Mathics somewhat translates Mathematica to SymPy. https://github.com/poeschko/mathics
http://www.mathics.org/ I tried to play with Mathics' code a while ago, unfortunately there is very little documentation. In any case, Mathics has a parser for Mathematica's code, it builds its own AST object mapping, and calls SymPy to perform the algorithmic evaluation. You could try to use it and put a debugging breakpoint on the SymPy calls, but that's still hard to do. Mathics has also a Mathematica compatible pattern matcher. Technically, I think that if we add assumptions-awareness and specificity sorting to the patterns containing wild cards, and default/optional wildcards, SymPy's patterns would become much more similar to those of Mathematica (except for excessive matches like inverse functions and powers). On Thursday, October 9, 2014 6:15:22 AM UTC+2, Richard Fateman wrote: > > There are at least 2 open source parsers for Mathematica code. > > The trivial stuff -- parsing x Sin[x] into x*sin(x) equivalent could > be > done by following directions in any intro to compilers book. > > The rest of the stuff, which requires pattern matching, simplification, and > a whole collection of specific commands all sitting inside the Mathematica > rule-based evaluation mechanism ... well you won't get that by translating > naively into python. > > It's like saying you can translate Lisp into C. Sure. But if you want to > execute it, you need a whole bunch of stuff, like a garbage collector, > arbitrary-precision integers, etc etc etc that you don't get by > translating > (+ a b) into a+b. > > > > On Wednesday, October 8, 2014 7:16:02 PM UTC-7, [email protected] wrote: >> >> you can try this expression for a test >> (-6x^2-x-7)(2x^3+3x^2-2x-5) >> >> 在 2012年5月2日星期三UTC+8上午12时19分18秒，Aaron Meurer写道： >>> >>> This probably is doable, but we would need an actual parser in SymPy, >>> which is the difficult part. If we had that, adding rules for >>> Mathematica functions would ideally not be hard. See >>> https://github.com/sympy/sympy/wiki/parsing for some ideas on parsing. >>> >>> More realistically, in >>> http://code.google.com/p/sympy/issues/detail?id=161, it is suggested >>> that for Maxima, Sage can be used. So I'm wondering if Sage, or maybe >>> some other project has a parser for Mathematica that can put it in a >>> form that SymPy can read, or at least on close. >>> http://code.google.com/p/sympy/issues/detail?id=2864 is also related >>> to this. >>> >>> Can you give an example of a Mathematica expression that you want to >>> parse? >>> >>> Aaron Meurer >>> >>> On Tue, May 1, 2012 at 9:00 AM, [email protected] >>> <[email protected]> wrote: >>> > I doubt that any automated translation will produce quality code. The >>> > style of Mathematica code is much more functional than the object >>> > oriented python style. What you get from automatic translation will >>> > not be human-readable (it will be python but very obfuscated). >>> > >>> > So I am very pessimistic. However if you have any success with this it >>> > will be great for sympy. >>> > >>> > -- >>> > You received this message because you are subscribed to the Google >>> Groups "sympy" group. >>> > To post to this group, send email to [email protected]. >>> > To unsubscribe from this group, send email to >>> [email protected]. >>> > For more options, visit this group at >>> http://groups.google.com/group/sympy?hl=en. >>> > >>> >> -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/84407fd0-1a16-4901-af10-1b1db8b4600b%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
