On 06/26/2018 08:57 AM, Takashi Yamamoto wrote: > On Tue, Jun 26, 2018 at 10:13 PM, Doug Hellmann <d...@doughellmann.com> wrote: >> Excerpts from Lance Bragstad's message of 2018-06-25 22:51:37 -0500: >>> Thanks a bunch for digging into this, Tony. I'll follow up with the >>> oauthlib maintainers and see if they'd be interested in these changes >>> upstream. If so, I can chip away at it. For now we'll have to settle for >>> not treating warnings as errors to unblock our documentation gate [0]. >>> >>> [0] https://review.openstack.org/#/c/577974/ >> How are docstrings from a third-party library making their way into the >> keystone docs and breaking the build? > in the same way that docstrings from os-vif affect networking-midonet docs. > i.e. via class inheritance
Correct, keystone relies in an interface from that library. I've reached out to their community to see if they would be interested in the fixes upstream [0], and they were receptive. Until then we might have to override the offending documentation strings somehow (per Doug's suggestion in IRC) or disable warning as errors in our build [1]. [0] https://github.com/oauthlib/oauthlib/issues/558 [1] https://review.openstack.org/#/c/577974/ > >> Doug >> >>> On 06/25/2018 07:27 PM, Tony Breeds wrote: >>>> On Mon, Jun 25, 2018 at 05:42:00PM -0500, Lance Bragstad wrote: >>>>> Keystone is hitting this, too [0]. I attempted the same solution that >>>>> Tony posted, but no luck. I've even gone so far as removing every >>>>> comment from the module to see if that helps narrow down the problem >>>>> area, but sphinx still trips. The output from the error message isn't >>>>> very descriptive either. Has anyone else had issues fixing this for >>>>> python comments, not just docstrings? >>>>> >>>>> [0] https://bugs.launchpad.net/keystone/+bug/1778603 >>>> I did a little digging for the keystone problem and it's due to a >>>> missing ':' in >>>> https://github.com/oauthlib/oauthlib/blob/master/oauthlib/oauth1/rfc5849/request_validator.py#L819-L820 >>>> >>>> So the correct way to fix this is to correct that in oauthlib, get it >>>> released and use that. >>>> >>>> I hit additional problems in that enabling -W in oauthlib, to pevent >>>> this happening in the future, lead me down a rabbit hole I don't really >>>> have cycles to dig out of. >>>> >>>> Here's a dump of where I got to[1]. Clearly it mixes "fixes" with >>>> debugging but it isn't too hard to reproduce and someone that knows more >>>> Sphinx will be able to understand the errors better than I can. >>>> >>>> >>>> [1] http://paste.openstack.org/show/724271/ >>>> >>>> Yours Tony. >>>> >>>> >>>> __________________________________________________________________________ >>>> OpenStack Development Mailing List (not for usage questions) >>>> Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe >>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev >> __________________________________________________________________________ >> OpenStack Development Mailing List (not for usage questions) >> Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe >> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev > __________________________________________________________________________ > OpenStack Development Mailing List (not for usage questions) > Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
signature.asc
Description: OpenPGP digital signature
__________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
