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

Bug#788659: autopkgtest: Provide option to not implicitly run undeclared tests

On Mon, Jun 15, 2015 at 04:48:11PM +0300, Niko Tyni wrote:
> On Mon, Jun 15, 2015 at 02:32:28PM +0100, Dominic Hargreaves wrote:
> > On Mon, Jun 15, 2015 at 07:30:34AM +0200, Martin Pitt wrote:
> > > So to clarify, you want an option to disable autodep8, right? If a
> > > source package doesn't have tests and autodep8 does not generate any
> > > either, then adt-run will already report "no tests in this package"
> > > and exit with 8. With a perl package and --no-autodep8 it would then
> > > do the same.
> > 
> > I thought about 'no autodep8', but the semantics aren't quite correct,
> > since a package could declare Testsuite: autopkgtest, but rely on the
> > autodep8 behaviour. This is in fact the common case for pkg-perl packages,
> > as far as I can tell.
> FWIW, the pkg-perl packages explicitly declare 
>  Testsuite: autopkgtest-pkg-perl

The Ruby ones declare

   Testsuite: autopkgtest-pkg-ruby

> Not sure if that makes any difference to the plain Testsuite: autopkgtest
> declaration with the current adt-run implementation, but at least it could.

It does, because autodep8 looks at the value of the Testsuite: field.

AFAICT how it all works today:

- debci searches for anything that matches 'autopkgtest' in the
  Testsuite: field, plus its configured whitelist, to see which packages
  have tests to be executed.
- autopkgtest does not care (or didn't, until recently) about the value
  of the Testsuite: field; if you give it a package, it will try to run
  tests for it.
- autodep8 uses the value of the Testsuite: field, together with other
  heuristics, to autogenerate a test control file

Antonio Terceiro <terceiro at debian.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/autopkgtest-devel/attachments/20150615/d64d6181/attachment.sig>

Reply to: