Compression and Decompression of multiple files

[Alex Rice]

On Feb 7, 2004, at 8:43 AM, Rob Cozens wrote:

> The design I have in mind would create a list of all files & their 
> folder locations via a tweaked Directory Walker command, compress each 
> file individually, append the individual compressed files, and append 
> an index of the archive listing each file's size & byte offset from 
> the beginning of the archive file.

Rob, just some thoughts:

It sounds like you would not be leveraging the structure of a "data 
stack" as the underlying format for the archive? If you aren't going to 
leverage a "data stack" then maybe it should just be a ZIP or TAR 
engine so as to be a more compatible format?

For a non-compatible format, I would use custom properties of a stack, 
to attach compressed files. instead of a byte offset. Don't you think 
that would be better, and easier to code?

Java has "jar" archives, which works kind of like tar plus compression. 
This kind of thing should really be standardized by Runrev (akin to 
Java's "jar" format).

--
Alex Rice | Mindlube Software | http://mindlube.com