From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:44247) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1d25Tb-0000Pu-1p for guix-patches@gnu.org; Sat, 22 Apr 2017 20:34:08 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1d25TW-0007Er-Nl for guix-patches@gnu.org; Sat, 22 Apr 2017 20:34:07 -0400 Received: from debbugs.gnu.org ([208.118.235.43]:36746) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1d25TW-0007El-If for guix-patches@gnu.org; Sat, 22 Apr 2017 20:34:02 -0400 Received: from Debian-debbugs by debbugs.gnu.org with local (Exim 4.84_2) (envelope-from ) id 1d25TW-0002uH-Cz for guix-patches@gnu.org; Sat, 22 Apr 2017 20:34:02 -0400 Subject: bug#26613: Add pootle Resent-Message-ID: Date: Sat, 22 Apr 2017 20:33:21 -0400 From: Leo Famulari Message-ID: <20170423003321.GA10153@jasmine> References: <20170422231314.075c7ba2@lepiller.eu> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha256; protocol="application/pgp-signature"; boundary="jI8keyz6grp/JLjh" Content-Disposition: inline In-Reply-To: <20170422231314.075c7ba2@lepiller.eu> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-patches-bounces+kyle=kyleam.com@gnu.org Sender: "Guix-patches" To: Julien Lepiller Cc: 26613@debbugs.gnu.org --jI8keyz6grp/JLjh Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Sat, Apr 22, 2017 at 11:13:14PM +0200, Julien Lepiller wrote: Thanks for all these patches! > here are patches that add pootle, a translation server written in > django. Some patches modify packages that have several dependents, so > it should probably go to core-updates first. We don't want to update foo and break foo's dependents on the master branch, but unless foo has > 1200 dependents, we can avoid waiting for core-updates. See: https://lists.gnu.org/archive/html/guix-devel/2016-10/msg00933.html For this patch series, the bigger changes are updates to pytest-runner and dateutil, right? That's less than 200 dependent packages. At that scale, all the packages could be built as a test on a developer's machine, or you could just check a few dependent packages and push it to master with fingers crossed ;) Patch-specific comments below. Whatever I don't comment on LGTM. > Subject: [PATCH 01/34] gnu: python-pytest-runner: Update to 2.11.1. >=20 > * gnu/packages/python.scm (python-pytest-runner) > (python2-pytest-runner): Update to 2.11.1. I'd check a handful of dependent packages and cross my fingers :) > Subject: [PATCH 02/34] gnu: Add python-pytest-warnings. >=20 > * gnu/packages/python.scm (python-pytest-warnings, python2-pytest-warning= s): > New variables. > + (description > + "Pytest plugin to list Python warnings in pytest report.") I know it can be hard to write detailed descriptions for these little modules, but please at least write a complete English sentence. For example, "Python-pytest-warnings is a pytest plugin to list Python warnings in the pytest report". > Subject: [PATCH 03/34] gnu: Add python-pytest-catchlog. >=20 > * gnu/packages/python.scm (python-pytest-catchlog, python2-pytest-catchlo= g): > New variables. > + (home-page "https://github.com/fschulze/pytest-warnings") Wrong home-page? > Subject: [PATCH 04/34] gnu: Add python-utils. >=20 > * gnu/packages/python.scm (python-utils, python2-utils): New variables. > + (version "2.0.1") We should use the latest release (2.1.0) or leave a code comment explaining why we are using an older version. I won't mention this again if it's an issue for the rest of the new packages. > Subject: [PATCH 05/34] gnu: Add python-webassets. >=20 > * gnu/packages/python.scm (python-webassets, python2-webassets): New vari= ables. > + (home-page "http://github.com/miracle2k/webassets") Please use the HTTPS URL. > Subject: [PATCH 07/34] gnu: python-rq: Update to 0.7.1. >=20 > * gnu/packages/python.scm (python-rq, python2-rq): Update to 0.7.1. We can add this to the commit message: [source]: Use pypi-uri. > Subject: [PATCH 11/34] gnu: Add python-nosexcover. >=20 > * gnu/packages/python.scm (python-nosexcover, python2-nosexcover): New va= riables. > + (description "Companion to the built-in nose.plugins.cover. This pl= ugin > +will write out an XML coverage report to a file named coverage.xml. "Nose-xcover is a companion to the built-in @code{nose.plugins.cover}. > + > +It will honor all the options you pass to the Nose coverage plugin, espe= cially > +???cover-package.") Is '???cover-package' intended? > Subject: [PATCH 17/34] gnu: Add python-factory-boy. >=20 > * gnu/packages/python.scm (python-factory-boy, python2-factory-boy): New = variables. > + (arguments > + `(#:tests? #f)) Why? > Subject: [PATCH 18/34] gnu: Add python-translate-toolkit. >=20 > * gnu/packages/python.scm (python-translate-toolkit, python2-translate-to= olkit): > New variables. > + (arguments > + ;; TODO: tests don't run properly > + `(#:tests? #f)) Does it indicate a real problem? If not, can you include a bit about how it fails in the comment? > Subject: [PATCH 20/34] gnu: Add python-django-assets. >=20 > * gnu/packages/django.scm (python-django-assets, python2-django-assets): > New variables. > + (snippet > + ;; https://github.com/miracle2k/django-assets/issues/87 > + `(begin > + (substitute* "tests/__init__.py" > + (("settings.configure.*") > + (string-append > + "settings.configure(\n" > + "INSTALLED_APPS=3D['django_assets', " > + "'django.contrib.staticfiles'],\n" > + "TEMPLATES=3D[{'BACKEND': " > + "'django.template.backends.django.DjangoTemplates= '}],\n" > + ")\n"))) > + ;; These tests fail > + (substitute* "tests/test_django.py" > + (("TestLoader") "NoTestLoader")))))) I think we should do this in a build phase so that the result of `guix build --source` is not affected. Origin snippets should be used for removing non-free components and fixing serious security bugs. > + (home-page "http://github.com/miracle2k/django-assets") HTTPS URL, please :) > Subject: [PATCH 21/34] gnu: Add python-django-jsonfield. >=20 > * gnu/packages/django.scm (python-django-jsonfield, python2-django-jsonfi= eld): > New variables. > + (snippet > + '(substitute* "jsonfield/tests.py" > + (("django.forms.util") "django.forms.utils"))))) I'd do this in a build phase, too. > Subject: [PATCH 22/34] gnu: Add python-django-bulk-update. >=20 > * gnu/packages/django.scm (python-django-bulk-update, python2-django-bulk= -update): > New variables. > + (arguments > + ;; tests seem to be broken > + `(#:tests? #f)) I assume the installed package seems to be working ;) > Subject: [PATCH 23/34] gnu: Add python-django-contact-form. >=20 > * gnu/packages/django.scm (python-django-contact-form, python2-django-con= tact-form): > New variables. > + (arguments > + `(#:tests? #f)) Why? > Subject: [PATCH 25/34] gnu: Add python-django-overextends. >=20 > * gnu/packages/django.scm (python-django-overextends, python2-django-over= extends): > New variables. > + (arguments > + `(#:tests? #f)) ? :) > Subject: [PATCH 26/34] gnu: Add python-django-redis. >=20 > * gnu/packages/django.scm (python-django-redis, python2-django-redis): > New variables. > + (arguments > + `(#:tests? #f)) ? :) > Subject: [PATCH 27/34] gnu: Add python-django-rq. >=20 > * gnu/packages/django.scm (python-django-rq, python2-django-rq): New vari= ables. > + (arguments > + `(#:tests? #f)) ? :) > Subject: [PATCH 28/34] gnu: Add python-django-sortedm2m. >=20 > * gnu/packages/django.scm (python-django-sortedm2m, python2-django-sorted= m2m): > New variables. > + (arguments > + `(#:tests? #f)) ? :) > Subject: [PATCH 33/34] gnu: Fix python2-django-allauth. >=20 > because python-openid is not the same package as python2-openid, > package-with-python2 doesn't work in packages that depend on it. I think this will need the python2-variant system, which was introduced to handle Python packages whose python-2 variants can't be automatically created by package-with-python2. See: https://bugs.gnu.org/22437 > Subject: [PATCH 34/34] gnu: Add pootle. >=20 > * gnu/packages/django.scm (pootle): New variable. > + (version "2.8.0rc5") Usually we don't package beta versions or release candidates, but sometimes it's the right thing to do. What's the story here? > + (modules '((guix build utils))) > + (snippet > + '(begin > + (mkdir-p "pytest_pootle/data/po/.tmp") > + (substitute* "Pootle.egg-info/requires.txt" > + (("1.7.3") "1.8.0") > + (("2.0.0") "2.1.0")) > + (substitute* "requirements/tests.txt" > + (("=3D=3D3.0.6") ">=3D3.0.6")) > + (substitute* "requirements/base.txt" > + (("1.7.3") "1.8.0") > + (("2.0.0") "2.1.0")))))) Let's do this in a build phase. > + (arguments > + `(#:python ,python-2 If it only supports python-2, can you mention that in a code comment? > + #:tests? #f)) Why skip the tests? > + (description "Community localization server.") How about copying the text from for the description? --jI8keyz6grp/JLjh Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- iQIzBAABCAAdFiEEsFFZSPHn08G5gDigJkb6MLrKfwgFAlj79k4ACgkQJkb6MLrK fwhtvhAAwoxMselt+wuJWh7kR9Rjm2FfcVjMFWaLVeRZjTjcyDzyeqzOSecTWzHy KH/4mGB+TXN8ql8Vp+kPt3psJqkBAMG/hxCj/wNi7FFYyNA1DuenjDd+YvoZCs+3 GCG8KivQwwMlynwBLr3IOpBg4rACxSdoihMAmjeRA/0PcNIOIWDIbfzMQ/v7+O7f u7mmu3oeGO0XxOBV4jba+Y3d6YlAjI4PIgQBjgZ5HwyUyTfWRBSmhRpYGBvhMl2B c9Om+rAD/3HaiqMTqt43e6jOvnRRyQ16RrTXnkLfs9dfEwBdn7NmifkkVeLKa7Xh Uc3lD42IpqJzN8aiub9KOUOR9bkmTZidI0SI64csrf1KtZJL3t5mBsI2hu1XCr9Q QA5uB8FPvicGNDKb4Q03kgHiFSrmpBw9LMw0O8KlZdgDwVwqGgQKqEbrCl29Gl13 kl2WkwBQxplZA9YnBJ5W9EgUQ1WAU8W8cIFTDwBM6HNTy5Qe/MsIwba7LJ8R8t2N lFCX+JVfX9L/cEUpED8AbZEuYtC3vS8pQxks5YU1O6dqrhRetB9RSjYMUJKWKi9E OpFgZ9Eh45+gBxhKKeE8+g+ZXDNA6EUkxrVuXeAn8pIocc8JSKo2qSAzSHCDqLZ2 skBGead16N9pdCU0ogv/Pv4AzdT9kd5QzftkZJotOTW1XvdNg5Q= =hl7b -----END PGP SIGNATURE----- --jI8keyz6grp/JLjh--