automake
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Option Consistency


From: Eric Siegerman
Subject: Re: Option Consistency
Date: Tue, 30 Jul 2002 12:51:43 -0400
User-agent: Mutt/1.2.5i

On Tue, Jul 30, 2002 at 07:46:18AM -0700, Bruce Korb wrote:
> [...] you still
> need to have both [suffix and compressor-program name] be
> specifiable so you are not boxed into
> cutting a new release in order to support a combination
> you didn't plan for.
> >
> >   for f in $DIST_FORMATS; do
> >     eval suffix=\$${f}_SUFFIX
   +    eval program=\$${f}_COMPRESSOR
> >     case $f in
> >       gzip)    ...
> >       bzip2) ...
> >       zip)   ...
> >       compress) ...
> >       none | cat) ...
> >       ...
> >     esac
> >   done

The example assumes that the compressor program is the same as
the case regexp.  For full generality, the latter should be a
generic name for the compression format (as, in the example, they
all are), with the program to produce that format specifyable
some other way.  After all, on DOS it's pkzip, and on Windows
it's probably WinZip with some funky command-line options.

Figuring out the compressor program for each desired format
(gzip_COMPRESSOR etc.) probably *should* be done by configure
(the operative instance being the one run on the release
builder's system of course).

N.B.: I called the new variables foo_COMPRESSOR, to leave room
for a foo_DECOMPRESSOR.  "make dist" doesn't need the latter, but
the autoconf foo-figure-outer macros should be (able to be) fully
generalized.

--

|  | /\
|-_|/  >   Eric Siegerman, Toronto, Ont.        address@hidden
|  |  /
Anyone who swims with the current will reach the big music steamship;
whoever swims against the current will perhaps reach the source.
        - Paul Schneider-Esleben



reply via email to

[Prev in Thread] Current Thread [Next in Thread]