Re: Dealing with very large blocks of character data

Hi David,

This sort of thing is the reason IBM increased the size of strings in 6.1. If you want to stick with RPG 5.4, you're going to have to either use the C functions, or write your own functions equivalent to the RPG (or C) ones.

Or do the whole project in a different language. (Java, maybe?)


On 11/30/2011 10:51 AM, David Gibbs wrote:

Folks:

I've got a problem where I need to analyze and manipulate very large
blocks of character data.

This character data could be well in excess of 64k.

Since the limit for RPG character fields on 5.4 is 64k ... I can't
use normal BIF's.

Upgrading to 6.1 isn't an option due to customer requirements (this
was my first idea, but it got shot down).

My current thought is to use pointers and C functions to analyze the
text in the space.

%alloc to create space and get a pointer

strstr in place of %scan(s)
strlen in place of %len(%trim(s))
strcat& %realloc in place of s1 + s2
strncpy in place of %subst

dealloc when I'm done with the space.

Obviously I would wrapper the C functions in procedures so I don't
have to deal with the individual C api's directly.

I'm not supper happy with the idea of having to deal with this data
using pointers and C api's ... Any suggestions on a different
approach to handling these large character variables?

Thanks!

david