On Sat, Dec 03, 2011 at 04:05:17PM +0100, Jakub Wilk wrote:
> * Miguel de Val-Borro <miguel.de...@gmail.com>, 2011-12-03, 13:16:
> >--- numexpr-1.4.2/debian/rules 2011-04-07 04:10:59.000000000 +0200
> >+++ numexpr-1.4.2/debian/rules 2011-12-03 12:13:42.000000000 +0100
> >@@ -19,4 +19,7 @@
> > install/$(DEB_PYTHON_MODULE_PACKAGES)::
> > sed -i 's#\#!/usr/bin/python[0-9].[0-9]#\#!/usr/bin/python#' \
> > $(cdbs_python_destdir)usr/bin/*
> >+
> >+build/python-numexpr::
> >+ dh_numpy
> > # rm -f $(cdbs_python_destdir)usr/share/doc/python-numexpr/LICENSE.txt
>
> That's an unusual place to call dh_numpy (or any helper that generates
> dependencies). The patch happens to do the trick for the moment, but
> only because dh_numpy is quite dumb and doesn't look at package
> contents. If dh_numpy ever grows smarter, this package will break.
Would it be correct to call dh_numpy under the install rule?
install/python-numexpr::
dh_numpy
When the package is built like that it also has the right dependency on
numpy.
Miguel
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
