Re: Re: ebuilds in XML

"Kevin Horn" <[email protected]> Mon, 30 Jun 2003 12:22:33 -0500
Newsgroups gmane.linux.zynot.general
Message-ID <055c01c33f2c$32287910$0301a8c0@maxilius>
----- Original Message -----
From: "Mark Bainter" <[email protected]>

> Kevin Horn [[email protected]] wrote:
> > I like it too.  I just get irritated at some of it's limitations.
> > Occaisionally I'll run into an ebuild that I can't read because that
> > package's build process depends on a config file or some such that the
> > Gentoo packager has tweaked using SED/AWK/other.
> >
> > sed and awk make my head hurt.  There's got to be a better way...
>
> A better way than using small commonly available shell tools to do
> the task they were designed for?  About all I can think of is maybe
> documenting what it's doing for those who aren't familiar with the
> tools.  I would be more than mildly annoyed to discover that an ebuild
> required me to install ruby because of an embedded script.
>

Hmmm...perhaps I wasn't being clear.  Scratch that, I WASN'T being clear.
What I would like to see in this area is a generalized method of doing this
sort of thing (mostly for consistency's sake).  Perhaps a bash function, or
a python function would work, but might be hard to make cover all cases.
The problem I have is that these little sed/awk hacks tend to be unique to
the particular ebuild they are found in.  Even providing documentation so
that they are at least implemented in a consistent manner would, IMO, be a
"better way".  Abstracting the "replace text in a given file" would be even
better.

I also agree that adding dependencies to ebuilds (like ruby, etc.) would be
a "bad thing".

Kevin Horn