Bigquery repeated data type (#236)

* Handle BigQuery repeated fields data_types

* include nested repated structs

* override repeated struct data_type with array

* Add trailing newline

* update changelog

* Update CHANGELOG.md

---------

Co-authored-by: Doug Beatty <[email protected]>

integration_tests/models/model_repeated.sql

@@ -0,0 +1,13 @@
+{% if target.type == "bigquery" %}
+
+    {#--- This exists to test the BigQuery-specific behavior requested in #190 -#}
+select
+  [1, 2] AS repeated_int,
+  [
+    STRUCT(1 as nested_int_field, [STRUCT("a" as string_field)] as nested_repeated_struct),
+    STRUCT(2 AS nested_int_field, [STRUCT("a" as string_field)] as nested_repeated_struct)
+  ] as repeated_struct
+
+{% else %}
+    select 1 as int_field
+{% endif %}

integration_tests/tests/test_generate_model_repeated_yaml.sql

@@ -0,0 +1,56 @@
+{% set raw_schema = generate_schema_name('raw_data') %}
+
+{% set actual_source_yaml = codegen.generate_model_yaml(
+    model_names=['model_repeated']
+  )
+%}
+
+{% if target.type == "bigquery" %}
+
+{% set expected_source_yaml %}
+version: 2
+
+models:
+  - name: model_repeated
+    description: ""
+    columns:
+      - name: repeated_int
+        data_type: array<int64>
+        description: ""
+
+      - name: repeated_struct
+        data_type: array
+        description: ""
+
+      - name: repeated_struct.nested_int_field
+        data_type: int64
+        description: ""
+
+      - name: repeated_struct.nested_repeated_struct
+        data_type: array
+        description: ""
+
+      - name: repeated_struct.nested_repeated_struct.string_field
+        data_type: string
+        description: ""
+
+{% endset %}
+
+{% else %}
+
+{% set expected_source_yaml %}
+version: 2
+
+models:
+  - name: model_repeated
+    description: ""
+    columns:
+      - name: int_field
+        data_type: {{ integer_type_value() }}
+        description: ""
+
+{% endset %}
+
+{% endif %}
+
+{{ assert_equal (actual_source_yaml | trim, expected_source_yaml | trim) }}

integration_tests/tests/test_helper_get_models.sql

@@ -7,6 +7,6 @@
 {% set actual_list = codegen.get_models(prefix='model_')|sort %}
 {% endif %}
 
-{% set expected_list = ['model_data_a', 'model_from_source', 'model_struct', 'model_without_any_ctes', 'model_without_import_ctes'] %}
+{% set expected_list = ['model_data_a', 'model_from_source', 'model_repeated', 'model_struct', 'model_without_any_ctes', 'model_without_import_ctes'] %}
 
 {{ assert_equal (actual_list, expected_list) }}

macros/vendored/dbt_core/format_column.sql

@@ -1,5 +1,21 @@
 {% macro format_column(column) -%}
+  {{ return(adapter.dispatch('format_column', 'codegen')(column)) }}
+{%- endmacro %}
+
+{# Vendored from: https://github.com/dbt-labs/dbt-adapters/blob/c7b12aee533184bad391a657d1753539d1dd496a/dbt/include/global_project/macros/relations/column/columns_spec_ddl.sql#L85-L89 #}
+{% macro default__format_column(column) -%}
   {% set data_type = column.dtype %}
   {% set formatted = column.column.lower() ~ " " ~ data_type %}
   {{ return({'name': column.name, 'data_type': data_type, 'formatted': formatted}) }}
 {%- endmacro -%}
+
+{# Vendored from: https://github.com/dbt-labs/dbt-bigquery/blob/4d255b2f854d21d5d8871bdaa8d7ab47e7e863a3/dbt/include/bigquery/macros/utils/get_columns_spec_ddl.sql#L1-L5 #}
+{# But modified to handle https://github.com/dbt-labs/dbt-codegen/issues/190 #}
+{% macro bigquery__format_column(column) -%}
+  {% set data_type = column.data_type %}
+  {% if column.mode.lower() == "repeated" and column.dtype.lower() == "record" %}
+    {% set data_type = "array" %}
+  {% endif %}
+  {% set formatted = column.column.lower() ~ " " ~ data_type %}
+  {{ return({'name': column.name, 'data_type': data_type, 'formatted': formatted}) }}
+{%- endmacro -%}