* check the proper field for dpkg installation state fixes#2006
Signed-off-by: Mathieu Sauve-Frankel <msf@kisoku.net>
* Properly handle held packages on dpkg-flavored OS
InSpec was looking at the wrong field in `dpkg -s` output to determine
whether a package was installed or not. An installed, held package was
incorrectly reported as uninstalled.
This adds the proper unit tests and also adds a `be_held` matcher.
Thanks to @kisoku for the initial work in #2007.
Signed-off-by: Adam Leff <adam@leff.co>
* Publish rubygem via expeditor
When InSpec is promoted to the `stable` channel, Expeditor will take care of the gem publishing!
Signed-off-by: Adam Leff <adam@leff.co>
With last weeks tag fix, `ourorg/container` ended up having its `repo` reported as `container`.
With this it'll be `ourorg/container` again.
Signed-off-by: Stephan Renatus <srenatus@chef.io>
Reuse blog post content on profile inheritance for web site
based on feedback from colleagues and community members.
Signed-off-by: Adam Leff <adam@leff.co>
No reason to keep the Gemfile.lock in the www directory under source
control. It adds extra steps to the www release process after an InSpec
release because the release process will fail of the git working tree
is not clean (and since the www Rake tasks rely on InSpec stuff, the
Gemfile.lock will be perpetually out of date).
Signed-off-by: Adam Leff <adam@leff.co>
This will allow for four different changelog categories (in addition to the default)
that will be used based on PR label.
Signed-off-by: Adam Leff <adam@leff.co>
* Changes resources/filename to match resource name
Signed-off-by: kgarmoe <kgarmoe@chef.io>
* Deletes original resource files.
Signed-off-by: kgarmoe <kgarmoe@chef.io>
Train 0.26.1 fixes a bug where missing mocked commands would return
a zero exit code. This resulted in some unit tests failing due to
missing mocked commands.
Signed-off-by: Adam Leff <adam@leff.co>
Bumping InSpec's minor version to 1.33 because a recent PR added
new functionality. Also bumping train to 0.26 to pick up a recent
bug fix.
Signed-off-by: Adam Leff <adam@leff.co>
* New matcher 'be_in'
Fixes#2018
Signed-off-by: Rony Xavier <rx294@nyu.edu>
* small fixes to wording.
Signed-off-by: Aaron Lippold <lippold@gmail.com>
* Added code to use be_in for with the following use case:
describe nginx do
its(module_list) { should be_in AUTHORIZED_MODULE_LIST }
end
Fixes#2018
Signed-off-by: Rony Xavier <rx294@nyu.edu>
* Updates to the matcher
Fixes#2018
Signed-off-by: Rony Xavier <rx294@nyu.edu>
* Added tests for the be_in matcher
Signed-off-by: Rony Xavier <rx294@nyu.edu>
* Requested updates completed
Signed-off-by: Rony Xavier <rx294@nyu.edu>
In the event we have a docs-only change or similar which does not necessitate a
version bump, changelog update, or omnibus build, a single label named
`Expeditor: Skip All` can be used rather than adding individual labels.
Signed-off-by: Adam Leff <adam@leff.co>
Version bumping and changelog generation has been moved to
Chef's expeditor tool. The github_changelog_generator bits
and the Rake tasks that use them are no longer necessary.
Also, the old historical changelog has been merged back into
the main CHANGELOG.md file since Expeditor does not alter
any existing changelog content like github_changelog_generator
does, so it's safe to bring this stuff back.
Signed-off-by: Adam Leff <adam@leff.co>
The fix is already provided here: https://github.com/chef/inspec/pull/2040
This PR only adds a unit test to catch it
Signed-off-by: Dominik Richter <dominik.richter@gmail.com>
Add a newline symbol to the end of the parsed input.
Sample hexdump of a file deployed by xinetd cookbook:
$ hexdump -C /var/chef/cache/cookbooks/xinetd/templates/default/xinetd.conf.erb | tail -2
000000b0 72 20 2f 65 74 63 2f 78 69 6e 65 74 64 2e 64 |r /etc/xinetd.d|
000000bf
Signed-off-by: Simonas Kareiva <simonas@5grupe.lt>
With this configuration change, expeditor will add a "latest stable"
section to the changelog automatically whenever we promote InSpec
to the stable channel. All existing changelog entries will remain,
and any additional changelog entries that have been made in newer
versions that are newer than the artifact being promoted will
remain intact.
Signed-off-by: Adam Leff <adam@leff.co>
