home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group03a.txt
/
000087_icon-group-sender_Wed Aug 13 07:36:26 2003.msg
< prev
next >
Wrap
Internet Message Format
|
2003-12-22
|
2KB
Return-Path: <icon-group-sender>
Received: (from root@localhost)
by baskerville.CS.Arizona.EDU (8.11.1/8.11.1) id h7DEYiM04035
for icon-group-addresses; Wed, 13 Aug 2003 07:34:44 -0700 (MST)
Message-Id: <200308131434.h7DEYiM04035@baskerville.CS.Arizona.EDU>
Date: Tue, 12 Aug 2003 17:06:46 -0700 (MST)
From: Gregg Townsend <gmt>
To: ahamm@mail.com
Cc: icon-group
Subject: Re: pre-processing?
Errors-To: icon-group-errors@cs.arizona.edu
Status: RO
> OK - thanks for the reply. I've virtually settled on just poking the version
> details into the .u1 straight after compiling. I suppose I could latch onto
> the first procedure header in each file and inject a string.
Injecting a trivial procedure at end of file might be even easier.
> Looking in a basic .u1 emitted from 9.4.1, I see:
>
> proc thingy
> filen tester.icn
> local 0,001000,a
> local 1,001000,args
> local 2,000020,b
> local 3,000020,c
> con 0,010000,6,163,164,162,151,156,147
> con 1,002000,2,10
> declend
>
> so it seems reasonable to find the first proc, then count the "con"s and add
> a new con with the next available serial number just before the declend?
>
> Do you, from your position of great inner knowledge, see any downside to
> this technique being applied in the Makefile?
I *think* that should work, assuming you have a way to get it back out.
But it's been a long time since I've mucked with the icode.
---------------------------------------------------------------------------
Gregg Townsend Staff Scientist The University of Arizona
gmt@cs.arizona.edu Computer Science Tucson, Arizona, USA