Data tables for a software tool

[steve ravet]

Ludis Langens wrote:
> 
> The data could be written as initialized C data structures.  However, I
> haven't encountered an environment where this is space and speed
> efficient.  Also, I expect the tables to enventually be bigger than some
> environments can handle in their global data space.

It's more difficult in C to make data structures both packed and
portable but you can do it, even to the extent of packing bitfields into
integer fields.  It's more common for really large data to be stored in
a file and manipulated dynamically than to be compiled into the program,
why did you decide to do it that way?  Compilers and linkers used to
support overlays for data in the past but not anymore, because code and
data spaces have gotten much larger.  How much data do you think you are
talking about?

> I would prefer to create the data tables using a sort of "data compiler"
> I have.  This would allow for the most editable table source code.
> Unfortunately, this compiler is specific to my development environment.
> This would mean that the data table source could only be edited by
> someone with access to the same development environment I have.  This
> compiler's output could be converted to a raw data block which could be
> moved to other development platforms though.  Another negative is that
> the raw data would have a specific integer "endianess" baked into it,
> although it is possible to write a special purpose tool to flip all
> multibyte values in the data.

You could distribute the data compiler also, assuming it's written in C
or something portable.

--steve

-- 
Steve Ravet
steve.ravet at arm.com
ARM,Inc.
www.arm.com
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org