On 11/12/07, Bill Page wrote:
>
> After looking at the Axiom source code I see that 'unparse' is
> actually internally calling functions which convert an expression in
> Axiom form to Fortran form. The necessary code is found here:
>
> http://wiki.axiom-developer.org/axiom--test--1/src/algebra/MkfuncSpad
> http://wiki.axiom-developer.org/axiom--test--1/src/interp/NewfortBoot
> http://wiki.axiom-developer.org/axiom--test--1/src/interp/FormatBoot
>
> But it isn't really necessary for you to understand all that. Here is
> a distillation of the important parts into a fairly simple function
> that I think probably does what you want:
>
> (1) -> output1d(e)== _
> message( _
> concat map(string,destruct _
> fortranCleanUp( _
> exp2Fort1( _
> exp2FortOptimize( _
> outputTran(e::INFORM)$Lisp _
> )$Lisp _
> )$Lisp _
> )$Lisp _
> ) _
> )$OutputForm
> Type: Void
> (2) -> output1d(cos(x2) + 2*x1)
> Compiling function output1d with type Expression Integer ->
> OutputForm
> `-mcpu=' is deprecated. Use `-mtune=' or '-march=' instead.
>
> (2) cos(x2)+2*x1
> Type: OutputForm
> (3) ->
>
Oops, the version about fails for expressions like:
output1d(x2/x2)
because of some additional unnecessary Fortran processing by
the routine 'outputTran'. Please use this slightly simpler version:
(1) -> output1d(e)== _
message( _
concat map(string,destruct _
fortranCleanUp( _
exp2Fort1( _
exp2FortOptimize( _
e::INFORM _
)$Lisp _
)$Lisp _
)$Lisp _
) _
)$OutputForm
Type: Void
(2) -> output1d(cos(x2) + 2*x1)
Compiling function output1d with type Expression Integer ->
OutputForm
`-mcpu=' is deprecated. Use `-mtune=' or '-march=' instead.
(2) cos(x2)+2*x1
Type: OutputForm
(3) -> output1d(x2/x1)
Compiling function output1d with type Fraction Polynomial Integer
-> OutputForm
`-mcpu=' is deprecated. Use `-mtune=' or '-march=' instead.
(3) x2/x1
Type: OutputForm
(4) -> output1d(sqrt(x1))
(4) x1**(1/2)
Type: OutputForm
(5) ->
-------------
I have recorded this issue at:
http://axiom-wiki.newsynthesis.org/401UnparseGeneratesFortranFunctionNames
since I think that really this is due to a bug in 'unparse'. It really
should not generate those Fortran-type names if it's purpose is really
to unparse Axiom commands in such a manner that they can be re-parsed
by the Axiom interpreter.
At the above link I also proposed a more general solution that
corrects this problem for other parts of Axiom as well.
Regards,
Bill Page.
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
open-axiom-devel mailing list
open-axiom-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/open-axiom-devel
