use "direct" write for non-partitioned python model materializations

[colin-rogers-dbt]

resolves #1318

In order to support partitioned materializations in BQ python models we we switched from "direct" to "indirect" mode when writing model results in Dataproc back to BigQuery. As the naming implies "indirect" temporarily stages data in the provided GCS bucket. If a user has a retention policy on the bucket this will fail as the bucket won't allow Dataproc to delete these temp files as it goes.

This PR sidesteps that issue to ensure backwards compatibility with <1.7 created models by using "direct" write when a partitioned config is not provided.

Note: I have not added any new testing to cover the case where a user has set a retention policy. Ultimately I think this is an edge case we don't need to test against but we should document that a bucket retention policy cannot be used with a partitioned python model as a follow up

docs dbt-labs/docs.getdbt.com/#

Problem

Solution

Checklist

• I have read the contributing guide and understand what's expected of me
• I have run this code in development and it appears to resolve the stated issue
• This PR includes tests, or tests are not required/relevant for this PR
• This PR has no interface changes (e.g. macros, cli, logs, json artifacts, config files, adapter interface, etc) or this PR has already received feedback and approval from Product or DX

[VersusFacit]

one nit