Re: Some wishlist feature suggestions

[Bill Allombert <allomber@math.u-bordeaux.fr>]

On Tue, Dec 14, 2004 at 10:17:21AM -0800, Phil Carmody wrote:
> > For the case at hand: probably it would be better to add an optional
> > parameter to znorder (a multiple of the order) than adding a whole
> > new function.
> 
> That would certainly work, the default being eulerphi of the 2nd 
> component to the Mod (presumably that's what order()'s current 
> "o=phi((GEN) x[1]);" represents). I assume that decomp() returns 

Yes, for some historical reason C functions an GP function name do
not match. You can use ctags/etags to map the two.

> the factorisation of its parameter as some kind of matrix. 
> In which case, it should be a simple alteration to order(), once
> I work out how optional parameters work. The only optional parameters 
> I can find are flags, not GENs; I don't know if that matters.

The convention for optional parameters:
NULL for GEN, prototype code DG
0 for long,   protoype code D0,L,

So you will need to change the Prototype field in 
src/functions/number_theoretical/znorder from G
to GDG.

Cheers,
Bill.