Re: Improvement of the MathML editor

Hello Frederic,

As a background for your research, you may be interested in
http://unicode.org/notes/tn28/UTN28-PlainTextMath.pdf
(which is very explicitly not restricted to ASCII, but
also tries to linearize input). Or do you already know this?

Regards,    Martin.

At 23:47 09/03/07, Fr$B".(Bd$B".(Bric WANG wrote:
>Hi all,
>
>As I said yesterday, I've started to improve the way one can produce formulae using Amaya. I write this mail because I suppose users may want to know what features I plan to add and give suggestions and comments. I've two main ideas:
>
>1) Improving the way formulae are typed: extending the scope of "ASCII strings" that are automatically converted in MathML.
>2) Improving the MathML panel: adding new pre-defined construction and allowing user to customize the panel.
>
>1) Using LaTeX-like fragment has been requested by some teachers from the PALETTE project. Many people who are used to write math with LaTeX are likely to be interested in this option. I'm not a TeX user and I think that for complex formulae, it is better to use the Wysiwyg interface: things like "\sqrt{\frac{\alpha}{2} x}" (25 characters) can be written faster with "Ctrl m q, Ctrl m f, Ctr g a, arrow, 2, arrow, x" (7 commands). Moreover you immediatly get the result displayed and don't need to care about syntax. However, I believe a "ASCII syntax" is sometimes useful, especially when it is "natural". For instance everybody have already used x^2 or a_0 to indicate superscript/subscript. Another example is chemistry: for instance one would like "HCl + H2O <-> H3O+ + Cl$BchC(B to be immediatly splitted.
>
>Currently, I've study how a general string parser could be integrated in Amaya and experiment it in the special case of chemical molecules. For the moment, the only issue with this parser is that only ASCII encoding is accepted for the source language. How the different modes of edition would be selected? First I thought switching modes using shortcuts was a good idea, but now I would prefer insertion of "placeholder" with an (internal) attribute that indicates how the element will be parsed. If the parsing fails (error of syntax), then the element is transformed in <mtext/> and preserves its attribute, so that the user can fix the syntax and the string is parsed again.
>
>First, I'll continue my work on the "chemical mode" and then look for "units mode" (for physics, chemistry...) and "LaTeX-like mode" (for basic constructions).
>
>2) The current panel covers all the "Content MathML" notations but it would be useful to add pre-defined constructions for Physics/Chemistry. Following the idea of "SVG templates" that are stored in external .svg files, I think it would be possible to create "MathML templates" with external .mml files, so that adding new constructions would be easier. Additionnaly, people should be able to select their "favorite constructions" in the panel so that they are directly accessible. This is a long term plan.
>
>Fr$B%F%%(Bd$B%F%%(Bric Wang
>


#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst@it.aoyama.ac.jp

Received on Sunday, 8 March 2009 08:05:36 UTC