated person_pskel parser skeleton looks like this:

class person_pskel: public xml_schema::parser_complex_content
{
public:
  // Parser callbacks. Override them in your implementation.
  //
  virtual void
  pre ();

  virtual void
  first_name (const std::string&);

  virtual void
  last_name (const std::string&);

  virtual void
  gender ();

  virtual void
  age (short);

  virtual void
  post_person ();

  // Parser construction API.
  //
  void
  first_name_parser (xml_schema::string_pskel&);

  void
  last_name_parser (xml_schema::string_pskel&);

  void
  gender_parser (gender_pskel&);

  void
  age_parser (xml_schema::short_pskel&);

  void
  parsers (xml_schema::string_pskel& /* first-name */,
           xml_schema::string_pskel& /* last-name */,
           gender_pskel&             /* gender */,
           xml_schema::short_pskel&  /* age */);
};
  

As you can see, we have a parser callback for each of the nested elements found in the person XML Schema type. The implementation of this parser is straightforward:

class person_pimpl: public person_pskel
{
public:
  virtual void
  first_name (const std::string& n)
  {
    cout << "first: " << f << endl;
  }

  virtual void
  last_name (const std::string& l)
  {
    cout << "last: " << l << endl;
  }

  virtual void
  age (short a)
  {
    cout << "age: " << a << endl;
  }
};
  

Notice that we didn't override the gender() callback because all the printing is done by gender_pimpl.

3.3 Implementing the People Parser

The generated people_pskel parser skeleton looks like this:

class people_pskel: public xml_schema::parser_complex_content
{
public:
  // Parser callbacks. Override them in your implementation.
  //
  virtual void
  pre ();

  virtual void
  person ();

  virtual void
  post_people ();

  // Parser construction API.
  //
  void
  person_parser (person_pskel&);

  void
  parsers (person_pskel& /* person */);
};
  

The person() callback will be called after parsing each person element. While person_pimpl does all the printing, one useful thing we can do in this callback is to print an extra newline after each person record so that our output is more readable:

class people_pimpl: public people_pskel
{
public:
  virtual void
  person ()
  {
    cout << endl;
  }
};
  

Now it is time to put everything together.

3.4 Connecting the Parsers Together

At this point we have all the individual parsers implemented and can proceed to assemble them into a complete parser for our XML vocabulary. The first step is to instantiate all the individual parsers that we will need:

xml_schema::short_pimpl short_p;
xml_schema::string_pimpl string_p;

gender_pimpl gender_p;
person_pimpl person_p;
people_pimpl people_p;
  

Notice that our schema uses two built-in XML Schema types: string for the first-name and last-name elements as well as short for age. We will use predefined parsers that come with the XSD/e runtime to handle these types. The next step is to connect all the individual parsers. We do this with the help of functions defined in the parser skeletons and marked with the "Parser Construction API" comment. One way to do it is to connect each individual parser by calling the *_parser() functions:

person_p.first_name_parser (string_p);
person_p.last_name_parser (string_p);
person_p.gender_parser (gender_p);
person_p.age_parser (short_p);

people_p.person_parser (person_p);
  

You might be wondering what happens if you do not provide a parser by not calling one of the *_parser() functions. In that case the corresponding XML content will be skipped, including validation. This is an efficient way to ignore parts of the document that you are not interested in.

An alternative, shorter, way to connect the parsers is by using the parsers() functions which connects all the parsers for a given type at once:

person_p.parsers (string_p, string_p, gender_p, short_p);
people_p.parsers (person_p);
  

The following figure illustrates the resulting connections. Notice the correspondence between return types of the post_*() functions and argument types of element callbacks that are connected by the arrows.

The last step is the construction of the document parser and invocation of the complete parser on our sample XML instance:

xml_schema::document_pimpl doc_p (people_p, "people");

people_p.pre ();
doc_p.parse ("people.xml");
people_p.post_people ();
  

Let's consider xml_schema::document_pimpl in more detail. While the exact definition of this class varies depending on the mapping configuration, here is the part relevant to our example:

namespace xml_schema
{
  class document_pimpl
  {
  public:
    document_pimpl (xml_schema::parser_base&,
                    const std::string& root_element_name);

    document_pimpl (xml_schema::parser_base&,
                    const std::string& root_element_namespace,
                    const std::string& root_element_name);

    void
    parse (const std::string& file);

    void
    parse (std::istream&);

    void
    parse (const void* data, size_t size, bool last);
  };
}
  

xml_schema::document_pimpl is a root parser for the vocabulary. The first argument to its constructors is the parser for the type of the root element (people_pimpl in our case). Because a type parser is only concerned with the element's content and not with the element's name, we need to specify the root element name somewhere. That's what is passed as the second and third arguments to the document_pimpl's constructors.

There are also three overloaded parse() function defined in the document_pimpl class. The first version parses a local file identified by a name. The second version reads the data from an input stream. The last version allows you to parse the data directly from a buffer, one chunk at a time. You can call this function multiple times with the final call having the last argument set to true. For more information on the xml_schema::document_pimpl class refer to Chapter 7, "Document Parser and Error Handling".

Let's now consider a step-by-step list of actions that happen as we parse through people.xml. The content of people.xml is repeated below for convenience.

<?xml version="1.0"?>
<people>
  <person>
    <first-name>John</first-name>
    <last-name>Doe</last-name>
    <gender>male</gender>
    <age>32</age>
  </person>
  <person>
    <first-name>Jane</first-name>
    <last-name>Doe</last-name>
    <gender>female</gender>
    <age>28</age>
  </person>
</people>
  

1. people_p.pre() is called from main(). We did not provide any implementation for this callback so this call is a no-op.
2. doc_p.parse("people.xml") is called from main(). The parser opens the file and starts parsing its content.
3. The parser encounters the root element. doc_p verifies that the root element is correct and calls _pre() on people_p which is also a no-op. Parsing is now delegated to people_p.
4. The parser encounters the person element. people_p determines that person_p is responsible for parsing this element. pre() and _pre() callbacks are called on person_p. Parsing is now delegated to person_p.
5. The parser encounters the first-name element. person_p determines that string_p is responsible for parsing this element. pre() and _pre() callbacks are called on string_p. Parsing is now delegated to string_p.
6. The parser encounters character content consisting of "John". The _characters() callback is called on string_p.
7. The parser encounters the end of first-name element. The _post() and post_string() callbacks are called on string_p. The first_name() callback is called on person_p with the return value of post_string(). The first_name() implementation prints "first: John" to cout. Parsing is now returned to person_p.
8. Steps analogous to 5-7 are performed for the last-name, gender, and age elements.
9. The parser encounters the end of person element. The _post() and post_person() callbacks are called on person_p. The person() callback is called on people_p. The person() implementation prints a new line to cout. Parsing is now returned to people_p.
10. Steps 4-9 are performed for the second person element.
11. The parser encounters the end of people element. The _post() callback is called on people_p. The doc_p.parse("people.xml") call returns to main().
12. people_p.post_people() is called from main() which is a no-op.

4 Type Maps

There are many useful things you can do inside parser callbacks as they are right now. There are, however, times when you want to propagate some information from one parser to another or to the caller of the parser. One common task that would greatly benefit from such a possibility is building a tree-like in-memory object model of the data stored in XML. During execution, each individual sub-parser would create a sub-tree and return it to its parent parser which can then incorporate this sub-tree into the whole tree.

In this chapter we will discuss the mechanisms offered by the C++/Parser mapping for returning information from individual parsers and see how to use them to build an object model of our people vocabulary.

4.1 Object Model

An object model for our person record example could look like this (saved in the people.hxx file):

#include <string>
#include <vector>

enum gender
{
  male,
  female
};

class person
{
public:
  person (const std::string& first,
          const std::string& last,
          ::gender gender,
          short age)
    : first_ (first), last_ (last),
      gender_ (gender), age_ (age)
  {
  }

  const std::string&
  first () const
  {
    return first_;
  }

  const std::string&
  last () const
  {
    return last_;
  }

  ::gender
  gender () const
  {
    return gender_;
  }

  short
  age () const
  {
    return age_;
  }

private:
  std::string first_;
  std::string last_;
  ::gender gender_;
  short age_;
};

typedef std::vector<person> people;
  

While it is clear which parser is responsible for which part of the object model, it is not exactly clear how, for example, gender_pimpl will deliver gender to person_pimpl. You might have noticed that string_pimpl manages to deliver its value to the first_name() callback of person_pimpl. Let's see how we can utilize the same mechanism to propagate our own data.

There is a way to tell the XSD/e compiler that you want to exchange data between parsers. More precisely, for each type defined in XML Schema, you can tell the compiler two things. First, the return type of the post_*() callback in the parser skeleton generated for this type. And, second, the argument type for callbacks corresponding to elements and attributes of this type. For example, for XML Schema type gender we can specify the return type for post_gender() in the gender_pskel skeleton and the argument type for the gender() callback in the person_pskel skeleton. As you might have guessed, the generated code will then pass the return value from the post_*() callback as an argument to the element or attribute callback.

The way to tell the XSD/e compiler about these XML Schema to C++ mappings is with type map files. Here is a simple type map for the gender type from the previous paragraph.

include "people.hxx";
gender ::gender ::gender;
  

The first line indicates that the generated code must include people.hxx in order to get the definition for the gender type. The second line specifies that both argument and return types for the gender XML Schema type should be the ::gender C++ enum (we use fully-qualified C++ names to avoid name clashes). The next section will describe the type map format in detail. We save this type map in people.map and then translate our schemas with the --type-map option to let the XSD/e compiler know about our type map:

$ xsde cxx-parser --type-map people.map people.xsd
  

If we now look at the generated people-pskel.hxx, we will see the following changes in the gender_pskel and person_pskel skeletons:

#include "people.hxx"

class gender_pskel: public xml_schema::string_pskel
{
  virtual ::gender
  post_gender () = 0;

  ...
};

class person_pskel: public xml_schema::parser_complex_content
{
  virtual void
  gender (::gender);

  ...
};
  

Notice that #include "people.hxx" was added to the generated header file from the type map to provide the definition for the gender enum.

4.2 Type Map File Format

Type map files are used to define a mapping between XML Schema and C++ types. The compiler uses this information to determine return types of post_*() callbacks in parser skeletons corresponding to XML Schema types as well as argument types for callbacks corresponding to elements and attributes of these types.

The compiler has a set of predefined mapping rules that map the built-in XML Schema types to suitable C++ types (discussed below) and all other types to void. By providing your own type maps you can override these predefined rules. The format of the type map file is presented below:

namespace <schema-namespace> [<cxx-namespace>]
{
  (include <file-name>;)*
  ([type] <schema-type> <cxx-ret-type> [<cxx-arg-type>];)*
}
  

Both <schema-namespace> and <schema-type> are regex patterns while <cxx-namespace>, <cxx-ret-type>, and <cxx-arg-type> are regex pattern substitutions. All names can be optionally enclosed in " ", for example, to include white-spaces.

<schema-namespace> determines XML Schema namespace. Optional <cxx-namespace> is prefixed to every C++ type name in this namespace declaration. <cxx-ret-type> is a C++ type name that is used as a return type for the post_*() callback. Optional <cxx-arg-type> is an argument type for callbacks corresponding to elements and attributes of this type. If <cxx-arg-type> is not specified, it defaults to <cxx-ret-type> if <cxx-ret-type> ends with * or & (that is, it is a pointer or a reference) and const <cxx-ret-type>& otherwise. <file-name> is a file name either in the " " or < > format and is added with the #include directive to the generated code.

The # character starts a comment that ends with a new line or end of file. To specify a name that contains # enclose it in " ". For example:

namespace http://www.example.com/xmlns/my my
{
  include "my.hxx";

  # Pass apples by value.
  #
  apple apple;

  # Pass oranges as pointers.
  #
  orange orange_t*;
}
  

In the example above, for the http://www.example.com/xmlns/my#orange XML Schema type, the my::orange_t* C++ type will be used as both return and argument types.

Several namespace declarations can be specified in a single file. The namespace declaration can also be completely omitted to map types in a schema without a namespace. For instance:

include "my.hxx";
apple apple;

namespace http://www.example.com/xmlns/my
{
  orange "const orange_t*";
}
  

The compiler has a number of predefined mapping rules for the built-in XML Schema types which can be presented as the following map files:

namespace http://www.w3.org/2001/XMLSchema
{
  boolean bool bool;

  byte "signed char" "signed char";
  unsignedByte "unsigned char" "unsigned char";

  short short short;
  unsignedShort "unsigned short" "unsigned short";

  int int int;
  unsignedInt "unsigned int" "unsigned int";

  long "long long" "long long";
  unsignedLong "unsigned long long" "unsigned long long";

  integer long long;

  negativeInteger long long;
  nonPositiveInteger long long;

  positiveInteger "unsigned long" "unsigned long";
  nonNegativeInteger "unsigned long" "unsigned long";

  float float float;
  double double double;
  decimal double double;

  NMTOKENS xml_schema::string_sequence*;
  IDREFS xml_schema::string_sequence*;

  base64Binary xml_schema::buffer*;
  hexBinary xml_schema::buffer*;

  date xml_schema::date;
  dateTime xml_schema::date_time;
  duration xml_schema::duration;
  gDay xml_schema::gday;
  gMonth xml_schema::gmonth;
  gMonthDay xml_schema::gmonth_day;
  gYear xml_schema::gyear;
  gYearMonth xml_schema::gyear_month;
  time xml_schema::time;
}
  

If STL is enabled (Section 5.1, "Standard Template Library"), the following mapping is used for the string-based XML Schema built-in types:

namespace http://www.w3.org/2001/XMLSchema
{
  include <string>;

  anySimpleType std::string;

  string std::string;
  normalizedString std::string;
  token std::string;
  Name std::string;
  NMTOKEN std::string;
  NCName std::string;
  ID std::string;
  IDREF std::string;
  language std::string;
  anyURI std::string;

  QName xml_schema::qname;
}
  

Otherwise, a C string-based mapping is used:

namespace http://www.w3.org/2001/XMLSchema
{
  anySimpleType char*;

  string char*;
  normalizedString char*;
  token char*;
  Name char*;
  NMTOKEN char*;
  NCName char*;
  ID char*;
  IDREF char*;
  language char*;
  anyURI char*;

  QName xml_schema::qname*;
}
  

For more information about the mapping of the built-in XML Schema types to C++ types refer to Chapter 6, "Built-In XML Schema Type Parsers". The last predefined rule maps anything that wasn't mapped by previous rules to void:

namespace .*
{
  .* void void;
}
  

When you provide your own type maps with the --type-map option, they are evaluated first. This allows you to selectively override any of the predefined rules. Note also that if you change the mapping of a built-in XML Schema type then it becomes your responsibility to provide the corresponding parser skeleton and implementation in the xml_schema namespace. You can include the custom definitions into the generated header file using the --hxx-prologue-* options.

4.3 Parser Implementations

With the knowledge from the previous section, we can proceed with creating a type map that maps types in the people.xsd schema to our object model classes in people.hxx. In fact, we already have the beginning of our type map file in people.map. Let's extend it with the rest of the types:

include "people.hxx";

gender ::gender ::gender;
person ::person;
people ::people;
  

A few things to note about this type map. We did not provide the argument types for person and people because the default constant reference is exactly what we need. We also did not provide any mappings for built-in XML Schema types string and short because they are handled by the predefined rules and we are happy with the result. Note also that all C++ types are fully qualified. This is done to avoid potential name conflicts in the generated code. Now we can recompile our schema and move on to implementing the parsers:

$ xsde cxx-parser --type-map people.map people.xsd
  

Here is the implementation of our three parsers in full. One way to save typing when implementing your own parsers is to open the generated code and copy the signatures of parser callbacks into your code. Or you could always auto generate the sample implementations and fill them with your code.

#include "people-pskel.hxx"

class gender_pimpl: public gender_pskel
{
public:
  gender_pimpl ()
    : gender_pskel (&base_impl_)
  {
  }

  virtual ::gender
  post_gender ()
  {
    return post_string () == "male" ? male : female;
  }

private:
  xml_schema::string_pimpl base_impl_;
};

class person_pimpl: public person_pskel
{
public:
  virtual void
  first_name (const std::string& f)
  {
    first_ = f;
  }

  virtual void
  last_name (const std::string& l)
  {
    last_ = l;
  }

  virtual void
  gender (::gender g)
  {
    gender_ = g;
  }

  virtual void
  age (short a)
  {
    age_ = a;
  }

  virtual ::person
  post_person ()
  {
    return ::person (first_, last_, gender_, age_);
  }

private:
  std::string first_;
  std::string last_;
  ::gender gender_;
  short age_;
};

class people_pimpl: public people_pskel
{
public:
  virtual void
  person (const ::person& p)
  {
    people_.push_back (p);
  }

  virtual ::people
  post_people ()
  {
    ::people r;
    r.swap (people_);
    return r;
  }

private:
  ::people people_;
};
  

This code fragment should look familiar by now. Just note that all the post_*() callbacks now have return types instead of void. Here is the implementation of the test driver for this example:

#include <iostream>

using namespace std;

int
main (int argc, char* argv[])
{
  // Construct the parser.
  //
  xml_schema::short_pimpl short_p;
  xml_schema::string_pimpl string_p;

  gender_pimpl gender_p;
  person_pimpl person_p;
  people_pimpl people_p;

  person_p.parsers (string_p, string_p, gender_p, short_p);
  people_p.parsers (person_p);

  // Parse the document to obtain the object model.
  //
  xml_schema::document_pimpl doc_p (people_p, "people");

  people_p.pre ();
  doc_p.parse (argv[1]);
  people ppl = people_p.post_people ();

  // Print the object model.
  //
  for (people::iterator i (ppl.begin ()); i != ppl.end (); ++i)
  {
    cout << "first:  " << i->first () << endl
         << "last:   " << i->last () << endl
         << "gender: " << (i->gender () == male ? "male" : "female") << endl
         << "age:    " << i->age () << endl
         << endl;
  }
}
  

The parser creation and assembly part is exactly the same as in the previous chapter. The parsing part is a bit different: post_people() now has a return value which is the complete object model. We store it in the ppl variable. The last bit of the code simply iterates over the people vector and prints the information for each person. We save the last two code fragments to driver.cxx and proceed to compile and test our new application:

$ c++ -I.../libxsde -c driver.cxx people-pskel.cxx
$ c++ -o driver driver.o people-pskel.o .../libxsde/xsde/libxsde.a
$ ./driver people.xml
first:  John
last:   Doe
gender: male
age:    32

first:  Jane
last:   Doe
gender: female
age:    28
  

5 Mapping Configuration

The Embedded C++/Parser mapping has a number of configuration parameters that determine the overall properties and behavior of the generated code, such as the use of Standard Template Library (STL), Input/Output Stream Library (iostream), C++ exceptions, XML Schema validation, 64-bit integer types, parser implementation reuse styles, and support for XML Schema polymorphism. Previous chapters assumed that the use of STL, iostream, C++ exceptions, and XML Schema validation were enabled. This chapter will discuss the changes in the Embedded C++/Parser programming model that result from the changes to these configuration parameters. A complete example that uses the minimal mapping configuration is presented at the end of this chapter.

In order to enable or disable a particular feature, the corresponding configuration parameter should be set accordingly in the XSD/e runtime library as well as specified during schema compilation with the XSD/e command line options as described in the XSD/e Compiler Command Line Manual.

While the XML documents can use various encodings, the Embedded C++/Parser mapping always delivers character data to the application in the same encoding. The application encoding can either be UTF-8 (default) or ISO-8859-1. To select a particular encoding, configure the XSD/e runtime library accordingly and pass the --char-encoding option to the XSD/e compiler when translating your schemas.

When using ISO-8859-1 as the application encoding, XML documents being parsed may contain characters with Unicode values greater than 0xFF which are unrepresentable in the ISO-8859-1 encoding. By default, in such situations parsing will terminate with an error. However, you can suppress the error by providing a replacement character that should be used instead of unrepresentable characters, for example:

xml_schema::iso8859_1::unrep_char ('?');
  

To revert to the default behavior, set the replacement character to '\0'.

The Embedded C++/Parser mapping includes built-in support for XML documents encoded in UTF-8, UTF-16, ISO-8859-1, and US-ASCII. Other encodings can be supported by providing application-specific decoder functions.

5.1 Standard Template Library

To disable the use of STL you will need to configure the XSD/e runtime without support for STL as well as pass the --no-stl option to the XSD/e compiler when translating your schemas. When STL is disabled, all string-based XML Schema types are mapped to C-style char* instead of std::string, as described in Section 4.2, "Type Map File Format". The following code fragment shows changes in the signatures of first_name() and last_name() callbacks from the person record example.

class person_pskel
{
public:
  virtual void
  first_name (char*);

  virtual void
  last_name (char*);

  ...
};
  

Note that it is your responsibility to eventually release the memory associated with these strings using operator delete[].

5.2 Input/Output Stream Library

To disable the use of iostream you will need to configure the XSD/e runtime library without support for iostream as well as pass the --no-iostream option to the XSD/e compiler when translating your schemas. When iostream is disabled, the following two parse() functions in the xml_schema::document_pimpl class become unavailable:

  void
  parse (const std::string& file);

  void
  parse (std::istream&);
  

Leaving you with only one function in the form:

  void
  parse (const void* data, size_t size, bool last);
  

See Section 7.1, "Document Parser" for more information on the semantics of these functions.

5.3 C++ Exceptions

To disable the use of C++ exceptions, you will need to configure the XSD/e runtime without support for exceptions as well as pass the --no-exceptions option to the XSD/e compiler when translating your schemas. When C++ exceptions are disabled, the error conditions are indicated with error codes instead of exceptions, as described in Section 7.3, "Error Codes".

5.4 XML Schema Validation

To disable support for XML Schema validation, you will need to configure the XSD/e runtime accordingly as well as pass the --suppress-validation option to the XSD/e compiler when translating your schemas. Disabling XML Schema validation allows to further increase the parsing performance and reduce footprint in cases where XML instances are known to be valid.

5.5 64-bit Integer Type

By default the 64-bit long and unsignedLong XML Schema built-in types are mapped to the 64-bit long long and unsigned long long fundamental C++ types. To disable the use of these types in the mapping you will need to configure the XSD/e runtime accordingly as well as pass the --no-long-long option to the XSD/e compiler when translating your schemas. When the use of 64-bit integral C++ types is disabled the long and unsignedLong XML Schema built-in types are mapped to long and unsigned long fundamental C++ types.

5.6 Parser Reuse

When one type in XML Schema inherits from another, it is often desirable to be able to reuse the parser implementation corresponding to the base type in the parser implementation corresponding to the derived type. XSD/e provides support for two parser reuse styles: the so-called mixin (generated when the --reuse-style-mixin option is specified) and tiein (generated by default) styles.

The compiler can also be instructed not to generate any support for parser reuse with the --reuse-style-none option. This is mainly useful to further reduce the generated code size when your vocabulary does not use inheritance or when you plan to implement each parser from scratch. Note also that the XSD/e runtime should be configured in accordance with the parser reuse style used in the generated code. The remainder of this section discusses the mixin and tiein parser reuse styles in more detail.

To provide concrete examples for each reuse style we will use the following schema fragment:

<xs:complexType name="person">
  <xs:sequence>
    <xs:element name="first-name" type="xs:string"/>
    <xs:element name="last-name" type="xs:string"/>
    <xs:element name="age" type="xs:short"/>
  </xs:sequence>
</xs:complexType>

<xs:complexType name="emplyee">
  <complexContent>
    <extension base="person">
      <xs:sequence>
        <xs:element name="position" type="xs:string"/>
        <xs:element name="salary" type="xs:unsignedLong"/>
      </xs:sequence>
    </extension>
  </complexContent>
</xs:complexType>
  

The mixin parser reuse style uses the C++ mixin idiom that relies on multiple and virtual inheritance. Because virtual inheritance can result in a significant object code size increase, this reuse style should be considered when such an overhead is acceptable and/or the vocabulary consists of only a handful of types. When the mixin reuse style is used, the generated parser skeletons use virtual inheritance, for example:

class person_pskel: public virtual parser_complex_content
{
  ...
};

class employee_pskel: public virtual person_pskel
{
  ...
};
  

When you implement the base parser you also need to use virtual inheritance. The derived parser is implemented by inheriting from both the derived parser skeleton and the base parser implementation (that is, mixing in the base parser implementation), for example:

class person_pimpl: public virtual person_pskel
{
  ...
};

class employee_pimpl: public employee_pskel,
                      public person_pimpl
{
  ...
};
  

The tiein parser reuse style uses delegation and normally results in a significantly smaller object code while being almost as convenient to use as the mixin style. When the tiein reuse style is used, the generated derived parser skeleton declares a constructor which allows you to specify the implementation of the base parser:

class person_pskel: public parser_complex_content
{
  ...
};

class employee_pskel: public person_pskel
{
public:
  employee_pskel (person_pskel* base_impl)

  ...
};
  

If you pass the implementation of the base parser to this constructor then the generated code will transparently forward all the callbacks corresponding to the base parser skeleton to this implementation. You can also pass 0 to this constructor in which case you will need to implement the derived parser from scratch. The following example shows how we could implement the person and employee parsers using the tiein style:

class person_pimpl: public person_pskel
{
  ...
};

class employee_pimpl: public employee_pskel
{
public:
  employee_pimpl ()
    : employee_pskel (&base_impl_)
  {
  }

  ...

private:
  person_pimpl base_impl_;
};
  

Note that you cannot use the tied in base parser instance (base_impl_ in the above code) for parsing anything except the derived type.

The ability to override the base parser callbacks in the derived parser is also available in the tiein style. For example, the following code fragment shows how we can override the age() callback if we didn't like the implementation provided by the base parser:

class employee_pimpl: public employee_pskel
{
public:
  employee_pimpl ()
    : employee_pskel (&base_impl_)
  {
  }

  virtual void
  age (short a)
  {
    ...
  }

  ...

private:
  person_pimpl base_impl_;
};
  

In the above example the age element will be handled by emplyee_pimpl while the first-name and last-name callbacks will still go to base_impl_.

It is also possible to inherit from the base parser implementation instead of declaring it as a member variable. This can be useful if you need to access protected members in the base implementation or need to override a virtual function that is not part of the parser skeleton interface. Note, however, that in this case you will need to resolve a number of ambiguities with explicit qualifications or using-declarations. For example:

class person_pimpl: public person_pskel
{
  ...
protected:
  virtual person*
  create ()
  {
    return new person ();
  }
};

class employee_pimpl: public employee_pskel,
                      public person_pimpl
{
public:
  employee_pimpl ()
    : employee_pskel (static_cast<person_pimpl*> (this))
  {
  }

  // Resolve ambiguities.
  //
  using emplyee_pskel::parsers;

  ...

protected:
  virtual employee*
  create ()
  {
    return new employee ();
  }
};
  

5.7 Support for Polymorphism

By default the XSD/e compiler generates non-polymorphic code. If your vocabulary uses XML Schema polymorphism in the form of xsi:type and/or substitution groups, then you will need to configure the XSD/e runtime with support for polymorphism, compile your schemas with the --generate-polymorphic option to produce polymorphism-aware code, as well as pass true as the last argument to the xml_schema::document_pimpl's constructors. If some of your schemas do not require support for polymorphism then you can compile them with the --runtime-polymorphic option and still use the XSD/e runtime configured with polymorphism support.

When using the polymorphism-aware generated code, you can specify several parsers for a single element by passing a parser map instead of an individual parser to the parser connection function for the element. One of the parsers will then be looked up and used depending on the xsi:type attribute value or an element name from a substitution group. Consider the following schema as an example:

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">

  <xs:complexType name="person">
    <xs:sequence>
      <xs:element name="name" type="xs:string"/>
    </xs:sequence>
  </xs:complexType>

  <!-- substitution group root -->
  <xs:element name="person" type="person"/>

  <xs:complexType name="superman">
    <xs:complexContent>
      <xs:extension base="person">
        <xs:attribute name="can-fly" type="xs:boolean"/>
      </xs:extension>
    </xs:complexContent>
  </xs:complexType>

  <xs:element name="superman"
              type="superman"
              substitutionGroup="person"/>

  <xs:complexType name="batman">
    <xs:complexContent>
      <xs:extension base="superman">
        <xs:attribute name="wing-span" type="xs:unsignedInt"/>
      </xs:extension>
    </xs:complexContent>
  </xs:complexType>

  <xs:element name="batman"
              type="batman"
              substitutionGroup="superman"/>

  <xs:complexType name="supermen">
    <xs:sequence>
      <xs:element ref="person" maxOccurs="unbounded"/>
    </xs:sequence>
  </xs:complexType>

  <xs:element name="supermen" type="supermen"/>

</xs:schema>
  

Conforming XML documents can use the superman and batman types in place of the person type either by specifying the type with the xsi:type attributes or by using the elements from the substitution group, for instance:

<supermen xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">

  <person>
    <name>John Doe</name>
  </person>

  <superman can-fly="false">
    <name>James "007" Bond</name>
  </superman>

  <superman can-fly="true" wing-span="10" xsi:type="batman">
    <name>Bruce Wayne</name>
  </superman>

</supermen>
  

To print the data stored in such XML documents we can implement the parsers as follows:

class person_pimpl: public person_pskel
{
public:
  virtual void
  pre ()
  {
    cout << "starting to parse person" << endl;
  }

  virtual void
  name (const std::string& v)
  {
    cout << "name: " << v << endl;
  }

  virtual void
  post_person ()
  {
    cout << "finished parsing person" << endl;
  }
};

class superman_pimpl: public superman_pskel
{
public:
  superman_pimpl ()
    : superman_pskel (&base_impl_)
  {
  }

  virtual void
  pre ()
  {
    cout << "starting to parse superman" << endl;
  }

  virtual void
  can_fly (bool v)
  {
    cout << "can-fly: " << v << endl;
  }

  virtual void
  post_person ()
  {
    post_superman ();
  }

  virtual void
  post_superman ()
  {
    cout << "finished parsing superman" << endl
  }

private:
  person_pimpl base_impl_;
};

class batman_pimpl: public batman_pskel
{
public:
  batman_pimpl ()
    : batman_pskel (&base_impl_)
  {
  }

  virtual void
  pre ()
  {
    cout << "starting to parse batman" << endl;
  }

  virtual void
  wing_span (unsigned int v)
  {
    cout << "wing-span: " << v << endl;
  }

  virtual void
  post_person ()
  {
    post_superman ();
  }

  virtual void
  post_superman ()
  {
    post_batman ();
  }

  virtual void
  post_batman ()
  {
    cout << "finished parsing batman" << endl;
  }

private:
  superman_pimpl base_impl_;
};
  

Note that because the derived type parsers (superman_pskel and batman_pskel) are called via the person_pskel interface, we have to override the post_person() virtual function in superman_pimpl and batman_pimpl to call post_superman() and the post_superman() virtual function in batman_pimpl to call post_batman() (when the mixin parser reuse style is used it is not necessary to override post_person() in batman_pimpl since the suitable implementation is inherited from superman_pimpl).

The following code fragment shows how to connect the parsers together. Notice that for the person element in the supermen_p parser we specify a parser map instead of a specific parser and we pass true as the last argument to the document parser constructor to indicate that we are parsing potentially-polymorphic XML documents:

int
main (int argc, char* argv[])
{
  // Construct the parser.
  //
  xml_schema::string_pimpl string_p;
  xml_schema::boolean_pimpl boolean_p;
  xml_schema::unsigned_int_pimpl unsigned_int_p;

  person_pimpl person_p;
  superman_pimpl superman_p;
  batman_pimpl batman_p;

  xml_schema::parser_map_impl person_map (5); // 5 hashtable buckets
  supermen_pimpl supermen_p;

  person_p.parsers (string_p);
  superman_p.parsers (string_p, boolean_p);
  batman_p.parsers (string_p, boolean_p, unsigned_int_p);

  // Here we are specifying several parsers that can be used to
  // parse the person element.
  //
  person_map.insert (person_p);
  person_map.insert (superman_p);
  person_map.insert (batman_p);

  supermen_p.person_parser (person_map);

  // Parse the XML document. The last argument to the document's
  // constructor indicates that we are parsing polymorphic XML
  // documents.
  //
  xml_schema::document_pimpl doc_p (supermen_p, "supermen", true);

  supermen_p.pre ();
  doc_p.parse (argv[1]);
  supermen_p.post_supermen ();
}
  

When polymorphism-aware code is generated, each element's *_parser() function is overloaded to also accept an object of the xml_schema::parser_map type. For example, the supermen_pskel class from the above example looks like this:

class supermen_pskel: public xml_schema::parser_complex_content
{
public:

  ...

  // Parser construction API.
  //
  void
  parsers (person_pskel&);

  // Individual element parsers.
  //
  void
  person_parser (person_pskel&);

  void
  person_parser (xml_schema::parser_map&);

  ...
};
  

Note that you can specify both the individual (static) parser and the parser map. The individual parser will be used when the static element type and the dynamic type of the object being parsed are the same. This is the case, for example, when there is no xsi:type attribute and the element hasn't been substituted. Because the individual parser for an element is cached and no map lookup is necessary, it makes sense to specify both the individual parser and the parser map when most of the objects being parsed are of the static type and optimal performance is important. The following code fragment shows how to change the above example to set both the individual parser and the parser map:

int
main (int argc, char* argv[])
{
  ...

  // Here we are specifying several parsers that can be used to
  // parse the person element.
  //
  person_map.insert (superman_p);
  person_map.insert (batman_p);

  supermen_p.person_parser (person_p);
  supermen_p.person_parser (person_map);

  ...
}
  

The xml_schema::parser_map interface and the xml_schema::parser_map_impl default implementation are presented below:

namespace xml_schema
{
  class parser_map
  {
  public:
    virtual parser_base*
    find (const char* type) const = 0;

    virtual void
    reset () const = 0;
  };

  class parser_map_impl: public parser_map
  {
  public:
    parser_map_impl (size_t buckets);

    void
    insert (parser_base&);

    virtual parser_base*
    find (const char* type) const;

    virtual void
    reset () const;

  private:
    parser_map_impl (const parser_map_impl&);

    parser_map_impl&
    operator= (const parser_map_impl&);

    ...
  };
}
  

The type argument in the find() virtual function is the type name and namespace from the xsi:type attribute (the namespace prefix is resolved to the actual XML namespace) or the type of an element from the substitution group in the form "<name> <namespace>" with the space and the namespace part absent if the type does not have a namespace. You can obtain a parser's dynamic type in the same format using the _dynamic_type() function. The static type can be obtained by calling the static _static_type() function, for example person_pskel::_static_type(). Both functions return a C string (const char*) which is valid for as long as the application is running. The reset() virtual function is used to reset the parsers contained in the map (as opposed to resetting or clearing the map itself). For more information on parser resetting refer to Section 7.4, "Reusing Parsers after an Error". The following example shows how we can implement our own parser map using std::map:

#include <map>
#include <string>

class parser_map: public xml_schema::parser_map
{
public:
 void
 insert (xml_schema::parser_base& p)
 {
   map_[p._dynamic_type ()] = &p;
 }

 virtual xml_schema::parser_base*
 find (const char* type) const
 {
   map::const_iterator i = map_.find (type);
   return i != map_.end () ? i->second : 0;
 }

 virtual void
 reset () const
 {
   for (map::const_iterator i (map_.begin ()), e (map_.end ());
        i != e; ++i)
   {
     xml_schema::parser_base* p = i->second;
     p->_reset ();
   }
 }

private:
  typedef std::map<std::string, xml_schema::parser_base*> map;
  map map_;
};
  

The XSD/e runtime provides the default implementation for the xml_schema::parser_map interface, xml_schema::parser_map_impl, which is a hashmap. It requires that you specify the number of buckets it will contain and it does not support automatic table resizing. To obtain good performance the elements to buckets ratio should be between 0.7 and 0.9. It is also recommended to use prime numbers for bucket counts: 53, 97, 193, 389, 769, 1543, 3079, 6151, 12289, 24593, 49157, 98317, 196613, 393241.

If C++ exceptions are disabled (Section 5.3, "C++ Exceptions"), the xml_schema::parser_map_impl class has the following additional error querying API. It can be used to detect the out of memory errors after calls to the parser_map_impl's constructor and insert() function.

namespace xml_schema
{
  class parser_map_impl: public parser_map
  {
  public:
    enum error
    {
      error_none,
      error_no_memory
    };

    error
    _error () const;

    ...
  };
}
  

To support polymorphic parsing the XSD/e runtime and generated code maintain a number of hashmaps that contain substitution and, if XML Schema validation is enabled (Section 5.4, "XML Schema Validation"), inheritance information. Because the number of elements in these hashmaps depends on the schemas being compiled and thus is fairly static, these hashmaps do not perform automatic table resizing and instead the number of buckets is specified when the XSD/e runtime is configured. To obtain good performance the elements to buckets ratio in these hashmaps should be between 0.7 and 0.9. The recommended way to ensure this range is to add diagnostics code to your application as shown in the following example:

int
main ()
{
  // Check that the load in substitution and inheritance hashmaps
  // is not too high.
  //
#ifndef NDEBUG
  float load = xml_schema::parser_smap_elements ();
  load /= xml_schema::parser_smap_buckets ();

  if (load > 0.8)
  {
    cerr << "substitution hashmap load is " << load << endl;
    cerr << "time to increase XSDE_PARSER_SMAP_BUCKETS" << endl;
  }

  load = xml_schema::parser_imap_elements ();
  load /= xml_schema::parser_imap_buckets ();

  if (load > 0.8)
  {
    cerr << "inheritance hashmap load is " << load << endl;
    cerr << "time to increase XSDE_PARSER_IMAP_BUCKETS" << endl;
  }
#endif

  ...
}
  

Most of the code presented in this section is taken from the polymorphism example which can be found in the examples/cxx/parser/ directory of the XSD/e distribution. Handling of xsi:type and substitution groups when used on root elements requires a number of special actions as shown in the polyroot example.

5.8 Custom Allocators

By default the XSD/e runtime and generated code use the standard operators new and delete to manage dynamic memory. However, it is possible to instead use custom allocator functions provided by your application. To achieve this, configure the XSD/e runtime library to use custom allocator functions as well as pass the --custom-allocator option to the XSD/e compiler when translating your schemas. The signatures of the custom allocator functions that should be provided by your application are listed below. Their semantics should be equivalent to the standard C malloc(), realloc(), and free() functions.

extern "C" void*
xsde_alloc (size_t);

extern "C" void*
xsde_realloc (void*, size_t);

extern "C" void
xsde_free (void*);
  

Note also that when custom allocators are enabled, any dynamically-allocated object of which the XSD/e runtime or generated code assume ownership should be allocated using the custom allocation function. Similarly, if your application assumes ownership of any dynamically-allocated object returned by the XSD/e runtime or the generated code, then such an object should be disposed of using the custom deallocation function. To help with these tasks the generated xml_schema namespace defines the following two helper functions and, if C++ exceptions are enabled, automatic pointer class:

namespace xml_schema
{
  void*
  alloc (size_t);

  void
  free (void*);

  struct alloc_guard
  {
    alloc_guard (void*);
    ~alloc_guard ();

    void*
    get () const;

    void
    release ();

  private:
    ...
  };
}
  

If C++ exceptions are disabled, these functions are equivalent to xsde_alloc() and xsde_free(). If exceptions are enabled, xml_schema::alloc() throws std::bad_alloc on memory allocation failure.

The following code fragment shows how to create and destroy a dynamically-allocated object with custom allocators when C++ exceptions are disabled:

void* v = xml_schema::alloc (sizeof (type));

if (v == 0)
{
  // Handle out of memory condition.
}

type* x = new (v) type (1, 2);

...

if (x)
{
  x->~type ();
  xml_schema::free (x);
}
  

The equivalent code fragment for configurations with C++ exceptions enabled is shown below:

xml_schema::alloc_guard g (xml_schema::alloc (sizeof (type)));
type* x = new (g.get ()) type (1, 2);
g.release ();

...

if (x)
{
  x->~type ();
  xml_schema::free (x);
}
  

5.9 A Minimal Example

The following example is a re-implementation of the person records example presented in Chapter 3, "Parser Skeletons". It is intended to work without STL, iostream, and C++ exceptions. It can be found in the examples/cxx/parser/minimal/ directory of the XSD/e distribution. The people.xsd schema is compiled with the --no-stl, --no-iostream, and --no-exceptions options. The following listing presents the implementation of parser skeletons and the test driver in full.

#include <stdio.h>

#include "people-pskel.hxx"

class gender_pimpl: public gender_pskel
{
public:
  gender_pimpl ()
    : gender_pskel (&base_impl_)
  {
  }

  virtual void
  post_gender ()
  {
    char* s = post_string ();
    printf ("gender: %s\n", s);
    delete[] s;
  }

private:
  xml_schema::string_pimpl base_impl_;
};

class person_pimpl: public person_pskel
{
public:
  virtual void
  first_name (char* n)
  {
    printf ("first: %s\n", n);
    delete[] n;
  }

  virtual void
  last_name (char* n)
  {
    printf ("last: %s\n", n);
    delete[] n;
  }

  virtual void
  age (short a)
  {
    printf ("age: %hd\n", a);
  }
};

class people_pimpl: public people_pskel
{
public:
  virtual void
  person ()
  {
    // Add an extra newline after each person record.
    //
    printf ("\n");
  }
};

int
main (int argc, char* argv[])
{
  // Construct the parser.
  //
  xml_schema::short_pimpl short_p;
  xml_schema::string_pimpl string_p;

  gender_pimpl gender_p;
  person_pimpl person_p;
  people_pimpl people_p;

  person_p.parsers (string_p, string_p, gender_p, short_p);
  people_p.parsers (person_p);

  // Open the file.
  //
  FILE* f = fopen (argv[1], "rb");

  if (f == 0)
  {
    fprintf (stderr, "%s: unable to open\n", argv[1]);
    return 1;
  }

  // Parse.
  //
  typedef xml_schema::parser_error error;
  error e;
  bool io_error = false;

  do
  {
    xml_schema::document_pimpl doc_p (people_p, "people");
    if (e = doc_p._error ())
      break;

    people_p.pre ();
    if (e = people_p._error ())
      break;

    char buf[4096];
    do
    {
      size_t s = fread (buf, 1, sizeof (buf), f);

      if (s != sizeof (buf) && ferror (f))
      {
        io_error = true;
        break;
      }

      doc_p.parse (buf, s, feof (f) != 0);
      e = doc_p._error ();

    } while (!e && !feof (f));

    if (io_error || e)
      break;

    people_p.post_people ();
    e = people_p._error ();

  } while (false);

  fclose (f);

  // Handle errors.
  //

  if (io_error)
  {
    fprintf (stderr, "%s: read failure\n", argv[1]);
    return 1;
  }

  if (e)
  {
    switch (e.type ())
    {
    case error::sys:
      {
        fprintf (stderr, "%s: %s\n", argv[1], e.sys_text ());
        break;
      }
    case error::xml:
      {
        fprintf (stderr, "%s:%lu:%lu: %s\n",
                 argv[1], e.line (), e.column (), e.xml_text ());
        break;
      }
    case error::schema:
      {
        fprintf (stderr, "%s:%lu:%lu: %s\n",
                 argv[1], e.line (), e.column (), e.schema_text ());
        break;
      }
    case error::app:
      {
        fprintf (stderr, "%s:%lu:%lu: application error %d\n",
                 argv[1], e.line (), e.column (), e.app_code ());
        break;
      }
    default:
      break;
    }
    return 1;
  }
  return 0;
}
  

6 Built-In XML Schema Type Parsers

The XSD/e runtime provides parser implementations for all built-in XML Schema types as summarized in the following table. Declarations for these types are automatically included into each generated header file. As a result you don't need to include any headers to gain access to these parser implementations.

XML Schema type	Parser implementation in the xml_schema namespace	Parser return type
anyType and anySimpleType types
anyType	any_type_pimpl	void
anySimpleType	any_simple_type_pimpl	std::string or char* Section 5.1, "Standard Template Library"
fixed-length integral types
byte	byte_pimpl	signed char
unsignedByte	unsigned_byte_pimpl	unsigned char
short	short_pimpl	short
unsignedShort	unsigned_short_pimpl	unsigned short
int	int_pimpl	int
unsignedInt	unsigned_int_pimpl	unsigned int
long	long_pimpl	long long or long Section 5.5, "64-bit Integer Type"
unsignedLong	unsigned_long_pimpl	unsigned long long or unsigned long Section 5.5, "64-bit Integer Type"
arbitrary-length integral types
integer	integer_pimpl	long
nonPositiveInteger	non_positive_integer_pimpl	long
nonNegativeInteger	non_negative_integer_pimpl	unsigned long
positiveInteger	positive_integer_pimpl	unsigned long
negativeInteger	negative_integer_pimpl	long
boolean types
boolean	boolean_pimpl	bool
fixed-precision floating-point types
float	float_pimpl	float
double	double_pimpl	double
arbitrary-precision floating-point types
decimal	decimal_pimpl	double
string-based types
string	string_pimpl	std::string or char* Section 5.1, "Standard Template Library"
normalizedString	normalized_string_pimpl	std::string or char* Section 5.1, "Standard Template Library"
token	token_pimpl	std::string or char* Section 5.1, "Standard Template Library"
Name	name_pimpl	std::string or char* Section 5.1, "Standard Template Library"
NMTOKEN	nmtoken_pimpl	std::string or char* Section 5.1, "Standard Template Library"
NCName	ncname_pimpl	std::string or char* Section 5.1, "Standard Template Library"
language	language_pimpl	std::string or char* Section 5.1, "Standard Template Library"
qualified name
QName	qname_pimpl	xml_schema::qname or xml_schema::qname* Section 6.1, "QName Parser"
ID/IDREF types
ID	id_pimpl	std::string or char* Section 5.1, "Standard Template Library"
IDREF	idref_pimpl	std::string or char* Section 5.1, "Standard Template Library"
list types
NMTOKENS	nmtokens_pimpl	xml_schema::string_sequence* Section 6.2, "NMTOKENS and IDREFS Parsers"
IDREFS	idrefs_pimpl	xml_schema::string_sequence* Section 6.2, "NMTOKENS and IDREFS Parsers"
URI types
anyURI	uri_pimpl	std::string or char* Section 5.1, "Standard Template Library"
binary types
base64Binary	base64_binary_pimpl	xml_schema::buffer* Section 6.3, "base64Binary and hexBinary Parsers"
hexBinary	hex_binary_pimpl	xml_schema::buffer* Section 6.3, "base64Binary and hexBinary Parsers"
date/time types
date	date_pimpl	xml_schema::date Section 6.5, "date Parser"
dateTime	date_time_pimpl	xml_schema::date_time Section 6.6, "dateTime Parser"
duration	duration_pimpl	xml_schema::duration Section 6.7, "duration Parser"
gDay	gday_pimpl	xml_schema::gday Section 6.8, "gDay Parser"
gMonth	gmonth_pimpl	xml_schema::gmonth Section 6.9, "gMonth Parser"
gMonthDay	gmonth_day_pimpl	xml_schema::gmonth_day Section 6.10, "gMonthDay Parser"
gYear	gyear_pimpl	xml_schema::gyear Section 6.11, "gYear Parser"
gYearMonth	gyear_month_pimpl	xml_schema::gyear_month Section 6.12, "gYearMonth Parser"
time	time_pimpl	xml_schema::time Section 6.13, "time Parser"

6.1 QName Parser

The return type of the qname_pimpl parser implementation is either xml_schema::qname when STL is enabled (Section 5.1, "Standard Template Library") or xml_schema::qname* when STL is disabled. The qname class represents an XML qualified name. When the return type is xml_schema::qname*, the returned object is dynamically allocated with operator new and should eventually be deallocated with operator delete. With STL enabled, the qname type has the following interface:

namespace xml_schema
{
  class qname
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    qname ();

    explicit
    qname (const std::string& name);
    qname (const std::string& prefix, const std::string& name);

    void
    swap (qname&);

    const std::string&
    prefix () const;

    std::string&
    prefix ();

    void
    prefix (const std::string&);

    const std::string&
    name () const;

    std::string&
    name ();

    void
    name (const std::string&);
  };

  bool
  operator== (const qname&, const qname&);

  bool
  operator!= (const qname&, const qname&);
}
  

When STL is disabled and C++ exceptions are enabled (Section 5.3, "C++ Exceptions"), the qname type has the following interface:

namespace xml_schema
{
  class qname
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    qname ();

    explicit
    qname (char* name);
    qname (char* prefix, char* name);

    void
    swap (qname&);

  private:
    qname (const qname&);

    qname&
    operator= (const qname&);

  public:
    char*
    prefix ();

    const char*
    prefix () const;

    void
    prefix (char*);

    void
    prefix_copy (const char*);

    char*
    prefix_detach ();

  public:
    char*
    name ();

    const char*
    name () const;

    void
    name (char*);

    void
    name_copy (const char*);

    char*
    name_detach ();
  };

  bool
  operator== (const qname&, const qname&);

  bool
  operator!= (const qname&, const qname&);
}

The modifier functions and constructors that have the char* argument assume ownership of the passed strings which should be allocated with operator new char[] and will be deallocated with operator delete[] by the qname object. If you detach the underlying prefix or name strings, then they should eventually be deallocated with operator delete[].

Finally, if both STL and C++ exceptions are disabled, the qname type has the following interface:

namespace xml_schema
{
  class qname
  {
  public:
    enum error
    {
      error_none,
      error_no_memory
    };

    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    qname ();

    explicit
    qname (char* name);
    qname (char* prefix, char* name);

    void
    swap (qname&);

  private:
    qname (const qname&);

    qname&
    operator= (const qname&);

  public:
    char*
    prefix ();

    const char*
    prefix () const;

    void
    prefix (char*);

    error
    prefix_copy (const char*);

    char*
    prefix_detach ();

  public:
    char*
    name ();

    const char*
    name () const;

    void
    name (char*);

    error
    name_copy (const char*);

    char*
    name_detach ();
  };

  bool
  operator== (const qname&, const qname&);

  bool
  operator!= (const qname&, const qname&);
}
  

6.2 NMTOKENS and IDREFS Parsers

The return type of the nmtokens_pimpl and idrefs_pimpl parser implementations is xml_schema::string_sequence*. The returned object is dynamically allocated with operator new and should eventually be deallocated with operator delete. With STL and C++ exceptions enabled (Section 5.1, "Standard Template Library", Section 5.3, "C++ Exceptions"), the string_sequence type has the following interface:

namespace xml_schema
{
  class string_sequence
  {
  public:
    typedef std::string         value_type;
    typedef std::string*        pointer;
    typedef const std::string*  const_pointer;
    typedef std::string&        reference;
    typedef const std::string&  const_reference;

    typedef size_t              size_type;
    typedef ptrdiff_t           difference_type;

    typedef std::string*        iterator;
    typedef const std::string*  const_iterator;

  public:
    string_sequence ();

    void
    swap (string_sequence&);

  private:
    string_sequence (string_sequence&);

    string_sequence&
    operator= (string_sequence&);

  public:
    iterator
    begin ();

    const_iterator
    begin () const;

    iterator
    end ();

    const_iterator
    end () const;

    std::string&
    front ();

    const std::string&
    front () const;

    std::string&
    back ();

    const std::string&
    back () const;

    std::string&
    operator[] (size_t);

    const std::string&
    operator[] (size_t) const;

  public:
    bool
    empty () const;

    size_t
    size () const;

    size_t
    capacity () const;

    size_t
    max_size () const;

  public:
    void
    clear ();

    void
    pop_back ();

    iterator
    erase (iterator);

    void
    push_back (const std::string&);

    iterator
    insert (iterator, const std::string&);

    void
    reserve (size_t);
  };

  bool
  operator== (const string_sequence&, const string_sequence&);

  bool
  operator!= (const string_sequence&, const string_sequence&);
}
  

When STL is enabled and C++ exceptions are disabled, the signatures of the push_back(), insert(), and reserve() functions change as follows:

namespace xml_schema
{
  class string_sequence
  {
  public:
    enum error
    {
      error_none,
      error_no_memory
    };

    ...

  public:
    error
    push_back (const std::string&);

    error
    insert (iterator, const std::string&);

    error
    insert (iterator, const std::string&, iterator& result);

    error
    reserve (size_t);
  };
}
  

When STL is disabled and C++ exceptions are enabled, the string_sequence type has the following interface:

namespace xml_schema
{
  class string_sequence
  {
  public:
    typedef char*         value_type;
    typedef char**        pointer;
    typedef const char**  const_pointer;
    typedef char*         reference;
    typedef const char*   const_reference;

    typedef size_t        size_type;
    typedef ptrdiff_t     difference_type;

    typedef char** iterator;
    typedef const char* const* const_iterator;

    string_sequence ();

    void
    swap (string_sequence&);

  private:
    string_sequence (string_sequence&);

    string_sequence&
    operator= (string_sequence&);

  public:
    iterator
    begin ();

    const_iterator
    begin () const;

    iterator
    end ();

    const_iterator
    end () const;

    char*
    front ();

    const char*
    front () const;

    char*
    back ();

    const char*
    back () const;

    char*
    operator[] (size_t);

    const char*
    operator[] (size_t) const;

  public:
    bool
    empty () const;

    size_t
    size () const;

    size_t
    capacity () const;

    size_t
    max_size () const;

  public:
    void
    clear ();

    void
    pop_back ();

    iterator
    erase (iterator);

    void
    push_back (char*);

    void
    push_back_copy (const char*);

    iterator
    insert (iterator, char*);

    void
    reserve (size_t);

    // Detach a string from the sequence at a given position.
    // The string pointer at this position in the sequence is
    // set to 0.
    //
    char*
    detach (iterator);
  };

  bool
  operator== (const string_sequence&, const string_sequence&);

  bool
  operator!= (const string_sequence&, const string_sequence&);
}
  

The push_back() and insert() functions assume ownership of the passed string which should be allocated with operator new char[] and will be deallocated with operator delete[] by the string_sequence object. These two functions free the passed object if the reallocation of the underlying sequence buffer fails. The push_back_copy() function makes a copy of the passed string. If you detach the underlying element string, then it should eventually be deallocated with operator delete[].

When both STL and C++ exceptions are disabled, the signatures of the push_back(), push_back_copy(), insert(), and reserve() functions change as follows:

namespace xml_schema
{
  class string_sequence
  {
  public:
    enum error
    {
      error_none,
      error_no_memory
    };

    ...

  public:
    error
    push_back (char*);

    error
    push_back_copy (const char*);

    error
    insert (iterator, char*);

    error
    insert (iterator, char*, iterator& result);

    error
    reserve (size_t);
  };
}
  

6.3 base64Binary and hexBinary Parsers

The return type of the base64_binary_pimpl and hex_binary_pimpl parser implementations is xml_schema::buffer*. The returned object is dynamically allocated with operator new and should eventually be deallocated with operator delete. With C++ exceptions enabled (Section 5.3, "C++ Exceptions"), the buffer type has the following interface:

namespace xml_schema
{
  class buffer
  {
  public:
    class bounds {}; // Out of bounds exception.

  public:
    buffer ();

    explicit
    buffer (size_t size);
    buffer (size_t size, size_t capacity);
    buffer (const void* data, size_t size);
    buffer (const void* data, size_t size, size_t capacity);

    enum ownership_value { assume_ownership };

    // This constructor assumes ownership of the memory passed.
    //
    buffer (void* data, size_t size, size_t capacity, ownership_value);

  private:
    buffer (const buffer&);

    buffer&
    operator= (const buffer&);

  public:
    void
    attach (void* data, size_t size, size_t capacity);

    void*
    detach ();

    void
    swap (buffer&);

  public:
    size_t
    capacity () const;

    bool
    capacity (size_t);

  public:
    size_t
    size () const;

    bool
    size (size_t);

  public:
    const char*
    data () const;

    char*
    data ();

    const char*
    begin () const;

    char*
    begin ();

    const char*
    end () const;

    char*
    end ();
  };

  bool
  operator== (const buffer&, const buffer&);

  bool
  operator!= (const buffer&, const buffer&);
}
  

The last constructor and the attach() member function make the buffer instance assume the ownership of the memory block pointed to by the data argument and eventually release it by calling operator delete(). The detach() member function detaches and returns the underlying memory block which should eventually be released by calling operator delete().

The capacity() and size() modifier functions return true if the underlying buffer has moved. The bounds exception is thrown if the constructor or attach() member function arguments violate the (size <= capacity) constraint.

If C++ exceptions are disabled, the buffer type has the following interface:

namespace xml_schema
{
  class buffer
  {
  public:
    enum error
    {
      error_none,
      error_bounds,
      error_no_memory
    };

    buffer ();

  private:
    buffer (const buffer&);

    buffer&
    operator= (const buffer&);

  public:
    error
    attach (void* data, size_t size, size_t capacity);

    void*
    detach ();

    void
    swap (buffer&);

  public:
    size_t
    capacity () const;

    error
    capacity (size_t);

    error
    capacity (size_t, bool& moved);

  public:
    size_t
    size () const;

    error
    size (size_t);

    error
    size (size_t, bool& moved);

  public:
    const char*
    data () const;

    char*
    data ();

    const char*
    begin () const;

    char*
    begin ();

    const char*
    end () const;

    char*
    end ();
  };

  bool
  operator== (const buffer&, const buffer&);

  bool
  operator!= (const buffer&, const buffer&);
}
  

6.4 Time Zone Representation

The date, dateTime, gDay, gMonth, gMonthDay, gYear, gYearMonth, and time XML Schema built-in types all include an optional time zone component. The following xml_schema::time_zone base class is used to represent this information:

namespace xml_schema
{
  class time_zone
  {
  public:
    time_zone ();
    time_zone (short hours, short minutes);

    bool
    zone_present () const;

    void
    zone_reset ();

    short
    zone_hours () const;

    void
    zone_hours (short);

    short
    zone_minutes () const;

    void
    zone_minutes (short);
  };

  bool
  operator== (const time_zone&, const time_zone&);

  bool
  operator!= (const time_zone&, const time_zone&);
}
  

The zone_present() accessor function returns true if the time zone is specified. The zone_reset() modifier function resets the time zone object to the not specified state. If the time zone offset is negative then both hours and minutes components are represented as negative integers.

6.5 date Parser

The return type of the date_pimpl parser implementation is xml_schema::date which represents a year, a day, and a month with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class date: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    date ();

    date (int year, unsigned short month, unsigned short day);

    date (int year, unsigned short month, unsigned short day,
          short zone_hours, short zone_minutes);

    int
    year () const;

    void
    year (int);

    unsigned short
    month () const;

    void
    month (unsigned short);

    unsigned short
    day () const;

    void
    day (unsigned short);
  };

  bool
  operator== (const date&, const date&);

  bool
  operator!= (const date&, const date&);
}
  

6.6 dateTime Parser

The return type of the date_time_pimpl parser implementation is xml_schema::date_time which represents a year, a month, a day, hours, minutes, and seconds with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class date_time: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    date_time ();

    date_time (int year, unsigned short month, unsigned short day,
               unsigned short hours, unsigned short minutes,
               double seconds);

    date_time (int year, unsigned short month, unsigned short day,
               unsigned short hours, unsigned short minutes,
               double seconds, short zone_hours, short zone_minutes);

    int
    year () const;

    void
    year (int);

    unsigned short
    month () const;

    void
    month (unsigned short);

    unsigned short
    day () const;

    void
    day (unsigned short);

    unsigned short
    hours () const;

    void
    hours (unsigned short);

    unsigned short
    minutes () const;

    void
    minutes (unsigned short);

    double
    seconds () const;

    void
    seconds (double);
  };

  bool
  operator== (const date_time&, const date_time&);

  bool
  operator!= (const date_time&, const date_time&);
}
  

6.7 duration Parser

The return type of the duration_pimpl parser implementation is xml_schema::duration which represents a potentially negative duration in the form of years, months, days, hours, minutes, and seconds. Its interface is presented below.

namespace xml_schema
{
  class duration
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    duration ();

    duration (bool negative,
              unsigned int years, unsigned int months, unsigned int days,
              unsigned int hours, unsigned int minutes, double seconds);

    bool
    negative () const;

    void
    negative (bool);

    unsigned int
    years () const;

    void
    years (unsigned int);

    unsigned int
    months () const;

    void
    months (unsigned int);

    unsigned int
    days () const;

    void
    days (unsigned int);

    unsigned int
    hours () const;

    void
    hours (unsigned int);

    unsigned int
    minutes () const;

    void
    minutes (unsigned int);

    double
    seconds () const;

    void
    seconds (double);
  };

  bool
  operator== (const duration&, const duration&);

  bool
  operator!= (const duration&, const duration&);
}
  

6.8 gDay Parser

The return type of the gday_pimpl parser implementation is xml_schema::gday which represents a day of the month with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class gday: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    gday ();

    explicit
    gday (unsigned short day);

    gday (unsigned short day, short zone_hours, short zone_minutes);

    unsigned short
    day () const;

    void
    day (unsigned short);
  };

  bool
  operator== (const gday&, const gday&);

  bool
  operator!= (const gday&, const gday&);
}
  

6.9 gMonth Parser

The return type of the gmonth_pimpl parser implementation is xml_schema::gmonth which represents a month of the year with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class gmonth: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    gmonth ();

    explicit
    gmonth (unsigned short month);

    gmonth (unsigned short month,
            short zone_hours, short zone_minutes);

    unsigned short
    month () const;

    void
    month (unsigned short);
  };

  bool
  operator== (const gmonth&, const gmonth&);

  bool
  operator!= (const gmonth&, const gmonth&);
}
  

6.10 gMonthDay Parser

The return type of the gmonth_day_pimpl parser implementation is xml_schema::gmonth_day which represents a day and a month of the year with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class gmonth_day: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    gmonth_day ();

    gmonth_day (unsigned short month, unsigned short day);

    gmonth_day (unsigned short month, unsigned short day,
                short zone_hours, short zone_minutes);

    unsigned short
    month () const;

    void
    month (unsigned short);

    unsigned short
    day () const;

    void
    day (unsigned short);
  };

  bool
  operator== (const gmonth_day&, const gmonth_day&);

  bool
  operator!= (const gmonth_day&, const gmonth_day&);
}
  

6.11 gYear Parser

The return type of the gyear_pimpl parser implementation is xml_schema::gyear which represents a year with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class gyear: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    gyear ();

    explicit
    gyear (int year);

    gyear (int year, short zone_hours, short zone_minutes);

    int
    year () const;

    void
    year (int);
  };

  bool
  operator== (const gyear&, const gyear&);

  bool
  operator!= (const gyear&, const gyear&);
}
  

6.12 gYearMonth Parser

The return type of the gyear_month_pimpl parser implementation is xml_schema::gyear_month which represents a year and a month with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class gyear_month: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    gyear_month ();

    gyear_month (int year, unsigned short month);

    gyear_month (int year, unsigned short month,
                 short zone_hours, short zone_minutes);

    int
    year () const;

    void
    year (int);

    unsigned short
    month () const;

    void
    month (unsigned short);
  };

  bool
  operator== (const gyear_month&, const gyear_month&);

  bool
  operator!= (const gyear_month&, const gyear_month&);
}
  

6.13 time Parser

The return type of the time_pimpl parser implementation is xml_schema::time which represents hours, minutes, and seconds with an optional time zone. Its interface is presented below. For more information on the base xml_schema::time_zone class refer to Section 6.4, "Time Zone Representation".

namespace xml_schema
{
  class time: public time_zone
  {
  public:
    // The default constructor creates an uninitialized object.
    // Use modifiers to initialize it.
    //
    time ();

    time (unsigned short hours, unsigned short minutes, double seconds);

    time (unsigned short hours, unsigned short minutes, double seconds,
          short zone_hours, short zone_minutes);

    unsigned short
    hours () const;

    void
    hours (unsigned short);

    unsigned short
    minutes () const;

    void
    minutes (unsigned short);

    double
    seconds () const;

    void
    seconds (double);
  };

  bool
  operator== (const time&, const time&);

  bool
  operator!= (const time&, const time&);
}
  

7 Document Parser and Error Handling

In this chapter we will discuss the xml_schema::document_pimpl type, the error handling mechanisms provided by the mapping, as well as how to reuse a parser after an error has occurred.

There are four categories of errors that can result from running a parser on an XML instance: system, xml, schema, and application. The system category contains memory allocation and file/stream operation errors. The xml category is for XML parsing and well-formedness checking errors. Similarly, the schema category is for XML Schema validation errors. Finally, the application category is for application logic errors that you may want to propagate from parser implementations to the caller of the parser.

The C++/Parser mapping supports two methods of reporting errors: using C++ exceptions and with error codes. The method used depends on whether or not you have configured the XSD/e runtime and the generated code with C++ exceptions enabled, as described in Section 5.3, "C++ Exceptions".

7.1 Document Parser

The xml_schema::document_pimpl parser is a root parser for the vocabulary. As mentioned in Section 3.4, "Connecting the Parsers Together", its interface varies depending on the mapping configuration (Chapter 5, "Mapping Configuration"). When STL and the iostream library are enabled, the xml_schema::document_pimpl class has the following interface:

namespace xml_schema
{
  class parser_base;

  class document_pimpl
  {
  public:
    document_pimpl (parser_base&,
                    const char* root_element_name);

    document_pimpl (parser_base&,
                    const char* root_element_namespace,
                    const char* root_element_name);

    document_pimpl (parser_base&,
                    const std::string& root_element_name);

    document_pimpl (parser_base&,
                    const std::string& root_element_namespace,
                    const std::string& root_element_name);


  public:
    // Parse a local file. The file is accessed with std::ifstream
    // in binary mode. The std::ios_base::failure exception is used
    // to report io errors (badbit and failbit) if exceptions are
    // enabled. Otherwise error codes are used.
    //
    void
    parse (const char* file);

    void
    parse (const std::string& file);

    // Parse std::istream. std::ios_base::failure exception is used
    // to report io errors (badbit and failbit) if exceptions are
    // enabled. Otherwise error codes are used.
    //
    void
    parse (std::istream&);

    // Parse a chunk of input. You can call this function multiple
    // times with the last call having the last argument true.
    //
    void
    parse (const void* data, size_t size, bool last);

    // Low-level Expat-specific parsing API.
    //
    void
    parse_begin (XML_Parser);

    void
    parse_end ();
  };
}
  

When the use of STL is disabled, the constructors and the parse() function that use std::string in their signatures are not available. When the use of iostream is disabled, the parse() functions that parse a local file and std::istream are not available.

When support for XML Schema polymorphism is enabled, the overloaded document_pimpl constructors have additional arguments which control polymorphic parsing. For more information refer to Section 5.7, "Support for Polymorphism".

The first argument to all overloaded constructors is the parser for the type of the root element. The parser_base class is the base type for all parser skeletons. The second and third arguments to the document_pimpl's constructors are the root element's name and namespace.

The parse_begin() and parse_end() functions present a low-level, Expat-specific parsing API for maximum control. A typical use case would look like this (pseudo-code):

xxx_pimpl root_p;
document_pimpl doc_p (root_p, "root");

root_p.pre ();
doc_p.parse_begin (xml_parser);

while (more_stuff_to_parse)
{
   // Call XML_Parse or XML_ParseBuffer:
   //
   if (XML_Parse (...) != XML_STATUS_ERROR)
     break;
}

doc_p.parse_end ();
result_type result (root_p.post_xxx ());
  

Note that if your vocabulary use XML namespaces, the XML_ParserCreateNS() functions should be used to create the XML parser. Space (XML_Char (' ')) should be used as a separator (the second argument to XML_ParserCreateNS()). Furthermore, if XML_Parse or XML_ParseBuffer fail, call parse_end() to determine the error which is indicated either via exception or set as an error code.

The error handling mechanisms employed by the document_pimpl parser are described in Section 7.2, "Exceptions" and Section 7.3, "Error Codes".

7.2 Exceptions

When C++ exceptions are used for error reporting, the system errors are mapped to the standard exceptions. The out of memory condition is indicated by throwing an instance of std::bad_alloc. The stream operation errors are reported by throwing an instance of std::ios_base::failure.

The xml and schema errors are reported by throwing the xml_schema::parser_xml and xml_schema::parser_schema exceptions, respectively. These two exceptions derive from xml_schema::parser_exception which, in turn, derives from std::exception. As a result, you can handle any error from these two categories by either catching std::exception, xml_schema::parser_exception, or individual exceptions. The further down the hierarchy you go the more detailed error information is available to you. The following listing shows the definitions of these exceptions:

namespace xml_schema
{
  class parser_exception: public std::exception
  {
  public:
    unsigned long
    line () const;

    unsigned long
    column () const;

    virtual const char*
    text () const = 0;

    ...
  };

  std::ostream&
  operator<< (std::ostream&, const parser_exception&);


  typedef <implementation-details> parser_xml_error;

  class parser_xml: public parser_exception
  {
  public:
    parser_xml_error
    code () const;

    virtual const char*
    text () const;

    virtual const char*
    what () const throw ();

    ...
  };


  typedef <implementation-details> parser_schema_error;

  class parser_schema: public parser_exception
  {
  public:
    parser_schema_error
    code () const;

    virtual const char*
    text () const;

    virtual const char*
    what () const throw ();

    ...
  };
}
  

The parser_xml_error and parser_schema_error are implementation-specific error code types. The operator<< defined for the parser_exception class simply prints the error description as returned by the text() function. The following example shows how we can catch these exceptions:

int
main (int argc, char* argv[])
{
  try
  {
    // Parse argv[1].
  }
  catch (const xml_schema::parser_exception& e)
  {
    cout << argv[1] << ":" << e.line () << ":" << e.column ()
         << ": error: " << e.text () << endl;
    return 1;
  }
}
  

Finally, for reporting application errors from parsing callbacks, you can throw any exceptions of your choice. They are propagated to the caller of the parser without any alterations.

7.3 Error Codes

When C++ exceptions are not available, error codes are used to report error conditions. Each parser skeleton and the root document_pimpl parser have the following member function for querying the error status:

xml_schema::parser_error
_error () const;
  

To handle all possible error conditions, you will need to obtain the error status after calls to: the document_pimpl's constructor (it performs memory allocations which may fail), the root parser pre() callback, each call to the parse() function, and, finally, the call to the root parser post_*() callback. The definition of xml_schema::parser_error class is presented below:

namespace xml_schema
{
  class sys_error
  {
  public:
    enum value
    {
      none,
      no_memory,
      open_failed,
      read_failed,
      write_failed
    };

    sys_error (value);

    operator value () const;

    static const char*
    text (value);

    ...
  };

  typedef <implementation-details> parser_xml_error;
  typedef <implementation-details> parser_schema_error;

  class parser_error
  {
  public:
    enum error_type
    {
      none,
      sys,
      xml,
      schema,
      app
    };

    error_type
    type () const;

    // Line and column are only available for xml, schema, and
    // app errors.
    //
    unsigned long
    line () const;

    unsigned long
    column () const;

    // Returns true if there is an error so that you can write
    // if (p.error ()) or if (error e = p.error ()).
    //
    typedef void (error::*bool_convertible) ();
    operator bool_convertible () const;

    // system
    //
    sys_error
    sys_code () const;

    const char*
    sys_text () const;

    // xml
    //
    parser_xml_error
    xml_code () const;

    const char*
    xml_text () const;

    // schema
    //
    parser_schema_error
    schema_code () const;

    const char*
    schema_text () const;

    // app
    //
    int
    app_code () const;

    ...
  };
}
  

The parser_xml_error and parser_schema_error are implementation-specific error code types. The parser_error class incorporates four categories of errors which you can query by calling the type() function. The following example shows how to handle error conditions with error codes. It is based on the person record example presented in Chapter 3, "Parser Skeletons".

int
main (int argc, char* argv[])
{
  // Construct the parser.
  //
  xml_schema::short_pimpl short_p;
  xml_schema::string_pimpl string_p;

  gender_pimpl gender_p;
  person_pimpl person_p;
  people_pimpl people_p;

  person_p.parsers (string_p, string_p, gender_p, short_p);
  people_p.parsers (person_p);

  // Parse.
  //
  using xml_schema::parser_error;
  parser_error e;

  do
  {
    xml_schema::document_pimpl doc_p (people_p, "people");
    if (e = doc_p._error ())
      break;

    people_p.pre ();
    if (e = people_p._error ())
      break;

    doc_p.parse (argv[1]);
    if (e = doc_p._error ())
      break;

    people_p.post_people ();
    e = people_p._error ();

  } while (false);

  // Handle errors.
  //
  if (e)
  {
    switch (e.type ())
    {
    case parser_error::sys:
      {
        cerr << argv[1] << ": error: " << e.sys_text () << endl;
        break;
      }
    case parser_error::xml:
      {
        cerr << argv[1] << ":" << e.line () << ":" << e.column ()
             << ": error: " << e.xml_text () << endl;
        break;
      }
    case parser_error::schema:
      {
        cerr << argv[1] << ":" << e.line () << ":" << e.column ()
             << ": error: " << e.schema_text () << endl;
        break;
      }
    case parser_error::app:
      {
        cerr << argv[1] << ":" << e.line () << ":" << e.column ()
             << ": application error " << e.app_code () << endl;
        break;
      }
    }
    return 1;
  }
}
  

The error type for application errors is int with the value 0 indicated the absence of error. You can set the application error by calling the _app_error() function inside a parser callback. For example, if it was invalid to have a person younger than 18 in our people catalog, then we could have implemented this check as follows:

class person_pimpl: public person_pskel
{
public:
  virtual void
  age (short a)
  {
    if (a < 18)
      _app_error (1);
  }

  ...
};
  

You can also set a system error by calling the _sys_error() function inside a parser callback. This function has one argument of type xml_schema::sys_error which was presented above. For example:

class person_pimpl: public person_pskel
{
public:
  virtual void
  pre ()
  {
    p_ = new person ();

    if (p_ == 0)
      _sys_error (xml_schema::sys_error::no_memory);
  }

  ...

private:
  person* p_;
};
  

7.4 Reusing Parsers after an Error

After a successful execution a parser returns into the initial state and can be used to parse another document without any extra actions. On the other hand, if an error occurred during parsing and you would like to reuse the parser to parse another document, you need to explicitly reset it into the initial state as shown in the following code fragment:

int
main ()
{
  ...

  std::vector<std::string> files = ...

  xml_schema::document_pimpl doc_p (people_p, "people");

  for (size_t i = 0; i < files.size (); ++i)
  {
    try
    {
      people_p.pre ();
      doc_p.parse (files[i]);
      people_p.post_people ();
    }
    catch (const xml_schema::parser_exception&)
    {
      doc_p.reset ();
    }
  }
}
  

If you do not need to reuse parsers after an error for example because your application terminates or you create a new parser instance in such situations, then you can avoid generating parser reset code by specifying the --suppress-reset XSD/e compiler option.

Your individual parser implementations may also require extra actions in order to bring them into a usable state after an error. To accomplish this you can override the _reset() virtual function as shown below. Notice that when you override the _reset() function in your implementation, you should always call the base skeleton version to allow it to reset its state:

class person_pimpl: public person_pskel
{
public:
  virtual void
  pre ()
  {
    p_ = new person ();
  }

  virtual void
  _reset ()
  {
    person_pskel::_reset ();
    delete p_;
    p_ = 0;
  }

  ...

private:
  person* p_;
};
  

Note also that the _reset() mechanism is used only when an error has occurred. To make sure that your parser implementations arrive at the initial state during successful execution, use the initialization (pre() and _pre()) and finalization (post_*() and _post()) callbacks.

Appendix A — Supported XML Schema Constructs

The Embedded C++/Parser mapping supports validation of the following W3C XML Schema constructs in the generated code.

Construct	Notes
Structure
element
attribute
any
anyAttribute
all
sequence
choice
complex type, empty content
complex type, mixed content
complex type, simple content extension
complex type, simple content restriction
complex type, complex content extension
complex type, complex content restriction
list
Facets
length	String-based types.
minLength	String-based types.
maxLength	String-based types.
pattern	String-based types.
whiteSpace	String-based types.
enumeration	String-based types.
minExclusive	Integer and floating-point types.
minInclusive	Integer and floating-point types.
maxExclusive	Integer and floating-point types.
maxInclusive	Integer and floating-point types.
Datatypes
byte
unsignedByte
short
unsignedShort
int
unsignedInt
long
unsignedLong
integer
nonPositiveInteger
nonNegativeInteger
positiveInteger
negativeInteger
boolean
float
double
decimal
string
normalizedString
token
Name
NMTOKEN
NCName
language
anyURI
ID	Identity constraint is not enforced.
IDREF	Identity constraint is not enforced.
NMTOKENS
IDREFS	Identity constraint is not enforced.
QName
base64Binary
hexBinary
date
dateTime
duration
gDay
gMonth
gMonthDay
gYear
gYearMonth
time