Project pxml2ods

Converts an XML document with a simplified schema into an ODS spreadsheet document.
It will work in concert with other projects to convert scraped web pages into an
easily-imported spreadsheet.

Project Purpose

This project exists to serve as a prototype for a C-language version with the same
basic utility. It will provide

A testing place for XSL transformation of the data document
A testing place for building an ODS document
A place to discuss design decisions that will be carried over to the C version
A repository for online references that help develop the application

Design Elements

ODS File Format

This file format is much more complicated than a CSV (comma-separated-values)
document that might also serve the same purpose, but its advantage is that the
separation between rows and columns is unambiguous. Unlike CSV files, an ODS
file can define non-string data-types for more accurate and usable data transfer.

Advantages of the ODS file format

The ODS file format is directly-readable (need not be imported) by Microsoft Excel
as well as most other open-source spreadsheet programs.
It is much easier to produce that other well-supported spreadsheet formats. The
data is stored as an XML document, with other XML and text documents, in a zipped
file.

XML Source Format

The XML document that is the source for the ODS file must conform to some
simple rules. The rules are a greatly-simplified version of the XML document
created by the Schema Server component.

The document must include a schema element as a child of the document element,
and the table rows must be elements that are also direct children of the document
element.
The schema must include a row-name attribute, the value of which is the
name of the row elements.
The schema must contain a collection of field elements containing a name
attribute that corresponds to an attribute of the row elements.
The spreadsheet columns will consist of the field elements of the schema,
in document order.

<?xml version=1.0?>
<list>
   <schema row-name="row">
      <field name="id" ></field>
      <field name="fname" ></field>
      <field name="lname" ></field>
   </schema>
   <row id="1" fname="Tom" lname="Swift" ></row>
   <row id="2" fname="Phyllis" lname="Newton" ></row>
</list>

XSL Transformation

Using libxml2 to parse HTML pages makes even very-badly
formatted HTML usable as well-formatted XML. XSL is the ideal tool for converting
XML to other forms because it is a mature design with good conversion tools
available.

References

OpenDocument technical specification