[Bug]: test "failures" has a misleading definition

[Stephen986]

Overview

The definition of "failures" is misleading: https://github.com/brooklyn-data/dbt_artifacts/blob/main/models/docs.md?plain=1#L184. It makes it sound like a field that is either 1 or 0, but failures can be much higher than 0 if multiple records fail the test.

How to reproduce

1. Read the docs on "failures"
2. Query your dbt artifacts data:
3. SELECT * FROM dbt-artifacts.mart__dbt_artifacts.fct_dbt__test_executions where failures > 1 LIMIT 10;

Expected behaviour

The docs should indicate that failures can be greater than 0. I'm not sure the best possible description, but maybe something like: "A single test execution can have multiple failures if multiple rows had incorrect results."

Environment

Results of running dbt --version:

1.6

[jared-rimmer]

@Stephen986 happy to accept a PR with a definition you propose