دانلود کتاب Literate programming in XML

Abstract Literate programming is a programming and documentation methodology. Its central tenet is that documentation is more important than source code and should be the focus of a programmer’s activity. Literate programming facilitates this approach by combining code and documentation into a single, unified source document. One interesting aspect of this combined form is that it is neither source code nor documentation. Instead, a literate programming system provides tools that allow a user to extract the source code or documentation automatically, but neither of these extracted forms is ever modified. Using a literate programming system offers some interesting benefits for many programming styles. Because the combined format is machine processed to produce the source code, the author is no longer required to maintain or write the code in the linear fashion that the computer ultimately expects. This is clearly advantageous for top-down and bottom-up design strategies. It may also have benefits for more modern programming methodologies, such as Extreme Programming. Typical literate programming systems are quite complex. They are built on top of some underlying documentation system (such as TeX) and described in terms of the macros and other documentation markup required to describe an xweb document. However, it quickly becomes apparent that XML can greatly simplify this situation. By stipulating that the documentation format include a few (namespaced) elements, it is possible to implement literate programming in XML on top of any format that the author chooses: DocBook, TEI, XHTML, you name it. In the past few years, the number of XML vocabularies has exploded. Where there used to be just a few, there are now hundreds. In addition, many of these new vocabularies have all sorts of sophisticated processing expectations: XSLT, W3C XML Schema, RELAX NG, Schematron, SAML, SVG, and MathML just to name a small handful. Luckily, Literate Programming with XML applies equally well to XML, so it is possible to apply a literate programming methodology to the development of XML vocabularies. This paper describes the design and implementation of a literate programming system using XML and XSLT. The resulting system is equally capable of authoring systems in traditional programming languages and systems that are themselves built from XML. The paper includes several examples to demonstrate these features and pointers to real-world systems that are actively exploiting the power it offers.