| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
| |
Also, remove the files associated with this that should not be pulled
from oslo-incubator any longer.
Closes bug 1179007
Change-Id: Id2e882837bdde51a60562e8cdd207c7ff0c88a6a
|
| |
|
|
|
|
| |
One code change, rest are in headers
Change-Id: I73f59681358629e1ad74e49d3d3ca13fcb5c2eb1
|
| |
|
|
|
|
|
|
| |
It was pointed out that we missed the use-case that running tests
in a fresh tree checkout should work - and we missed that by
reaching too far into the setup module interface. Shame on us.
Change-Id: Ic0056a3113b308f063d927484e1e4a44c9316a2d
|
| |
|
|
|
|
|
|
|
|
|
| |
glanceclient, and others, want to have a __version__ member in their
top-level __init__.py. This leads to attempting to process the
version information to realize the string at import time, rather
than at request time. If we make VersionInfo a thing which can
behave like a string, then, we should get the things we want without
the operational weirdness.
Change-Id: I3eb15006f50ed1155c85977aaae842677c5ec4bc
|
| |
|
|
|
|
|
|
|
|
| |
Running tests from the directory of an expanded tarball is a reasonable
thing to do, so support it by checking for PKG-INFO file if the pkg_resources
version grab fails.
Fixes bug 1103473
Change-Id: I2bf1fc39b070d3a2cdf6a3abe5e816c81d2916a0
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In discussions with ttx and zul, it became clear that the git describe info
for our intra-release case wasn't providing any value and was providing
confusion. Additionally, plain git sha's are not providing enough info.
So, the scheme settled on that breaks the least amount of things and yet
still reduces complexity is:
if pre-version: # server projects
try:
use current tag
except:
use pre-version+revcount+gitsha
else:
use git describe | s/-/./
Additionally, we'll use a as a prefix for the revcount, b as the prefix for
the milestone portion of a milestone tag, and rc as the prefix for rc's, so
that for our releases, python version sorting works as expected.
Change-Id: I6f0fe029d225afa8f435bc83216fc144c2402ae0
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In response to the cessation of per-commit tarballs produced by the CI
systems, simplify the version processing code. We did many many many
complicated things to support producing release artifacts that were not
actually supported release artifacts. After a bunch of issues emerged, it was
determined that the best way forward was to stop producing quasi-not-really
non-release tarballs. If people want to consume versions of openstack that
are not released versions, the git repos are always available.
Additionally, this patch removes versioninfo and just uses PKG-INFO and
pkg_resources to handle version processing.
Change-Id: I5c799f3520adb2ba5288d852543706d81a92f8a1
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This currently fails for client projects as python_package
is the name of the project, while package is where versioninfo
is written to. This doesn't affect server openstack projects
because python_package == package.
This hasn't affected any client projects yet because most likely
none of them call version.canonical_version_string() in their
setup.py
Change-Id: Ia97e5fae378c47909f3dbc842dc83e408af68671
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Need to get rid of the deferred string approach as
the problem was overriding string methods is hard and
flaky. There is just no other documented
argparse hook to allow deferring of the version
calculation as version passed in has to be a str
or buffer. Added a test to try longer string and even
run it in a loop just to be sure.
Fixes LP #1087575
Change-Id: Icf83944d67b8ce5d3230627f3de10501da3276f2
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The code for lazy loading of version string was broken. The way it
breaks in glance and oslo is different. This patch is just for oslo.
Here we end up overriding more methods in our lazy string class to make
argparse happy and added test cases to make sure we don't break again.
we also sure we cache the result of the lookup so we return the same
version everytime and for performance. Added a few test cases as well.
Fixes LP #1085984
Change-Id: I0635899c37fa8e5a7edf282907d2a5e04aa064cb
|
| |
|
|
| |
Change-Id: Ia2900e50263950ba40ba86e685640cacaa2137a7
|
| |
|
|
|
|
| |
also,this patch turns off pep8 E125 check.
Change-Id: I22d2a94d89530586d614af39af17ae542c5e0cbf
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Moved the version routines into a class for better structure.
Added caching/memoization logic so that pkg_resources/git doesn't get
called more than is necessary.
Added periodic-based caching that will only cause the meta info refs to
be checked at most once a day, to reduce network load for developers. (The
information in the meta refs should really only change once a month, and
this will only affect local developer workstation version reporting, so we
should be fine here.
Added a deferred version string class and function to allow for consuming of
the version string but doing the calculation only when it is requested,
based on test runs in glance.
Change-Id: I6a436c8817f6904959727a76d876b807ec545ee9
|
|
|
The new tag-based versioning produces files that we can consume that tell
us the versioning. That's all well and good, but what good is automatically
generated versioning if there are no apis to consume it. This is derived
from work to add those to nova and python-swiftclient.
Change-Id: I91568ef28a9ed7e7365f66bca2294670df6e0c41
|
