| Re: passing multiple keywords to subroutines [message #14078] |
Mon, 25 January 1999 00:00 |
Liam Gumley
Messages: 473
Registered: November 1994
|
Senior Member |
|
|
David Ritscher wrote:
> It's often desirable to redefine a function, and then have this new
> function behave almost like the original function. To cook up an
> example, let's say I need a new plot function, that plots the x-axis
> as .001 units:
>
> pro myplot, x
> plot, findgen(n_elements(x) * 0.001), x
> return
> end
>
> It would now be great to be able to pass any and all possible keywords
> into this new plotting function, so I might call it as:
> myplot, x, title='test', linestyle=3
>
> In IDL, this can be done by including all possible keywords (rather
> tedious!)
It can be done this way, and it is tedious indeed.
Alternatively, you can exploit the keyword inheritance features offered
by the _EXTRA keyword, e.g.
pro myplot, x, _extra = extra_keywords
plot, findgen(n_elements(x) * 0.001), x, _extra = extra_keywords
end
Now all keywords which go unrecognized by your routine MYPLOT are passed
into (but not out of) the PLOT procedure, so that
myplot, x, title='test', linestyle=3
will work the way you intended. In the online help, click the Contents
button, select 'Building IDL Applications', 'Defining Procedures and
Programs', 'Keyword Inheritance'.
Cheers,
Liam.
---
Liam E. Gumley
Space Science and Engineering Center, UW-Madison
1225 W. Dayton St., Madison WI 53706, USA
Phone (608) 265-5358, Fax (608) 262-5974
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
| Re: passing multiple keywords to subroutines [message #14079 is a reply to message #14078] |
Mon, 25 January 1999 00:00  |
J.D. Smith
Messages: 214
Registered: August 1996
|
Senior Member |
|
|
David Ritscher wrote:
>
> It's often desirable to redefine a function, and then have this new
> function behave almost like the original function. To cook up an
> example, let's say I need a new plot function, that plots the x-axis
> as .001 units:
>
> pro myplot, x
> plot, findgen(n_elements(x) * 0.001), x
> return
> end
>
> It would now be great to be able to pass any and all possible keywords
> into this new plotting function, so I might call it as:
> myplot, x, title='test', linestyle=3
>
> In IDL, this can be done by including all possible keywords (rather
> tedious!) In PV-Wave, it's almost impossible, since it is not
> possible to pass in an undefined keyword:
>
> (I have not defined the variable 'color' here:)
> WAVE> plot, [1,5], color=color
> % PLOT: Variable is undefined: COLOR.
> % Execution halted at $MAIN$ (PLOT).
>
> This same command works fine in IDL, and has the appropriate behavior
> (Here, with 'color' undefined, it behaves exactly the same as if the
> keyword was not specified).
>
> Thus in IDL it is possible to define my new function along these lines:
> pro myplot, x, color=color, linestyle=linestyle, ynozero=ynozero, $
> noclip=noclip, yrange=yrange,
> ...
> As I say, this becomes rather tedious.
>
> Does anyone see a simpler way to handle this sort of thing? It seems
> to me that that two things are needed:
> 1. both companies need to add a mechanism to handle keywords that will
> be passed on to subroutines, which might take on a form like the
> following:
> pro pass_keywords=pass_keywords myplot, x
> plot, x, pass_keywords=pass_keywords
>
> where I can now call 'myplot' with extra keywords:
> myplot, x, linestyle=3
>
> 2. The bugs in PV-Wave have to be fixed, such that undefined keywords
> can be handled by every function.
>
> Am I missing a better way to handle these issues?
>
> David Ritscher
>
uhhhhh....
Take a look at IDL's _EXTRA and _REF_EXTRA keyword inheritance
mechanisms...
or am I missing something?
JD
--
J.D. Smith |*| WORK: (607) 255-5842
Cornell University Dept. of Astronomy |*| (607) 255-6263
304 Space Sciences Bldg. |*| FAX: (607) 255-5875
Ithaca, NY 14853 |*|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
