kwarc / deprecated-llamapun Goto Github PK

Similarly to the way smartphones auto-correct spelling, there is a common need in arXiv to auto-correct the separation between text and mathematics. In TeX, there is a clear separation of "text mode" and "math mode", which LaTeXML conserves in a HTML vs MathML separation, after conversion.

There is an existing implementation from a semester project of mine at:
https://github.com/KWARC/LLaMaPUn/blob/master/lib/LLaMaPUn/Preprocessor/Purify.pm

The original ticket, with detailed description of progress and various phenomena covered is at the old Trac:
https://trac.kwarc.info/lamapun/ticket/1

I should get this ported to C, in order to improve the quality of the input data for our linguistic experiments.

You currently have createDNM defined on entire LibXML documents:

dnmPtr createDNM(xmlDocPtr doc, long parameters)

However, I want to create separate DNMs for individual paragraphs, so that I can work with the plain text of those XML elements exclusively. It should be easy to change the bookkeeping to support xmlNodePtr as the argument of createDNM, I don't think you really need the document itself for anything important.

So the signature I would like to use is:

dnmPtr createDNM(xmlNodePtr node, long parameters)