Support Poetry Config

[dehume]

Support setting the dbt version from the poetry configuration if set. Decided to just install their poetry config via the Github action rather than just parsing out the dbt version. Using the poetry.lock seems a little messy and while poetry does support getting information about individual packages, you would still need to parse the output. So something like poetry show dbt-snowflake would produce:

 name         : dbt-snowflake
 version      : 1.1.0
 description  : The Snowflake adapter plugin for dbt

dependencies
 - cryptography >=3.2,<4
 - dbt-core >=1.1.0,<1.2.0
 - requests <3.0.0
 - snowflake-connector-python >=2.4.1,<2.8.0

[dehume]

Removed creating and entering the virtual env which didn't seem necessary. For the poetry setup, those dependencies will also be installed outside of a virtual environment so the base dbt functions will behave the same.

[bobbyiliev]

I think that I added this initially as I was seeing some warnings suggesting a venv

[dehume]

Don't love this order of check if DBT_MZ_VERSION is set, then check POETRY_VERSION and then finally use latest DBT_MZ_VERSION.

[bobbyiliev]

I think that I added this initially as I was seeing some warnings suggesting a venv

[morsapaes]

Shouldn't this be conditional on whether POETRY_VERSION is set? It also looks like pyproject.toml and poetry.lock shouldn't be committed to the repo outside of /tests.

[dehume]

@bobbyiliev not sure on the best location for these files. I was treating these like the profiles.yml.template as an example of what it might look like in the repo being used. Though true the COPY should probably be conditional.

[morsapaes]

Suggested change

	| `POETRY_VERSION` | Poetry version if dbt project is managed with poetry. Cannot be used with `DBT_MZ_VERSION` | |
	| `POETRY_VERSION` | Poetry version, if the dbt project is managed with [Poetry](https://python-poetry.org/). Exclusive with `DBT_MZ_VERSION`. | |

[morsapaes]

Wouldn't the Poetry version also be available in the user's pyproject.toml? Sounds a bit like a chicken and egg problem. I'm wondering if we could go from that file and pip install . instead of requiring users to set the Poetry version on the side. IIUC, this would not install dev dependencies by default (see Poetry #3514), but dependency management seems to have changed quite a bit in 1.2.0.