untested code! just to give a sense of what this looks like

Based on / copied from:

• dbt-postgres → dbt-redshift inheritance (relevant files)
• file diff between this repo and dbt-sqlserver
• setup.py from dbt-spark for reporting version

SQL Strategies

If you need a Jinja adapter macro to do something genuinely different from the implementation in dbt-sqlserver, it's as simple as reimplementing it with a synapse__ prefix instead.

Today, if you want a macro to do exactly the same thing, you still need to reimplement it, but you can just wrap the sqlserver version:

{% macro synapse__get_columns_in_query(select_sql) %}
  {{ return(sqlserver__get_columns_in_query(select_sql)) }}
{% endmacro %}

There's a change I'd like to make for dispatched macros (dbt-labs/dbt-core#2923) which will enable you to completely avoid reimplementing macros if the synapse__ version is the same as the sqlserver__ version. (I currently have it under the v0.20.0 milestone.)