| Re: How to traverse/inquire a class object structure in IDL? [message #17387] |
Wed, 13 October 1999 00:00  |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
Paul van Delst (paul.vandelst@ssec.wisc.edu) writes:
> First off, thanks to David and Pavel for their insights. Before I could check the
> newsgroup for replies, one of our younger go-getter science types came and told me
> something about object oriented programming that made good sense:
>
> The data should be an attribute of the object, not the object itself.
Yeah, those young bucks know about object programming. :-(
> Hmm. Anyway, he and I sat down for about 15 minutes and came up with the following
> class structure definition and cleanup method:
>
> PRO nasti__define
>
> ; -- Define the NAMED data structure attribute
> data = { data, $
> wavenumber : PTR_NEW(), $
> radiance : PTR_NEW(), $
> altitude : PTR_NEW(), $
> fov_angle : PTR_NEW(), $
> fov_index : PTR_NEW(), $
> latitude : PTR_NEW(), $
> longitude : PTR_NEW(), $
> aircraft_roll : PTR_NEW(), $
> aircraft_pitch : PTR_NEW(), $
> scan_line_index : PTR_NEW(), $
> date : PTR_NEW(), $
> time : PTR_NEW(), $
> decimal_time : PTR_NEW() }
>
> ; -- Create object CLASS structure
> nasti = { nasti, $
> data : data }
>
> END
>
> I like this becuase now I can add additional attributes whenever I want, e.g.
> global attribute data read from the netCDF data file containing instrument
> calibration information and/or processing software CVS/RCS info etc.
Not too bad, although I thought the first one was OK too.
Actually, it is the data and the methods that manipulate
the data that should be encapsulated in the object, so I
don't see that this new construction gains us much of anything,
except more structure de-referencing. :-)
But I *would* change the name of the structure from DATA
to something just a tad less generic. I see plenty of
trouble coming down the road with a name like DATA.
> The cleanup method is now:
>
> PRO nasti::cleanup
>
> PRINT, FORMAT = '( /5x, "Clean up..." )'
>
> ; -- Free up pointers
> n_data_fields = N_TAGS( self.data )
> FOR i = 0, n_data_fields - 1 DO $
> IF ( PTR_VALID( self.data.(i) ) ) THEN $
> PTR_FREE, self.data.(i)
>
> END
Have to admit that this is compact. :-)
> I wish I'd "discovered" objects earlier......all that code I wrote that *needs* the
> data to be encapsulated. Crikey.
You are on a slippery slope here. Once you fall
for objects almost *everything* looks like a perfect
opportunity to use one. :-)
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
