feat: Inspect columns

docs/customization/querying-columns.md

@@ -0,0 +1,16 @@
+# Querying columns with SQL
+
+The model `stg_columns` ([source](https://github.com/dbt-labs/dbt-project-evaluator/tree/main/models/staging/graph/stg_columns.sql)), created with the package, lists all the columns from all the dbt nodes (models, sources, tests, snapshots)
+
+You can use this model to help with questions such as:
+
+- Are there columns with the same name in different nodes?
+- Do any columns in the YAML configuration lack descriptions?
+- Do any columns share the same name but have different descriptions?
+- Are there columns with names that match a specific pattern (regex)?
+- Have any prohibited names been used for columns?
+
+
+## Defining additional tests that match your exact requirements
+
+You can create a custom test against  `{{ ref(stg_columns) }}` to test for your specific check! When running the package you'd need to make sure to also include children of the package's models by using the `package:dbt_project_evalutator+` selector.

integration_tests/models/dbt_project_evaluator_schema_tests/core.yml

@@ -9,7 +9,7 @@ models:
         tests:
           - unique
           - not_null 
-          
+
   - name: int_all_graph_resources
     description: "This table shows one record for each enabled resource in the graph and information about that resource."
     columns:
@@ -34,4 +34,4 @@ models:
       - name: unique_id
         tests:
           - unique
-          - not_null
+          - not_null

integration_tests/models/dbt_project_evaluator_schema_tests/graph.yml

@@ -49,6 +49,14 @@ models:
           - unique
           - not_null
 
+  - name: stg_columns
+    description: "Staging model from the graph variable, one record per column resource."
+    tests:
+      - dbt_utils.unique_combination_of_columns:
+          combination_of_columns:
+            - node_unique_id
+            - name
+
   - name: stg_sources
     description: "Staging model from the graph variable, one record per source resource."
     columns:

macros/insert_resources_from_graph.sql

@@ -1,5 +1,5 @@
-{% macro insert_resources_from_graph(relation, resource_type='nodes', relationships=False, batch_size=var('insert_batch_size') | int) %}
-  {%- set values = get_resource_values(resource_type, relationships) -%}
+{% macro insert_resources_from_graph(relation, resource_type='nodes', relationships=False, columns=False, batch_size=var('insert_batch_size') | int) %}
+  {%- set values = get_resource_values(resource_type, relationships, columns) -%}
   {%- set values_length = values | length -%}
   {%- set loop_count = (values_length / batch_size) | round(0, 'ceil') | int -%}
 

macros/unpack/get_column_values.sql

@@ -0,0 +1,39 @@
+{%- macro get_column_values(node_type) -%}
+    {{ return(adapter.dispatch('get_column_values', 'dbt_project_evaluator')(node_type)) }}
+{%- endmacro -%}
+
+{%- macro default__get_column_values(node_type) -%}
+
+    {%- if execute -%}
+        {%- if node_type == 'nodes' %}
+            {% set nodes_list = graph.nodes.values() %}   
+        {%- elif node_type == 'sources' -%}
+            {% set nodes_list = graph.sources.values() %}
+        {%- else -%}
+            {{ exceptions.raise_compiler_error("node_type needs to be either nodes or sources, got " ~ node_type) }}
+        {% endif -%}
+
+        {%- set values = [] -%}
+
+        {%- for node in nodes_list -%}
+            {%- for column in node.columns.values() -%}
+
+                {%- set values_line  = 
+                    [
+                        wrap_string_with_quotes(node.unique_id),
+                        wrap_string_with_quotes(dbt.escape_single_quotes(column.name)),
+                        wrap_string_with_quotes(dbt.escape_single_quotes(column.description)),
+                        wrap_string_with_quotes(dbt.escape_single_quotes(column.data_type)),
+                        wrap_string_with_quotes(dbt.escape_single_quotes(column.quote))
+                    ]
+                %}
+
+                {%- do values.append(values_line) -%}
+
+            {%- endfor -%}
+        {%- endfor -%}
+    {{ return(values) }}
+
+    {%- endif -%}
+
+{%- endmacro -%}

macros/unpack/get_resource_values.sql

@@ -1,6 +1,8 @@
-{% macro get_resource_values(resource=None, relationships=None) %}
+{% macro get_resource_values(resource=None, relationships=None, columns=None) %}
   {% if relationships %}
     {{ return(adapter.dispatch('get_relationship_values', 'dbt_project_evaluator')(node_type=resource)) }}
+  {% elif columns %}
+    {{ return(adapter.dispatch('get_column_values', 'dbt_project_evaluator')(node_type=resource)) }}
   {% elif resource == 'exposures' %}
     {{ return(adapter.dispatch('get_exposure_values', 'dbt_project_evaluator')()) }}
   {% elif resource == 'sources' %}

mkdocs.yml

@@ -81,6 +81,7 @@ nav:
     - Configuring exceptions to the rules: customization/exceptions.md
     - Excluding packages and models/sources based on path: customization/excluding-packages-and-paths.md
     - Display issues in the logs: customization/issues-in-log.md
+    - Querying columns: customization/querying-columns.md
   - Run in CI Check: ci-check.md
   - Querying the DAG: querying-the-dag.md
   - Contributing: contributing.md

models/staging/graph/base/base_node_columns.sql

@@ -0,0 +1,24 @@
+{{
+    config(
+        materialized='table',
+        post_hook="{{ insert_resources_from_graph(this, resource_type='nodes', columns=True) }}"
+    )
+}}
+
+{% if execute %}
+    {{ check_model_is_table(model) }}
+{% endif %}
+/* Bigquery won't let us `where` without `from` so we use this workaround */
+with dummy_cte as (
+    select 1 as foo
+) 
+
+select 
+    cast(null as {{ dbt.type_string() }}) as node_unique_id,
+    cast(null as {{ dbt.type_string()}}) as name,
+    cast(null as {{ dbt.type_string()}}) as description,
+    cast(null as {{ dbt.type_string()}}) as data_type,
+    cast(null as {{ dbt.type_string()}}) as quote
+
+from dummy_cte
+where false

models/staging/graph/base/base_source_columns.sql

@@ -0,0 +1,24 @@
+{{
+    config(
+        materialized='table',
+        post_hook="{{ insert_resources_from_graph(this, resource_type='sources', columns=True) }}"
+    )
+}}
+
+{% if execute %}
+    {{ check_model_is_table(model) }}
+{% endif %}
+/* Bigquery won't let us `where` without `from` so we use this workaround */
+with dummy_cte as (
+    select 1 as foo
+) 
+
+select 
+    cast(null as {{ dbt.type_string() }}) as node_unique_id,
+    cast(null as {{ dbt.type_string()}}) as name,
+    cast(null as {{ dbt.type_string()}}) as description,
+    cast(null as {{ dbt.type_string()}}) as data_type,
+    cast(null as {{ dbt.type_string()}}) as quote
+
+from dummy_cte
+where false

models/staging/graph/stg_columns.sql

@@ -0,0 +1,11 @@
+with
+
+final as (
+
+    {{ dbt_utils.union_relations([
+        ref('base_node_columns'),
+        ref('base_source_columns')
+    ])}}
+)
+
+select * from final