[Chicago-talk] ASCII compression?

[Jim Thomason]

Oh, another monkey wrench I can throw into the works - it is unlikely
that there will be much repeated data. 2 number sequences are
reasonable, 3 number sequences are possible. Beyond that is very
unlikely to see duplicates.

The more that I think about it,t he more that I think my approach of
converting to base36 is the best.

-Jim....


On Thu, 28 Oct 2004 16:41:50 -0500, Jim Thomason <thomasoniii at gmail.com> wrote:
> Because my output is being fed into a single 3of9 barcode, which is
> alphanumeric only. Upper case letters, numbers, and spaces. All I got.
> 
> If my resulting compressed output contains anything different, it's worthless.
> 
> Presumably I could then encode that output into alphanumeric, but
> that's very likely to end up increasing the string size. So I'd need a
> very tight compression algorithm with a very tight ascii converter.
> Not saying that it doesn't exist, it just sounds like a longshot vs.
> something that just compresses into alphanumeric.
> 
> -Jim....
> 
> 
> 
> 
> On Thu, 28 Oct 2004 16:19:38 -0500, Don Drake <don at drakeconsult.com> wrote:
> > Why ASCII only?  Why not use your favorite compression and then pass it to
> > Uuencode?
> >
> > http://search.cpan.org/~mlehmann/Compress-LZF-1.0b/LZF.pm
> >
> > and
> >
> > http://search.cpan.org/dist/Convert-UU/lib/Convert/UU.pm
> >
> > Do the opposite when you want the data back.
> >
> > -Don
> >
> >
> >
> > -----Original Message-----
> > From: chicago-talk-bounces at mail.pm.org
> > [mailto:chicago-talk-bounces at mail.pm.org] On Behalf Of Jim Thomason
> > Sent: Thursday, October 28, 2004 3:46 PM
> > To: Chicago.pm chatter
> > Subject: [Chicago-talk] ASCII compression?
> >
> > Anybody know of any good algorithms to do ascii-only compression?
> >
> > That is, I'm going to spit in ascii data, and I want to read out ascii data.
> >
> > More specifically, I have 14 chunks of 6 numbers. And I want to
> > compress it all into a single string that contains only upper case
> > letters and numbers that can be later decompressed and broken back
> > into those 14 chunks.
> >
> > Currently, I'm base-36 encoding the chunks. It seems to save me 33% of
> > the space, 6 digits -> 4 digits.
> >
> > Further, ideally, I'd like to avoid using an external lookup table. If
> > I'm going to have an external table, I'd just dump my entire value
> > into it and then read it out as a single integer ID lookup value. :-)
> > I may end up just falling back on this, though.
> >
> > Any suggestions?
> >
> > -Jim.....
> > _______________________________________________
> > Chicago-talk mailing list
> > Chicago-talk at mail.pm.org
> > http://mail.pm.org/mailman/listinfo/chicago-talk
> >
> > _______________________________________________
> > Chicago-talk mailing list
> > Chicago-talk at mail.pm.org
> > http://mail.pm.org/mailman/listinfo/chicago-talk
> >
>