jbw wells, joe Dr. Joe Wells Joe Wells 2.05 3869 jbw jbw http://www.cee.hw.ac.uk/~jbw/ lachlan mackinnon, lachlan Dr. Lachlan MacKinnon Lachlan MacKinnon G.39 3410 lachlan lachlan http://www.cee.hw.ac.uk/~lachlan/ emacs-XML A Better Emacs XML-Editing Mode jbw Provide a way of editing XML inside Emacs which works on the tree structure but displays things in a semi-WYSIWYG manner

In many places, Emacs is the standard programmer's editor. Emacs is also frequently used for creating scientific papers. As more and more programming and document creation is done using XML, it would be good if Emacs had better support for editing XML.

One way of editing XML is to edit the raw XML code with an editor that works on the raw bytes. Emacs currently supports this method. So the person editing the XML sees something like this: 


 Some random text ...
]]>

Another approach is to use an editor that diagrams the XML as a tree, as in this screenshot from the Conglomerate editor:
[image of tree-based editor]
Emacs does not yet support this method. It would be nice if Emacs could work like this.

It would be even nicer if the editor could display the XML using styles for each element, like the way some HTML editors work. Emacs does not support this method.

  • Implement an XML parser within Emacs. This is fairly simple.
  • Implement an XML writer for Emacs. This is trivial.
  • Implement an XML style engine within Emacs to lay out an Emacs buffer according to some specified styles in some very simple style framework.
  • Implement an Emacs major mode for editing the displayed XML.
 Machine with recent Emacs installed The hardest part will be learning about the Emacs execution environment. Fortunately, Dr. Wells is an expert in this and will be able to provide substantial help. Things which will help: familiarity with XML, familiarity with LISP, familiarity with Emacs. XML-diff-tools Open-Source XML Comparing, Differencing, and Patching Tools jbw Provide open source tools for comparing XML files, computing the differences between XML files, and given the differences between two XML files, computing one of the files from the other and the differences There are well known UNIX tools for manipulating line-oriented files: cmp, diff, patch. For XML files, line-oriented tools produce suboptimal results for the tasks of comparing files, computing the differences between files, and applying differences between two files as a delta to one of the files to obtain the other. There exist tree-oriented tools for the tasks of comparing and computing differences, but they unfortunately are proprietary and thus are not part of the electronic commons. There does not seem to exist any tree-oriented tool for applying the differences between two XML files to obtain the other.
  • Define a format for representing differences between XML files.
  • Develop a tool for calculating differences between XML files.
  • Evaluate the quality of the calculated differences in practice.
  • Develop a tool for applying differences to obtain one file from another. [Optional, to be done last.]
 XML manipulating software It is easy to calculate differences. For example, the difference between XML elements A and B can be given by the pair (A,B). But this is expensive. One wants to include only the changed data. However, if one includes just the changed data, then the difference can not be used on a slightly changed A' to obtain the corresponding B'. The hard part is in getting good differences which have all of the desired qualities. XML experience. XML-HWU-CEE-abstracts-reports XML-Based System for Project/Dissertation Abstracts and Reports jbw lachlan Develop an XML-based system for the long-term archiving of B.Sc. dissertation abstracts and M.Sc. project abstracts and import previous years' abstracts from Microsoft Word format As part of the work towards various degrees, each student does a project. For years, the secretaries have been entering the project abstracts for each year/program in Microsoft Word documents. It would be better if the abstracts were archived in a more lasting format. The Microsoft Word format is undefined (except by the implementation), undocumented, and frequently changes in ways that are not backward compatible.
  • Determine requirements of the various parties involved (students, central university, external examiners, etc.).
  • Determine what data is needed to satisfy these requirements.
  • Develop XML encodings of the necessary data.
  • Make available convenient methods for the various parties to enter and view the data.
  • Import the legacy data from prior years.
  • Develop instructions telling staff how to switch over to new system if it is deemed acceptable.
 XML manipulating software. Software for decoding Microsoft Word files. Half of the difficulty here will be accurately determining the real requirements (i.e., not just what people say they want). The other work will mainly involve properly gluing together existing software packages. Things which will help: XML experience, experience decoding Microsoft Word files, experience in determining requirements. module-language-prototype XML-Based Prototype Implementation of Toy Module Language to Aid in Design Exploration jbw An XML-based prototype implementation of a language based on the m-calculus of Dr. Wells for manipulating and linking modules to aid in exploring the design of such a language.

There are many kinds of module systems in different programming languages. To attempt to explain the essence of modules with a few basic constructs (in fact, roughly 5 constructs), Dr. Wells has designed the m-calculus. The m-calculus can model many existing module systems as well as other programming constructs such as first-class functions and objects.

To help in research involving the m-calculus, it is proposed to implement a toy programming language directly based on the m-calculus. An implementation will also provide a useful platform for exploring improvements on the ideas of the m-calculus.

The motivation for this research comes from the fact that modules are an essential fact of programming technology. They serve many important purposes:

  • Modules are often a unit of separate compilation.
  • Modules can be a unit of software reuse.
  • Modules often represent a unit of functionality which can be comprehended at one time.
  • Modules are often a unit of configuration which is done by switching from one module to another one implementing the same interface.
  • Choose an XML-based abstract syntax for a language inspired by the m-calculus. This part is not hard and there will be lots of suggestions for the student.
  • Choose a concrete syntax (i.e., in the style of existing programming languages) for the language. This part is not hard and there will be lots of suggestions for the student.
  • Implement a pretty printer for the language.
  • If the student is up to the challenge, the project may also include implementing a parser for converting from the concrete syntax to the abstract syntax. But this is less urgent and can be delayed to the end or not done at all.
  • Implement transformations on abstract syntax trees which represent carrying out operations such as linking two modules, using a component of a module, using variables, etc. Provide some user interface to this. All of the needed transformations will be explained for the student.
 A computer The initial implementation should not be hard. The subsequent explorations of variations can involve many different levels of difficulty. The overall degree of difficulty will depend on the student's level of motivation to work. Things which will help: experience in symbolic programming, experience with a functional programming language.