On 11/20/2013 8:59 AM, Aditya Mahajan wrote:
As far as ConTeXt is concerned, you can process the above XML quite easily. Come to think of it, it may be a useful to provide a module that maps HTML5 to PDF.
Aditya
I would vote for that approach. It is pretty much analogous to what I have decided to do. I'm doing my actual writing in HTML. creating CSS for ebooks and ConTeXt environment files for PDF. I'll probably hack the html2latex Perl script to do the mapping. Pandoc will /not/ meet my needs because Markdown does not distinguish between emphasized text and italic text and Pandoc compels all other input markup to behave like Markdown. Textile would be perfect but only outputs HTML well. The RedCloth implementation does output LaTeX but ignores CSS-style classes applied to paragraphs et. al. My needs are much simpler than the majority of people on this list. I have no need for math, no need for indexes, no need for bibliographies, footnotes or citations. BUT I want really top-notch visual output whether in PDF or printed on dead trees. The latter is usually the result of giving the printer process PDF files anyway so for the visuals I want, PDF is the common denominator and TeX (whether LaTeX or ConTeXt) is the most reasonable path, InDesign is not for poverty-stricken wretches like me. Unfortunately (for many reasons) the market for the stuff I write is hell-bent to replace printed books and even replace high-quality electronic presentation with formats in which the reader chooses almost everything, regardless of whether it compliments the text or not. EPUB (especially EPUB3) does appear to /want/ to provide author/designer-determined presentation but it can still be ignored or overridden by the reader. I don't want to even imagine the visual discordance of reading something like /Jane Eyre/ with double-spaced Comic Sans but if the e-reader allows it.... I'll crawl back under my rock, now. -- Bill Meahan, Westland, Michigan