[pacman-dev] About makepkg dlagents

[Dan McGee]

On Mon, Oct 26, 2009 at 9:15 AM, Xavier <shiningxc at gmail.com> wrote:
> On Mon, Oct 26, 2009 at 3:03 PM, Ciprian Dorin, Craciun
> <ciprian.craciun at gmail.com> wrote:
>>
>>    So based on what I've seen a build system has the following
>> responsabilities:
>>    * compile-time dependency checking; (like ./configure does, but
>> without the actual configuration;)
>>    * build environment preparation; (this includes downloading needed
>> source code, checking it, patching, etc.;)
>>    * configuring the build according to the package specification;
>> (invoking ./configure;)
>>    * building the package; (make;)
>>    * creating a deployable package;
>>
>>    Now getting back with the problem where we want to reduce the
>> snipet of code you've highlighted. We can see that the purpose of that
>> code is to "prepare the build environment", more exactly fetching the
>> source code, but we are forced to showell it inside the "build stage"
>> function.
>>
>>    So the reason why I've created the patch is because I want to
>> tread source code comming from Git just like any other source-code
>> bundle, without beeing forced to put it inside the build function.
>>
>>
>
> That's a very good point, and an idea which has been floating for a
> while would be to introduce a prepare() function.
>
> I think this is actually a requirement for makepkg -e to work in many
> cases, where patches are applied or a scm is used.
>
> Then, for limiting code duplication, we could use another idea which
> has been floating for a while : library functions.
> http://bugs.archlinux.org/task/10375
>
> Just an example of how this could work :
>
> prepare() {
> source $libdir/scm
> gitprepare $_gitroot $_gitname
> }

Xavier found the thing I was looking for. I've actually thought about
this idea for a while as being a good one, it just needs to be done
right as once a function gets out there it is hard to pull it back.

I'm a much bigger fan of the prepare() approach (and providing some
solid utility functions) as opposed to hacking the source array. While
the simple case sometimes works, see this example for a time when a
bash function would be a lot easier to use:
http://repos.archlinux.org/wsvn/packages/libfetch/repos/core-i686/PKGBUILD
We're getting close to the point where a few more standard function
names could come into play: things like prepare(), build(), package(),
test().

With that said, I like your effort you have put into this. It is an
area we don't do the best right now, but providing solid footing for
development packages is always a good idea.

-Dan