To be able to reasonably guess in which project the job is defined
looking at its name rename it to something more reasonable at the same
time assuming there might be more of the publish jobs (well, we can
generate lot of different things).
Change-Id: I5c2779c9418c28bc4af4f31a80661b87b4d78848
Add a job that collects built OpenAPI specs and prepares a single
delivery. Those can be then places to some authority location similar to
the service-types-authority.
Change-Id: I18bfb2a5195f88153d3a23072bc8a4351a5b7f1b
- update domain schema
- fix domain.config schemas
- update generator templates to deal with dict of dicts in domain.config
case
Change-Id: I51a828d8b0729d2543186f68bc1e42c83b81d1ff
Cinder implements os-volume-transfer and volume-transfers API at the
same time. Since they have certain differences we should treat them
independently (also avoiding stripping os_ from the resource name).
Change-Id: I8c5350e3cc479f36ebcd11caa53ba5000182b5c7
- manageable snapshot is allowing nullable json object on input what
required small adaptions in the generator.
Change-Id: I08d0f2474d09a3177850afec2a70b4191fa3859a
inspect shows that cinder exposes PUT requests to hosts (and it is
present in the code). Those things are, however, absolutely
undocumented, so for now they are blacklisted in the metadata.
Change-Id: I042f85ce545ce387d388759bd5a637f0dd61bac1
python docstrings are assumed to be RST, while for OpenAPI we need MD.
Convert controller description from rst->html->md using docutils.
Change-Id: I04acab45f5cf9d05b5d7158e484a14c4343bdb24
by accident list operation was not framed in the "object" container for
block-storage volume attachments.
Change-Id: I13960a65b4b23cadc10d9547bfdc38c6d6b8d826
"type: int64" is wrong and "format: int64" was meant instead
Also ensure operation_type is selected as "show" instead of "list" in
the metadata since it returns object and not list.
Change-Id: I243c74094a000ecf61edab998ae01ad432ab889e
cinder_schemas becomes too big and so it is time to split it into
individual smaller modules similar to how it is done in
keystone_schemas. This helps keeping code more clear
Change-Id: Ibd65663b6c94b9037817d6a01feee92abfa2c099
We had octavia spec already, but once code generation been attempted on
it series of issues were identified (mostly coming from deeper nesting
and weird path naming).
Change-Id: I4ecbecb17ce5bfdd0c655644a8b8349060f7c79d
Octavia models uncover issues in the way of handling model name
conflicts. There is very deep nesting with same model schemas. Instead
of adding model type name try to add parent name as prefix.
Change-Id: I4a53a68777aceb1dfb6d2d4fc8ac9f0d536ec5e9
federation mapping schema in keystone is broken. Fix it locally while
waiting for the schema fix to land upstream
Change-Id: I240c463e859ff0224495ee6c97eec35b97b0f7f7
In order to be able at some point in time to get rid of api-ref it is
necessary to start including more information directly in the code.
Change-Id: Ie8ea581f89008762a5e97d4c039026d299365ce8
