Uh oh!
There was an error while loading. Please reload this page.
Spanner: Re-generate gapic and drop usage of gax - #4695
Conversation
theacodes
commented
Jan 4, 2018
Towards #4692 |
theacodes
commented
Jan 4, 2018
@lukesneeringer for co-located generated code, we need to make this line: into a template so we can make it say: Will manually fix for the time being. Do you wanna file a toolkit bug for it? |
Also note that the generator clobbered |
theacodes
commented
Jan 4, 2018
Additionally, the autogenerated tests failed to account for the true location of |
lukesneeringer
commented
Jan 4, 2018
I think that might be configurable already. I will check. |
| super(HTTPIterator, self).__init__( | ||
| client, item_to_value, page_token=page_token, | ||
| max_results=max_results) | ||
| self.client = client |
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
| # limitations under the License. | ||
| """Google namespace package.""" | ||
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
| # limitations under the License. | ||
| """Google Cloud namespace package.""" | ||
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
| ): | ||
| for name, message in get_messages(module).items(): | ||
| message.__module__ = 'google.cloud.spanner_admin_instance_v1.types' | ||
| setattr(sys.modules[__name__], name, message) |
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
This comment was marked as spam.
This comment was marked as spam.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
theacodes
commented
Jan 5, 2018
@lukesneeringer this should be ready to review and merge. While the changes to the library itself are minor, there had to be a lot of changes to the tests to remove gax as they were extremely tightly coupled. |
theacodes
commented
Jan 5, 2018
@lukesneeringer seems the generator spit out a bad docstring, see 78dc4b0. |
lukesneeringer
commented
Jan 8, 2018
My guess is that is the Markdown to Restructed Text conversion that we do, meaning it is a bug in pandoc (or pypandoc). I will attempt to prove this and then file a bug with them. |
theacodes
commented
Jan 8, 2018
In the meantime, can you file a bug on toolkit for tracking? |
lukesneeringer
commented
Jan 8, 2018
theacodes
commented
Jan 8, 2018
@lukesneeringer (cc @dhermes) so I've ran into something and I really don't know how to fix it. The protos contain docstrings that are incompatible with autodoc. Source string here: https://github.com/GoogleCloudPlatform/google-cloud-python/blob/a44f6ceff101b338ae5c3e1718b378887ec6fd12/spanner/google/cloud/spanner_v1/proto/transaction_pb2.py#L372 This isn't something I can easily just edit back in place - there's entire prose docs in the proto. What should we do here? |
dhermes
commented
Jan 8, 2018
I've run into this before and just edited the |
I this case I'd need to remove the entire thing. |
I actually copied that entire doc to an entirely-separate RST file, replaced that huge section of the docstring with a reference, and gave Spanner a bug to please change it (internal: 65243734). |
theacodes
commented
Jan 8, 2018
Thanks, @lukesneeringer. We should be good to review and merge after 8e11b2d |
theacodes
commented
Jan 9, 2018
@lukesneeringer gentle ping - this should be ready to merge. |
lukesneeringer
commented
Jan 10, 2018
Agreed. |
No description provided.