Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!jpl-devvax!lwall
From: lwall@jpl-devvax.jpl.nasa.gov (Larry Wall)
Newsgroups: comp.lang.perl
Subject: Re: 'H' and 'h' in pack/unpack.
Message-ID: <1991Jun10.180647.5694@jpl-devvax.jpl.nasa.gov>
Date: 10 Jun 91 18:06:47 GMT
References: <1991May27.043119.12173@Sandelman.OCUnix.on.ca>
Reply-To: lwall@jpl-devvax.JPL.NASA.GOV (Larry Wall)
Distribution: comp
Organization: Jet Propulsion Laboratory, Pasadena, CA
Lines: 65

In article <1991May27.043119.12173@Sandelman.OCUnix.on.ca> mcr@Sandelman.OCUnix.on.ca (Michael Richardson) writes:
:   I have some binary data which I'd like to turn into S-records. (My
: knowledge of S-records is far from complete though). I also have a
: 6809 assemblers that produce S-records in a format that I don't think
: is appropriate for the 6809 board I want to download them to.
:   I thought that Perl would be appropriate, with the use of
: pack/unpack for the S-record manipulation. Checking my copy of the
: Perl book, I see 
:   h - A hexadecimal string, low nybble first.
:   H - A hexadecimal string, high nybble first.
: 
:   (I've never seen hex where the low nybble was first..)

Me neither, but if someone had a vec($nums, $offset, 4) kind of string, they
might want it unpacked into the same order as the offsets would imply.

:   I've come to the conclusion that these options are NOT for extracting
: hexadecimal data to and from (binary) strings. I finally checked the
: code (in doarg.c and dolist.c) but that didn't help. I'm really not
: sure what that code does as it doesn't seem to have the loop in the
: right place. 
:   What are they for?? Is there a way to use pack/unpack to do what I
: want?

The hex pack/unpack options are indeed for converting a string of nybbles to 
a string of bytes, and versa vica.  There was a bug in 4.003 in packing
anything other than H*, but that's not how you'd use it here anyway.

:   S-record format, btw, is something like this:
: 
: 
: S113AABB00112233445566778899aabbccddeeffZZ
: 
: 	S1	start of data record
: 	13	number of data bytes in the record + 3 (16 decimal)
: 	AABB	data to be loaded at address AABB (16 bit address)
: 	00112233445566778899aabbccddeeff	data
: 	ZZ	checksum of all bytes (include 13, don't include S1)
: 
: 
:   I'm resorting to hex(), loops and sprintf("%02x",$val) right now.
: 
:   BTW: this is 4.003 as posted to comp.sources.unix.

Okay, I just ran this under 4.003:

    #!/usr/bin/perl

    for (0..15) {
	$bytes[$_] = $_ + ($_ << 4);
    }
    $addr = 0xaabb;
    $bytes = pack('C n C16', 0x13, $addr, @bytes);
    $csum = unpack('%8C*', $bytes);
    $bytes .= pack('C', $csum);
    $nybbles = unpack('H*', $bytes);
    print 'S1', $nybbles, "\n";;

and it printed out:

    S113aabb00112233445566778899aabbccddeeff70

I'm presuming you really wanted a checkSUM and not a checkXOR.

Larry