[pacman-dev] Checking for VSC/SCM tools in makepkg

[Allan McRae]

Xavier wrote:
> On Mon, Mar 9, 2009 at 9:36 AM, Allan McRae <allan at archlinux.org> wrote:
>   
>> Yeah, I was just wondering if the devel_check should come after the
>> installation of deps and makedeps so that if a VCS tool is missing, we can
>> make it an error.  At the moment, we can not update the pkgver during the
>> first devel_check.  So you end up with issues like this:
>>
>> ==> Making package: notify-sharp-svn 3009-2 i686 (Mon Mar  9 08:23:02 UTC
>> 2009)
>> ...
>> ==> Installing missing dependencies...
>> (subversion)
>> ==> Entering fakeroot environment...
>> ==> Determining latest svn revision...
>>  -> Version found: 3032
>> (note the different version there and how --forcever does not seem to have
>> been used/work)
>> ...
>> ==> Finished making: notify-sharp-svn 3009-2 i686 (Mon Mar  9 08:25:36 UTC
>> 2009)
>>
>>     
>
> I see, but in any cases, we won't be able to display this message
> where it is now :
>   
>> ==> Making package: notify-sharp-svn 3009-2 i686 (Mon Mar  9 08:23:02 UTC
>>     
> unless we either remove the version, or move this message after
> installing missing dependencies (could look weird).
>
> Here are the other things that happen between the current devel_check
> and the missing deps install, which relies on $pkgver :
> 1) check of an existing package (and install with -i)
> 2) creation of the source package
>
> For 2, I don't see why it would be needed to bump the pkgver to the last one.
> For 1, it might be ok to not bump it either. This would make us aware
> that there is already a package built for the pkgver in the pkgbuild,
> so that we might want to just install that one. Otherwise, we can
> force with -f, which may or may not result in a different $pkgver
> after devel_update or devel_check, but well, that's what --holdver is
> for.
>
> So it seems doable to delay devel_check, having the three points above in mind.
>
>   


I am going to file a bug report so I do not forget about this.  Given we 
may have a release soon, I don't think I want to make all the changes 
that this would require at the current time.


>> Hmmm..... while testing this I noticed this:
>> ==> Validating source files with md5sums...
>> ==> Validating source files with sha1sums...
>> ==> Validating source files with sha256sums...
>> ==> Validating source files with sha384sums...
>> ==> Validating source files with sha512sums...
>>
>> Note that svn packages tend to have "md5sums=()" in the PKGBUILD.  I wonder
>> if that is still an issue?
>>
>>     
>
> Ahah.
> What about [ ${#source[@]} -eq 0 ] && return 0 at the top of check_checksums ?

That would be a good solution.  I can patch this up later.  I have never 
actually looked at the check_checksums function before...