[reprotest] 03/03: Don't vary user_group in the tests because it's not usually possible in automated builders
Ximin Luo
infinity0 at debian.org
Mon Aug 21 18:37:44 UTC 2017
This is an automated email from the git hooks/post-receive script.
infinity0 pushed a commit to branch master
in repository reprotest.
commit 977181f46ec9acb2ac260dc9929285ee030de952
Author: Ximin Luo <infinity0 at debian.org>
Date: Mon Aug 21 20:35:38 2017 +0200
Don't vary user_group in the tests because it's not usually possible in automated builders
---
debian/rules | 5 ++++-
reprotest/__init__.py | 2 +-
2 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/debian/rules b/debian/rules
index f8de5eb..604b183 100755
--- a/debian/rules
+++ b/debian/rules
@@ -8,7 +8,10 @@ export PYBUILD_NAME = reprotest
# It's hard to make sure disorderfs works on buildds, so drop it.
# The build user needs fuse access and this is not easily arranged.
-export REPROTEST_TEST_DONTVARY = fileordering
+#
+# To be user-friendly the user_group variation defaults to ON but is a no-op.
+# This causes tests to fail since they expect something to be captured. So ignore it here
+export REPROTEST_TEST_DONTVARY = fileordering,user_group
override_dh_auto_configure:
test $$(python3 setup.py --version) = $$(dpkg-parsechangelog -SVersion)
diff --git a/reprotest/__init__.py b/reprotest/__init__.py
index 24eadc7..aac5b79 100644
--- a/reprotest/__init__.py
+++ b/reprotest/__init__.py
@@ -359,7 +359,7 @@ def umask(ctx, script, env, tree, *args):
# as the other user (e.g. due to permissions).
def user_group(ctx, script, env, tree, *args):
if not ctx.user_groups:
- logging.warn("IGNORING user_group variation, because no --user-groups were given.")
+ logging.warn("IGNORING user_group variation, because no --user-groups were given. To suppress this warning, give --dont-vary user_group")
return script, env, tree
olduser = getpass.getuser()
oldgroup = grp.getgrgid(os.getgid()).gr_name
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/reproducible/reprotest.git
More information about the Reproducible-commits
mailing list