Re: IDL 5.2 array definition question.

[Craig Markwardt <craigmnet(at)cow.physics.wisc.edu>]

Paul van Delst <pvandelst@ncep.noaa.gov> writes:
> The documentation (that I read before posting the original message)
> actually specifies this except the number that they guarenteed across
> versions was only 25!. My (badly formed) question can be rephrased as:
> 
> why?
> 
> Like I alluded to previously, IDL 5.3 raised the limit to 65536 (or
> thereabouts) which is way larger than any array initialisation anyone
> would want to type by hand, but I would still like to know what design
> decisions the pre-5.3 behaviour was based on. That's all.

My guess is that this is a limitation of the IDL parser, the part of
IDL that converts tokenized text into the compiled form.  

Most modern parsers today are written with assistence from programs
like lex and yacc.  These parsers *can* have problems handling large
lists of items, but not if they are written properly.  In principle
they can handle lists of unbounded size.

I would further guess that the IDL parser is more home-brew than
yacc-descended, and as such has some hardwired limits for things like
array size.  They probably just bumped up those limits for IDL v5.3.

Craig

-- 
--------------------------------------------------------------------------
Craig B. Markwardt, Ph.D.         EMAIL:    craigmnet@cow.physics.wisc.edu
Astrophysics, IDL, Finance, Derivatives | Remove "net" for better response
--------------------------------------------------------------------------