This was using the CLJS code at master, so type-checking fails if you
add new CLJS functions and use them from CLJS in the same PR.

* Migrate Mongo 5 ssl driver test to GitHub Actions

* Migrate Mongo 4 ssl driver test to GitHub Actions

* Make Mongo certificates available

* Try `docker exec` for Mongo 4.0

* Remove `TTY` from the `exec`

* Try overriding the entrypoint

* Include `mongod` in entrypoint

* Quote entrypoint command

* Quote only the command

* Omit `--dbpath`

* Try running the docker container manually

* Detach containers

* Reduce the timeout

* Remove Presto 186 config

We've already migrated it to GitHub actions.

* Migrate Presto JDBC driver test to GitHub Actions

* Wait for port to be ready

* Try `presto-jdbc` driver

* Wait for port before creating certificates

* Migrate BigQuery SDK driver test to GitHub Actions

* Migrate Google related classpath drivers test to GitHub Actions

* Multi-line strings are not working?

* Provide BQ envs

* Migrate Snowflake driver test to GitHub Actions

* Do not run test when PR is in a draft mode

* Dial down timeout to 60 minutes

* Migrate Google Analytics driver test to GitHub Actions

* Migrate Redshift driver test to GitHub Actions

* Migrate Athena driver test to GitHub Actions

* Migrate vertica driver test to GitHub Actions

* Bump runner for vertica

* Do not use matrix as it messes the output

* Add vertica port to the image config

* Go back to the official GitHub runner

This gets rid of the following GitHub Action warning message:

```
check-uberjar (17)
Node.js 12 actions are deprecated. For more information see:
https://github.blog/changelog/2022-09-22-github-actions-all-actions-will-begin-running-on-node16-instead-of-node12/.
Please update the following actions to use Node.js 16:
actions/download-artifact@v2
```

* Do not run driver tests when PR is in a draft mode

* Use the conditional directly on the job level

* Migrate MySQL latest driver test to GitHub Actions

* Increase the timeout for MySQL latest driver test

* Make env vars available to the host machine

* Increase the timeout again

* Bump timeout to 1h for all driver tests

* MongoDB test data config should not hardcode the user and password we use in CI

* Minor test data interface dox improvements

* Support standard deviation and variance for MongoDB

* Oops I got :expressions mixed up with :expression-aggregations

* Note about calculating variance

* `connection-type` instead of `db-or-server`

* Include Amazon Athena support for dacort/metabase-athena-driver

* Use Metabase Maven repo to fetch Athena driver

* Copy test extensions from Damon's Athena driver repo

* Some code cleanup.

* Move namespaces => metabase.driver.athena

* Clean up the test extensions namespace

* 42 failures, 16 errors

* Fix regex support; disable a few tests

* Minor tweaks

* Fix data-source-name

* Fix :week and :day-of-week impls (mostly): 2 failures, 3 errors

* Fix OFFSET, :week-of-year; skip test that has TIME column

* Add Athena to CircleCI config. Don't wait for Java 11 tests to finish before driver tests.

* ALL TESTS ARE PASSING! <3

* Copy fixes for https://github.com/dacort/metabase-athena-driver/issues/115;

 add test

* We don't need to prep source files or fetch dependencies before running backend tests.

* Fix Eastwood error.

* Tweak CircleCI config.

* Fix TIMESTAMP WITH TIME ZONE

* Include the Athena/Redshift repos in the build-drivers deps.edn

* Build and release scripts need to have the :mvn/repos as well

* Revert change that enabled test for Presto

* Un-enable failing test for Snowflake as well.

* moves all is clauses into the test

- previously only the first few tests were being run

* Sort ns in `metabase.driver.athena-test`

* Prevent athena's log4j2.properties file from becoming log config

athena includes log4j2.properties top-level with the properties:

```
status = debug
rootLogger.level=debug
```

And this kills our beautifully crafted logging. Set
"log4j2.configurationFile" in bootstrap to our own log4j2.xml.

Log4j2 looks in a few places for its logging config, the first of which
is the properties file. So when the jar is loaded, log4j2 considers this
a logging config change and we lose our logging.

Co-authored-by: Damon P. Cortesi <d.lifehacker@gmail.com>
Co-authored-by: Bryan Maass <bryan.maass@gmail.com>
Co-authored-by: dan sutton <dan@dpsutton.com>

* [CI] Trigger `uberjar` workflow manually

* Do not make commit SHA input required

* Continue building on push to `master`

This will keep `metabase-head` and `metabase-enterprise-head`
DockerHub repos working as they did before.

* Simplify by removing references to `main` branch

* Ignore unrelated paths for push on `master` branch

* Ignore only FE unit test changes in backend workflows

* Ignore both FE and E2E tests

* Remove `cy2` (SorryCypress library)

* Stop recording Cypress tests using Currents dashboard service

* Fix pre-release config typo

* Fix conditional containing environment variable

* Retrofit the release workflow to be useful for doing a pre-release

* Distinguish between `metabase` main repo and the custom ones

* Truncate commit id to specific hash length

We need to make sure all images are tagged with exactly the same hash length
in order to avoid having multiple images for the same commit.

* Enforce full commit SHA-1 hash is used for an input

`actions/checkout@v3` will fail if full SHA-1 hash is not provided.
We want to use this implicit mechanism as an enforcing function
for uniform manual inputs.

* Never run script from the main repo

- This greatly simplifies things.
- The script is meant to be run from forks of Metabase for testing purposes and/or for custom builds.

* Safeguard against missing `CUSTOM_RELEASE_REPO`

Co-authored-by: Nemanja <31325167+nemanjaglumac@users.noreply.github.com>

- run all EE tests against EE uberjar and only a small subset of OSS-specific tests against OSS uberjar

* Simplify filter glob pattern for frontend tests

Take a look at the cheat sheet:
https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#filter-pattern-cheat-sheet

'**/docs/**' matches `docs/hello.md` and `dir/docs/my-file.txt`

* Ignore unrelated paths for `backend` workflow even in `master`

* Ignore unrelated paths for `drivers` workflow even in `master`

* Ignore clojure tests in E2E workflows

* Increase the ignored paths scope for `uberjar` workflow

* [E2E] Fix grep spec filtering

* Add undocumented `grepIntegrationFolder`

* Migrate some driver tests to GitHub Actions

* Druid
* MariaDB 10.2
* MariaDB latest
* MongoDB 4.0
* MongoDB 5.0
* MongoDB latest
* MySQL 5.7
* Postgres 9.6
* Postgres latest
* Presto
* Spark
* SQLite
* SQL Server 2017

* Don't use Buildjet runner

* Don't run the driver tests for draft PRs

* Use Buildjet for Spark driver tests

* Use Buildjet for Druid driver tests