Xref: utzoo comp.mail.mh:537 comp.sys.hp:2378
Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!tut.cis.ohio-state.edu!att!cbnewsc!gregg
From: gregg@cbnewsc.ATT.COM (gregg.g.wonderly)
Newsgroups: comp.mail.mh,comp.sys.hp
Subject: Re: mh on hp 9000 series
Message-ID: <1596@cbnewsc.ATT.COM>
Date: 7 Jul 89 14:56:19 GMT
References: <1909@zaphod.axion.bt.co.uk>
Organization: AT&T Bell Laboratories
Lines: 45

From article <1909@zaphod.axion.bt.co.uk>, by crouch@crunchie.axion.bt.co.uk (the man whose head expanded):
> I've managed to get mh to run on our hp cluster; however we have a problem
> with the scan command. Scan doesn't seem able to extract the dat from the
> message, so when we type scan we get output such as:
> 
>  6  00/01 To:crouch@axion.b  starting getalias
> 
>     ^^^^^	This should be 07/06.
> 
> Has anyone seen this before or can suggest possible fixes.

One problem that I am aware of with date processing is in
~mh/zotnet/tws/dtimep.lex.  There are others there too when you do not
pass it a strictly conforming date.

change the lines near line 203 from

[0-2]{d}{d}{d}{w} {
					tw.tw_hour = CVT2;
					tw.tw_min  = CVT2;
					BEGIN Z;
				}

to

[0-2]{d}{d}{d}{w} {
					if (tw.tw_hour) {
						cp += 2;
						tw.tw_year = CVT2;
						tw.tw_zone = ZONE(-5 * 60); /*  HACK, force EST */
					} else {
						tw.tw_hour = CVT2;
						tw.tw_min  = CVT2;
						BEGIN Z;
					}
				}

Without seeing the exact date line that you are passing it, I can not
say for sure, but this will certainly fix some problems.  Particularly,
this fixes the case where the date line does not have a Time Zone
with it (it should, but some don't).

-- 
-----
gregg.g.wonderly@att.com   (AT&T bell laboratories)