On Friday, 22 August 2014 11:18:58 UTC+10, Jeffrey Kegler wrote: > > Good question, to which the fast answer is "no". >
Ah, well. I'm reluctant to write the C required to do this myself. I'm about the announce my almost-finished new parser of Graphviz DOT files on the [graphviz-interest] mailing list, and wanted to say that I develop in Perl but that everything can be translated into C. If no-one's (publicly) done that, I'll have to consider what to say/do. I confess I did not study the THIF's docs yet. > Someone capable of documenting how to turn SLIF into THIF might be better > off rewriting the upper layers in some other form, such as pure C or C++, > Python, better Perl, etc., etc. I'm doing that, in C and Lua, as the > When you say ' the upper layers', what exactly did you have in mind? My Perl or Marpa::R2? From your next comment, I assume the latter. Kollos project. (I'm overloaded, so Kollos won't happen quickly.) This > is following the principle that, when the code is not at all clean, you > rewrite first and THEN document. It would also be more fun for the > rewriter, because it would allow the rewriter to experiment with the > language/interface of their choice. (As to why the current XS/Perl > implementation is a mess, the answer is that it evolved incrementally in > response to the demand for new features, and in the face of a strict > requirement that every change had to preserve compatibility, even with > obsolete interfaces.) > I'm not complaining about the (existence) of the XS/Perl interface. It's fabulous you've made such a thing available, since it makes me incredibly productive. And I've never learned XS, so I'm now concerned I'd have to work in XS in order to work out to talk to the THIF. Yak-shaving, anyone? -- You received this message because you are subscribed to the Google Groups "marpa parser" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
