Re: ITP: darj - arj archive unpacking tool
On Sat, Apr 28, 2001 at 09:36:28PM -0700, Alexander Hvostov wrote:
> Why not put the core into a library (libdarj.so or something)? This
> sounds like something that could be used by a multitude of programs
> (eg, Nautilus). This would also allow you to make one program that
> wraps the library and behaves like a proper unix utility, and then
> another that mimics unarj.
I've been thinking about that. Unfortunately a library is harder
to write. For example, darj simply exits if it runs out of memory,
and this simplifies the internal interfaces a great deal. But it's
not polite in a library.
I will consider this after finishing the unarj clone. I'm a fan
of refactoring :) I'll have a much better idea of what should
go into the library after I've designed two different tools that
use it.
> Are you planning to make a compressor at some point? I imagine if you
> can figure out the format well enough to extract from it, it wouldn't
> be much harder to be able to create/modify archives as well.
A compressor tends to be much harder to write than a decompressor,
actually. And I don't think arj as an archive format has much
going for it these days -- it's designed for msdos file attributes
and file types. It can't represent symbolic links, for example.
So, no, I don't intend to write a compressor.
Richard Braakman
Reply to: