Re: undump tool (was: Re: Base85 for DOS

On Wed, 28 Aug 2024 08:53:42 +0100
"Kerr-Mudd, John" <admin@127.0.0.1> wrote:

On Wed, 28 Aug 2024 09:17:20 +0700
JJ <jj4public@outlook.com> wrote:
 

On Tue, 27 Aug 2024 08:12:15 +0100, Kerr-Mudd, John wrote:

Last chance: could it be that in your environment ds<>es? - here's the same
code with a push pop at the start & a few more exits if run out of text..

 
No. It wouldn't be a DOS compatible system if DS!=ES at COM program startup.
So, same result in Windows XP. And I did tested it.

Well something was different - ah well, we got there.

 

1E07B43F B5FDBAC3 01CD2191 E3F989D7 89D65781 3D2D6474 06B03EF2 AEE354B0
0AF2AEE3 26B020F2 AEE3208A 05473C30 7226244F D43788C2 8A054724 4FD43788
D4D51088 04464974 0FE2E05A 89F129D1 43B440CD 214BEBAA 3C2075ED 380574BF
EBE74E6F 74206120 64626720 66696C65 0D0A245A BA6201B4 09CD21C3

 
This one works

 
Phew!
 
 

but still require the `->` custom marker, which I don't want,
because it won't work for standard DEBUG dump from anyone.

 
 
Oh dear. I felt sure I'd catered for that.
So I just tested, yup, it's OK here.
 
 

The prog tests the first 2 chars of the input for '-d',
if so it is deemed to be a DOS DEBUG file, if not a scan for '>' is
performed to "check" for a GRDB one. if neither it puts out an error
message.
It is then assumed that the remaining lines of input are of
format:

[prelude][space] [set of hex pairs] [double space] [other stuff to eol]

where:
[prelude] is typically segaddr:offset but must have no embedded
spaces.
 
[set of hex pairs] is e.g. [01 02 03 04-AB CD ..] with either space or '-'
between the pairs (anything <'0' actually).

[other stuff to eol] *can* contain spaces, the prog just skips
until LF seen

{-q should get ignored}

Given that you have a H2B prog and 'debug' you can see for yourself the
simplicity of the code.

--
Bah, and indeed Humbug.