On Thu, Mar 08, 2012 at 06:45:34PM +0100, Ralf Hemmecke wrote:
>>> Even more is true. If you connect to FriCAS and keep the connection
>>> open. Then you probably just have to transfer the data once per session.
>
>>> In other words, you basically talk from Haskell with the FriCAS interpreter.
>>>
>>
>> I do not understand this.
>> I think, it is faster for DoCon to "talk" with
>>
>> the compiled Spad function parseCall + applying this call +
>> read/write of Axiom for String and the two _named pipes_.
>
> Oh, maybe I don't yet understand your design. Or maybe I don't
> understand FriCAS, but I wouldn't know how to talk to fricas other than
> via the interpreter.
The Axiom interpreter evaluates the below loop of `repeat' :
------------------------------------------------ fifoFromA.input -------
toA : TextFile := open("toA", "input")
fromA : TextFile := open("fromA", "output")
repeat
iStr := readLineIfCan! toA
if iStr case "failed" then
(output ""; output "input -> failed"; break)
output "iStr = " iStr "\n"
strings := split(iStr, char ";") $String
strings' := map(parseEvalShow, strings) --
str := concat intersperse(";", strings')
resStr := concat[str, newline() :: String]
writeLine!(fromA, resStr)
flush(fromA)
------------------------------------------------------------------------
`repeat'
(1) inputs a string from the named pipe toA;
(2) splits to to several call inputs, each of type String;
(3) applies parseEvalShow str to each call input str.
(4) writes back to the pipe "fromA" the result strings strings'.
(several calls are transferred at a time, in order to possibly reduce
the number of calls for input/output with pipes).
Here parseEvalShow : String -> String
is programmed in Spad, compiled and loaded before running
fifoFromA.input.
The main Axiom part is in parseEvalShow:
parseCall, evaluate it by applying the head function,
output to String.
This all I mean by saying
"DoCon talks with the compiled Spad function parseEvalShow:
parseCall + applying this call + read/write of Axiom".
For example, for factoring -4x^3 +5x^2 +6 : UP(x, INT), it is input
"(factor [(P = UP x INT)] -- type input table
(UnivariateFactorize P) -- package call
[(Factorize P) P] -- signature of `factor'
[([3 2 0] [-4 5 6])] -- polynomial
)"
At the stage of `evaluate', it applies in this example
factor(<prasedPolynomial>) $UnivariateFactorize(<parsedDomain>)
And I hope that after
D := parseDomain(<needed input>) pretend SuchAndSuchRing
(which expresses the above P ) is computed, the call
factor(<prasedPolynomial>) $UnivariateFactorize( D )
will work.
Will it?
If it will, then this would mean `evaluate'.
and the first problem will be the performance of parseCall.
And the idea to replace the "parse - interpret" part of FriCAS with
my parseEvalShow.
Such is the situation.
Regards,
------
Sergei
[email protected]
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" 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/fricas-devel?hl=en.
