Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!snorkelwacker.mit.edu!bloom-beacon!eru!hagbard!sunic!mcsun!ukc!axion!tharr!gtoal
From: gtoal@tharr.UUCP (Graham Toal)
Newsgroups: comp.compression
Subject: Re: Y coding
Message-ID: <2159@tharr.UUCP>
Date: 18 May 91 19:56:04 GMT
References: <1991May8.194136.3683@doc.ic.ac.uk> <26227:May1416:59:4091@kramden.acf.nyu.edu>
Reply-To: gtoal@tharr.UUCP (Graham Toal)
Organization: Public Usenet Access
Lines: 36

In article <26227:May1416:59:4091@kramden.acf.nyu.edu> brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes:
:In article <1991May8.194136.3683@doc.ic.ac.uk> ra2@doc.ic.ac.uk (Roger Attrill) writes:
:> Sorry if this has been asked before, but I've been absent from the group
:> for a while. Does anyone have some code - preferably C - implementing
:> the Y coding which was described here some time ago.
:
:The reference implementation of Y coding is my yabbawhap package, in
:comp.sources.unix volume 24. The source is in C, and it's been reported
:to work under a variety of UNIX and non-UNIX systems.

Dan - has anyone hacked yabba yet so that it can be used as an
encapsulated part of other programs - for instance, an interface
like  'init_yabba(); compress_yabba(inblock, insize, outblock, &outsize);
terminate_yabba()' would be ideal! (to fit in my program :-) )

I've almost finished a shell for compression algorithms, which is
a layer of procedures on top of stdio.h -- fopen(file, mode) for
example would open a compressed file, which you could then read from
with fread().  The compressed file is compressed in blocks, and
there is a spine-block for fast access to random bytes within the
file.  The compression is totally transparent; for instance seeking
to the end of the file and doing an ftell() returns the size of
the expanded file.

(If you use the fake fopen() (which is a macro calling qfopen() ), and
the file doesn't have a compressed magic number, you'll still be able
to read it with the fake calls - at the lowish cost of one extra
level of procedure entry)

I'd quite like to include yabba as one of the base compression algorithms,
hence why I'm asking if anyone has mangled it into a blocked interface...

Regards
Graham
-- 
(* Posted from tharr.uucp - Public Access Unix - +44 (234) 841503 *)