***********************************************************************
*  This file is about an extended description of the MathDox format.  *
***********************************************************************


As we all know, the MathDox is an XML based format and combines several different XML-based languages, i.e. DocBook, OpenMath
MONET, XForms, Jelly, XInclude, into one new format. However, for the structure of 
the MathDox document, as MathDox is used for Maths, there are many cases which is not specified by DocBook standard.

We will discuss these situations and propose an preliminary extension to the Docbook for MathDox in the following para.

For most cases, we just continue to use the format of mathbook in MathDox.

1. When a part of the document states a theorem or lemma or proposition or claim, .
Add a new tag <assertion/> with optional attributes of "role" and "id" to wrapper this part. And this assertion may has a title, so we use a child element <title>...</title> to represent it.

Demo:
<assertion role='theorem' id = '...'>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>

<assertion role='lemma' id = '...'>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>

<assertion role='corollary' id = '...'>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>

<assertion role='proposition' id = '...'>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>

<assertion role='claim' id = '...'>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>

<assertion>
  (option: <title>...</title>)
  ... the content of the assertion...
</assertion>


2. When a part of the document contains a list of examples or exercises, we add new tags of <examples> and <exercises/> with optional attributes of 
"id", "type", and "for" to wrapper this part. And a child element <title>...</title> is also optional.

Demo:
<examples id = "..." type = "..." for = "...">
  (option: <title>...</title>)
  ... a list of examples...
</examples>

<examples>
  (option: <title>...</title>)
  ... a list of examples...
</examples>

<exercises id = "..." type = "..." for = "...">
 (option: <title>...</title>)
  ... a list of exercises...
</exercises>

<exercises>
 (option: <title>...</title>)
  ... a list of exercises...
</exercises>


3. When a part of the document states an example or exercise, we add a new tag of <example/> or <exercise/> with an necessary attribute of "id" and optional attributes of 
"type" and "for" to wrapper this part. And a child element <title>...</title> is also optional.
 
Demo:
<example type = "static" id = "..." for = "...">
  (option: <title>...</title>)
  ... the content of example...
</example>

<example id = "...">
  (option: <title>...</title>)
  ... the content of example...
</example>

<exercise type = "static" id = "..." for = "...">
  (option: <title>...</title>)
  ... the content of exercise...
</exercise>

<exercise id = "...">
  (option: <title>...</title>)
  ... the content of exercise...
</exercise>


4. When a part of the document states a definition, we add a new tag of <definition/> with an attribute of "id" to wrapper this part.
And a child element <title>...</title> is also optional.

Demo:
<definition id = "...">
 (option: <title>...</title>)
  ... the content of definition...
</definition>


5. When a part of the document states an algorithm, we add a new tag of <algorithm/> with an attribute of "id" to wrapper this part.
And a child element <title>...</title> is also optional.

Demo:
<algorithm id = "...">
 (option: <title>...</title>)
  ... the content of algorithm...
</algorithm>


6. When a part of the document states a remark, we add a new tag of <remark/> with an attribute of "id" to wrapper this part.
And a child element <title>...</title> is also optional.

Demo:
<remark id = "...">
  (option: <title>...</title>)
  ... the content of remark...
</remark>

7. When a part of the document states an answer, we add a new tag of <answer/> with optional attributes of "role and "id" to wrapper this part.
And a child element <title>...</title> is also optional.

Demo:
<answer role = "solution" id = "...">
  (option: <title>...</title>)
  ... the content of answer...
</answer>

<answer role = "hint" id = "...">
  (option: <title>...</title>)
  ... the content of answer...
</answer>

<answer>
  (option: <title>...</title>)
  ... the content of answer...
</answer>

8. When a part of the document states a proof, we add a new tag of <proof/> with optional attributes "id" and "for" to wrapper this part.
And a child element <title>Proof</title> is also optional.

Demo:
<proof id = "..." for = "...">
  <title>Proof</title>
  ... the content of proof...
</proof>

<proof>
... the content of proof...
</proof>


9. For the content of an algorithm, we use the structure as follows:

<algorithm id = "...">
  <itemizedlist>the content of the input and output</itemizedlist>
  <function> the definition of the procedure </function>
  <para> the content of the procedure </para>
</algorithm>