The server will have a directory termed "the repository" that will contain all of the documents on that server. The files will be arranged as follows:

d) root_dir
|
+- f) doc1.xml
|
+- f) doc2.xml
|
+- f) doc_list.xml

d) here denotes a directory, and f) - a file.


  doc_list.xml will be a master document list that will contain a list
  of all documents and the names of the directories in which they
  reside. Each document will reside in a directory of its own. The
  doc_list.xml file will have the following structure:

<documents>
  <document id="0001" name="Introduction to NetPaint" />
  <document id="0002" name="Introduction to NetTxt" />
  <document id="0003" name="NetTxt: Advanced Topics" />
</documents>

  The ID will be the hash code of the name of the document and has to
  be unique (no two documents with the same name allowed).


  docNNNN.xml will be created for each document and will contain the
  document itself as well as the list of sections and their contents:

<document name="NetTxt Server File Structure" id="0001">
   <section id="0001" name="Introduction" editor="Vladimir" time="2004-02-15 13:39" position="1">
      The section stuff goes in here.
   </section>

   <section id="0002" name="Directory Structure" editor="Vladimir" time="2004-02-15 13:40" position="2">
      The section stuff goes in here.
   </section>
</document>


  The section tags will contain the text of each section. Each
  paragraph will be contained within the <p> tags which will specify
  the alignment of the paragraph. The <p> tags can contain <b>, <i>,
  and <u> tags to denote formatting, as well as the <font> tags which
  will allow one to choose font faces and <color> tags which will
  allow one to choose colors. The format of the file will be as
  follows:

<section>
  <p align="left">
    The server will have a directory termed <i>"The Repository"</i> that will contain all of the documents on that server. The files are arranged as follows:
  </p>
</section>


  This file structure will simplify the file reading/writing process
  quite a bit. Making a list of documents available from the server
  still wouldn't be a problem even, since we would have that list
  available form the master doc_list.
