add example of creating AutoML Tables client with non-default endpoint ('new' sdk)

[amygdala]

Torry, this is related to b/146659153. @sirtorry
I may not be using the right conventions for this sample code— LMK. (also not sure I'm doing the right things w.r.t. the tests...)

Edit: also, can you share what linter to run? It looks like you all require 4 spaces, not 2, for one thing.

/assign @sirtorry

[dinagraves]

Hi Amy - try running black on your files to fix your linting. ie

python -m black .

[telpirion]

@sirtorry Correct me if I'm wrong, but I think we want to consolidate all of our AutoML samples into automl/.

[amygdala]

@sirtorry Correct me if I'm wrong, but I think we want to consolidate all of our AutoML samples into automl/.

I was just following what appeared to be the convention for the location of the 'new sdk' examples— LMK.
Edit: if you have decided to move these examples, there are a number of other files in this dir linked to from the docs too, so I'd suggest pulling this in once it seems okay, then tackling the move separately. (to coordinate the docs updates).

[leahecole]

Would it be possible to have the create_client_with_endpoint method print something, and then assert on your print? Then, you would be testing your sample, rather than the actual AutoML underlying function.

Also, if testing is something you're not 100% comfortable with, do feel free to put some time on my calendar and we can take 20-30 min to have a faster back and forth about what you think is most appropriate for this sample :)

[telpirion]

+1 to @leahecole 's suggestion.

For the set endpoint kind of samples, we typically use a really simple API call to verify that the client was created and operates correctly.

See example here.

I suggest copying the output of the regular Tables list_datasets sample.

[amygdala]

In this case, no need to check anything re: the 'list_datasets' output specifics. There may or may not be any defined, but the details aren't relevant here. (I picked that call arbitrarily as a simple connectivity check, so I don't think that sample is particularly relevant).

In fact, I think we can remove an assertion on the output altogether-- if the call fails, it will throw an exception, which will fail the test. Would you prefer the test explicitly catch the exception + then raise an error, or just let it be thrown?

[busunkim96]

+1 to just executing list_datasets since the goal is to check that the client was constructed correctly.

[leahecole]

@sirtorry Correct me if I'm wrong, but I think we want to consolidate all of our AutoML samples into automl/.

I was just following what appeared to be the convention for the location of the 'new sdk' examples— LMK.
Edit: if you have decided to move these examples, there are a number of other files in this dir linked to from the docs too, so I'd suggest pulling this in once it seems okay, then tackling the move separately. (to coordinate the docs updates).

+1 to separate move

[telpirion]

I'm fine with moving these samples as part of a separate PR.

[gcf-merge-on-green]

Your PR has attempted to merge for 3 hours. Please check that all required checks have passed, you have an automerge label, and that all your reviewers have approved the PR

[gcf-merge-on-green]

Your PR has attempted to merge for 3 hours. Please check that all required checks have passed, you have an automerge label, and that all your reviewers have approved the PR

[gcf-merge-on-green]

Merge-on-green attempted to merge your PR for 6 hours, but it was not mergeable because either one of your required status checks failed, or one of your required reviews was not approved. Learn more about your required status checks here: https://help.github.com/en/github/administering-a-repository/enabling-required-status-checks. You can remove and reapply the label to re-run the bot.