Skip gcs integration tests to avoid egress fees

[bendnorman]

We got hit with some big egress fees for the gs://intake.catalyst.coop bucket in the last two days. We are pretty sure this is because the CI runs the integration tests 8 times for each push to a PR. 0.12 $/GB 8 runs 4 downloads of the epacems data * 5 GB (size of EPA cems data) = ~$20 per push!! That's no good.

This PR skips the GCS tests until we figure out how to download partitions of the data using intake. We didn't disable the s3 tests because AWS covers the egress fees and we'd still like to test our catalogs! We need to figure out how to pull partitions so our CI isn't downloading 160 GB of CEMS data every time there is a push.

I'm still not sure why we weren’t hit with big egress fees earlier in the catalog development.

[codecov[bot]]

Codecov Report

Base: 100.0% // Head: 100.0% // No change to project coverage :thumbsup:

Coverage data is based on head (338323b) compared to base (e4ad1ad). Patch has no changes to coverable lines.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## dev #81 +/- ## ======================================= Coverage 100.0% 100.0% ======================================= Files 2 2 Lines 44 44 ======================================= Hits 44 44 ``` Help us with your feedback. Take ten seconds to tell us [how you rate us](https://about.codecov.io/nps?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=catalyst-cooperative). Have a feature suggestion? [Share it here.](https://app.codecov.io/gh/feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=catalyst-cooperative)

:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.

[bendnorman]

I reran one of the 3.11 tox actions and the bucket still sent 1 - 4 GBs. I went over our tests a few times and I don't see where we are requesting the GCS bucket!?! The default PUDL_INTAKE_PATH is s3 and it's skipping the GCS tests...

[bendnorman]

The CI is no longer making requests to the GCS bucket so we won't be getting hit with large egress fees whenever we want to run the CI.

I'm still not sure the CI wasn't pulling dozens of GBs of data in the past but these are my ideas:

• The intake cache disabling workaround might have only disabled the cache for the first test so the CI only requested 4 GB of cems data. Now that we've dropped support for cache disabling, each integration test pulls the data to a temporary directory. The data is requested 4 times for each cloud provider so each tox run requests ~20 GB of data. We are still requesting 20 GB of data from the s3 bucket but the gcs tests have been disabled.
• Maybe the github runners were caching the data? I'm not super familiar with how caching works on github actions.

[zaneselvans]

I looked at the tox-pytest workflow and it wasn't doing github runner caching, so that isn't it. So weird.