Skip to main content

About adapter object

Your database communicates with dbt using an internal database adapter object. For example, BaseAdapter and SnowflakeAdapter. The Jinja object adapter is a wrapper around this internal database adapter object.

adapter grants the ability to invoke adapter methods of that internal class via:

  • {% do adapter.<method name> %} -- invoke internal adapter method
  • {{ adapter.<method name> }} -- invoke internal adapter method and capture its return value for use in materialization or other macros

For example, the adapter methods below will be translated into specific SQL statements depending on the type of adapter your project is using:

Deprecated adapter functions​

The following adapter functions are deprecated, and will be removed in a future release.

dispatch​

Moved to separate page: dispatch

get_missing_columns​

Args:

Returns a list of Columns that is the difference of the columns in the from_table and the columns in the to_table, i.e. (set(from_relation.columns) - set(to_table.columns)). Useful for detecting new columns in a source table.

Usage:

models/example.sql
{%- set target_relation = api.Relation.create(
      database='database_name',
      schema='schema_name',
      identifier='table_name') -%}


{% for col in adapter.get_missing_columns(target_relation, this) %}
  alter table {{this}} add column "{{col.name}}" {{col.data_type}};
{% endfor %}

expand_target_column_types​

Args:

  • from_relation: The source Relation to use as a template
  • to_relation: The Relation to mutate

Expand the to_relation table's column types to match the schema of from_relation. Column expansion is constrained to string and numeric types on supported databases. Typical usage involves expanding column types (from eg. varchar(16) to varchar(32)) to support insert statements.

Usage:

example.sql
{% set tmp_relation = adapter.get_relation(...) %}
{% set target_relation = adapter.get_relation(...) %}

{% do adapter.expand_target_column_types(tmp_relation, target_relation) %}

get_relation​

Args:

  • database: The database of the relation to fetch
  • schema: The schema of the relation to fetch
  • identifier: The identifier of the relation to fetch

Returns a cached Relation object identified by the database.schema.identifier provided to the method, or None if the relation does not exist.

Usage:

example.sql

{%- set source_relation = adapter.get_relation(
      database="analytics",
      schema="dbt_drew",
      identifier="orders") -%}

{{ log("Source Relation: " ~ source_relation, info=true) }}

load_relation​

Args:

A convenience wrapper for get_relation. Returns the cached version of the Relation object, or None if the relation does not exist.

Usage:

example.sql

{% set relation_exists = load_relation(ref('my_model')) is not none %}
{% if relation_exists %}
      {{ log("my_model has already been built", info=true) }}
{% else %}
      {{ log("my_model doesn't exist in the warehouse. Maybe it was dropped?", info=true) }}
{% endif %}

get_columns_in_relation​

Args:

  • relation: The Relation to find the columns for

Returns a list of Columns in a table.

Usage:

example.sql

{%- set columns = adapter.get_columns_in_relation(this) -%}

{% for column in columns %}
  {{ log("Column: " ~ column, info=true) }}
{% endfor %}

create_schema​

Args:

  • relation: A relation object with the database and schema to create. Any identifier on the relation will be ignored.

Creates a schema (or equivalent) in the target database. If the target schema already exists, then this method is a no-op.

Usage:

example.sql

{% do adapter.create_schema(api.Relation.create(database=target.database, schema="my_schema")) %}

drop_schema​

Args:

  • relation: A relation object with the database and schema to drop. Any identifier on the relation will be ignored.

Drops a schema (or equivalent) in the target database. If the target schema does not exist, then this method is a no-op. The specific implementation is adapter-dependent, but adapters should implement a cascading drop, such that objects in the schema are also dropped. Note: this adapter method is destructive, so please use it with care!

Usage:

example.sql

{% do adapter.drop_schema(api.Relation.create(database=target.database, schema="my_schema"))) %}

drop_relation​

Args:

  • relation: The Relation to drop

Drops a Relation in the database. If the target relation does not exist, then this method is a no-op. The specific implementation is adapter-dependent, but adapters should implement a cascading drop, such that bound views downstream of the dropped relation are also dropped. Note: this adapter method is destructive, so please use it with care!

The drop_relation method will remove the specified relation from dbt's relation cache.

Usage:

example.sql

{% do adapter.drop_relation(this) %}

rename_relation​

Args:

  • from_relation: The Relation to rename
  • to_relation: The destination Relation to rename from_relation to

Renames a Relation the database. The rename_relation method will rename the specified relation in dbt's relation cache.

Usage:

example.sql

{%- set old_relation = adapter.get_relation(
      database=this.database,
      schema=this.schema,
      identifier=this.identifier) -%}

{%- set backup_relation = adapter.get_relation(
      database=this.database,
      schema=this.schema,
      identifier=this.identifier ~ "__dbt_backup") -%}

{% do adapter.rename_relation(old_relation, backup_relation) %}

quote​

Args:

  • identifier: A string to quote

Encloses identifier in the correct quotes for the adapter when escaping reserved column names etc.

Usage:

example.sql
select 
      'abc' as {{ adapter.quote('table_name') }},
      'def' as {{ adapter.quote('group by') }}

get_columns_in_table​

Deprecated

This method is deprecated and will be removed in a future release. Please use get_columns_in_relation instead.

Args:

  • schema_name: The schema to test
  • table_name: The table (or view) from which to select columns

Returns a list of Columns in a table.

models/example.sql
{% set dest_columns = adapter.get_columns_in_table(schema, identifier) %}
{% set dest_cols_csv = dest_columns | map(attribute='quoted') | join(', ') %}

insert into {{ this }} ({{ dest_cols_csv }}) (
  select {{ dest_cols_csv }}
  from {{ref('another_table')}}
);

already_exists​

Deprecated

This method is deprecated and will be removed in a future release. Please use get_relation instead.

Args:

  • schema: The schema to test
  • table: The relation to look for

Returns true if a relation named like table exists in schema schema, false otherwise.

models/example.sql
select * from {{ref('raw_table')}}

{% if adapter.already_exists(this.schema, this.name) %}
  where id > (select max(id) from {{this}})
{% endif %}

adapter_macro​

Deprecated

This method is deprecated and will be removed in a future release. Please use adapter.dispatch instead.

Prior to v0.18.0, dbt supported a limited version of dispatch functionality via a macro named adapter_macro.

Args:

  • name: name of macro to implement
  • *args
  • **kwargs

Finds an adapter-appropriate version of a named macro and implements it with the positional and/or keyword arguments provided. This is most relevant for macros in open-source packages with cross-database support.

Usage:

macros/concat.sql
{% macro concat(fields) -%}
  {{ adapter_macro('concat', fields) }}
{%- endmacro %}


{% macro default__concat(fields) -%}
    concat({{ fields|join(', ') }})
{%- endmacro %}


{% macro redshift__concat(fields) %}
    {{ fields|join(' || ') }}
{% endmacro %}


{% macro snowflake__concat(fields) %}
    {{ fields|join(' || ') }}
{% endmacro %}
0
Edit this page