MathML - aware article conversion from L A TEX . A comparison study

Publishing in Mathematics and theoretical areas in Computer Science and Physics has been predominantly using TEX/LTEX as a formatting language in the last two decades. This large corpus of borndigital material is both a boon — LTEX is semi-semantic format where the source often contains indications of the author’s intentions — and a problem — TEX is Turing-complete and authors use this freedom to use thousands of styles and millions of user macros. Several tools have been developed to convert TEX/LTEX documents to XML-based — i.e. Web and DML-compatible formats. Different DML Projects use different tools, and the selection seems largely accidental. To put the choice of converters for DML projects onto a more solid footing and to encourage competition and feature convergence we survey the market. In this paper we investigate and compare five LTEX-to-XML transformers in three dimensions: a) ergonomic factors like documentation, ease of installation, b) coverage, and c) quality of the resulting documents (in particular the MathML parts).