diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..e7f63fa --- /dev/null +++ b/.gitignore @@ -0,0 +1,18 @@ +target/ +dbt_modules/ +# newer versions of dbt use this directory instead of dbt_modules for test dependencies +dbt_packages/ +logs/ + +.venv/ +.python-version + +# Visual Studio Code files +*/.vscode +*.code-workspace +.history/ +**/.DS_Store +.vscode/ +.env +.DS_Store +.user.yml \ No newline at end of file diff --git a/README.md b/README.md new file mode 100644 index 0000000..6e05bc8 --- /dev/null +++ b/README.md @@ -0,0 +1,75 @@ + +## Profile Set Up + +#### Use the following within profiles.yml +---- + +```yml +movement: + target: dev + outputs: + dev: + type: snowflake + account: + role: + user: + password: + region: + database: movement_dev + warehouse: + schema: silver + threads: 4 + client_session_keep_alive: False + query_tag: +``` + +### Resources: +- Learn more about dbt [in the docs](https://docs.getdbt.com/docs/introduction) +- Check out [Discourse](https://discourse.getdbt.com/) for commonly asked questions and answers +- Join the [chat](https://community.getdbt.com/) on Slack for live discussions and support +- Find [dbt events](https://events.getdbt.com) near you +- Check out [the blog](https://blog.getdbt.com/) for the latest news on dbt's development and best practices + +- Check out [the blog](https://blog.getdbt.com/) for the latest news on dbt's development and best practices + +## Applying Model Tags + +### Database / Schema level tags + +Database and schema tags are applied via the `add_database_or_schema_tags` macro. These tags are inherited by their downstream objects. To add/modify tags call the appropriate tag set function within the macro. + +``` +{{ set_database_tag_value('SOME_DATABASE_TAG_KEY','SOME_DATABASE_TAG_VALUE') }} +{{ set_schema_tag_value('SOME_SCHEMA_TAG_KEY','SOME_SCHEMA_TAG_VALUE') }} +``` + +### Model tags + +To add/update a model's snowflake tags, add/modify the `meta` model property under `config`. Only table level tags are supported at this time via DBT. + +``` +{{ config( + ..., + meta={ + 'database_tags':{ + 'table': { + 'PURPOSE': 'SOME_PURPOSE' + } + } + }, + ... +) }} +``` + +By default, model tags are not pushed to snowflake on each load. You can push a tag update for a model by specifying the `UPDATE_SNOWFLAKE_TAGS` project variable during a run. + +``` +dbt run --var '{"UPDATE_SNOWFLAKE_TAGS":True}' -s models/core/core__fact_swaps.sql +``` + +### Querying for existing tags on a model in snowflake + +``` +select * +from table(movement.information_schema.tag_references('movement.core.fact_blocks', 'table')); +``` diff --git a/analyses/.gitkeep b/analyses/.gitkeep new file mode 100644 index 0000000..e69de29 diff --git a/data/.gitkeep b/data/.gitkeep new file mode 100644 index 0000000..e69de29 diff --git a/data/github_actions__workflows.csv b/data/github_actions__workflows.csv new file mode 100644 index 0000000..32c758c --- /dev/null +++ b/data/github_actions__workflows.csv @@ -0,0 +1,6 @@ +workflow_name,workflow_schedule +dbt_run_streamline_blocks_tx_realtime,"0,20,40 * * * *" +dbt_run_streamline_transactions_realtime,"15,55 * * * *" +dbt_run_incremental_core,"15,35,55 * * * *" +dbt_test_intraday,"20 */4 * * *" +dbt_test_tasks,"0,30 * * * *" \ No newline at end of file diff --git a/dbt_project.yml b/dbt_project.yml new file mode 100644 index 0000000..c48e7c3 --- /dev/null +++ b/dbt_project.yml @@ -0,0 +1,113 @@ +# Name your project! Project names should contain only lowercase characters +# and underscores. A good package name should reflect your organization's +# name or the intended use of these models +name: "movement_models" +version: "1.0.0" +config-version: 2 + +require-dbt-version: ">=1.9.0" + +# This setting configures which "profile" dbt uses for this project. +profile: "movement" + +# These configurations specify where dbt should look for different types of files. +# The `source-paths` config, for example, states that models in this project can be +# found in the "models/" directory. You probably won't need to change these! +model-paths: ["models"] +analysis-paths: ["analysis"] +test-paths: ["tests"] +seed-paths: ["data"] +macro-paths: ["macros"] +snapshot-paths: ["snapshots"] + +target-path: "target" # directory which will store compiled SQL files +clean-targets: # directories to be removed by `dbt clean` + - "target" + - "dbt_modules" + - "dbt_packages" + +tests: + +store_failures: true # all tests + +on-run-start: + - "{{ create_sps() }}" + - "{{ create_udfs() }}" + +on-run-end: + - '{{ apply_meta_as_tags(results) }}' + +dispatch: + - macro_namespace: dbt + search_order: + - movement-models + - dbt_snowflake_query_tags + - dbt + +query-comment: + comment: '{{ dbt_snowflake_query_tags.get_query_comment(node) }}' + append: true # Snowflake removes prefixed comments. + +# Configuring models +# Full documentation: https://docs.getdbt.com/docs/configuring-models + +models: + +copy_grants: true + +persist_docs: + relation: true + columns: true + +on_schema_change: "append_new_columns" + + movement_models: + gold: # This will apply to ALL models under the gold directory + +tests: + - dbt_utils.recency: + datepart: hour + field: BLOCK_TIMESTAMP + interval: 3 + severity: error + tags: ['recency'] + +# In this example config, we tell dbt to build all models in the example/ directory +# as tables. These settings can be overridden in the individual model files +# using the `{{ config(...) }}` macro. + +vars: + "dbt_date:time_zone": GMT + STREAMLINE_INVOKE_STREAMS: false + STREAMLINE_USE_DEV_FOR_EXTERNAL_TABLES: false + UPDATE_UDFS_AND_SPS: false + UPDATE_SNOWFLAKE_TAGS: True + OBSERV_FULL_TEST: false + START_GHA_TASKS: false + BRONZE_LOOKBACK_DAYS: '{{ env_var("BRONZE_LOOKBACK_DAYS", 3) }}' + +#### STREAMLINE 2.0 BEGIN #### + + API_INTEGRATION: '{{ var("config")[target.name]["API_INTEGRATION"] if var("config")[target.name] else var("config")["dev"]["API_INTEGRATION"] }}' + EXTERNAL_FUNCTION_URI: '{{ var("config")[target.name]["EXTERNAL_FUNCTION_URI"] if var("config")[target.name] else var("config")["dev"]["EXTERNAL_FUNCTION_URI"] }}' + ROLES: '{{ var("config")[target.name]["ROLES"] }}' + + config: + # The keys correspond to dbt profiles and are case sensitive + dev: + API_INTEGRATION: aws_movement_api_stg_v2 + EXTERNAL_FUNCTION_URI: qjj5rutl05.execute-api.us-east-1.amazonaws.com/stg/ + ROLES: + - AWS_LAMBDA_MOVEMENT_API + - INTERNAL_DEV + dev-2xl: + API_INTEGRATION: aws_movement_api_stg_v2 + EXTERNAL_FUNCTION_URI: qjj5rutl05.execute-api.us-east-1.amazonaws.com/stg/ + ROLES: + - AWS_LAMBDA_MOVEMENT_API + - INTERNAL_DEV + + prod: + API_INTEGRATION: aws_movement_api_prod_v2 + EXTERNAL_FUNCTION_URI: d0t060jjxf.execute-api.us-east-1.amazonaws.com/prod/ + ROLES: + - AWS_LAMBDA_MOVEMENT_API + - INTERNAL_DEV + - DBT_CLOUD_MOVEMENT + +#### STREAMLINE 2.0 END #### \ No newline at end of file diff --git a/docs/.gitkeep b/docs/.gitkeep new file mode 100644 index 0000000..e69de29 diff --git a/macros/.gitkeep b/macros/.gitkeep new file mode 100644 index 0000000..e69de29 diff --git a/macros/create_sps.sql b/macros/create_sps.sql new file mode 100644 index 0000000..9fa7c66 --- /dev/null +++ b/macros/create_sps.sql @@ -0,0 +1,6 @@ +{% macro create_sps() %} + {% if target.database == 'MOVEMENT' %} + CREATE schema IF NOT EXISTS _internal; +{{ sp_create_prod_clone('_internal') }}; + {% endif %} +{% endmacro %} diff --git a/macros/create_udfs.sql b/macros/create_udfs.sql new file mode 100644 index 0000000..3515bbf --- /dev/null +++ b/macros/create_udfs.sql @@ -0,0 +1,10 @@ +{% macro create_udfs() %} + {% if var("UPDATE_UDFS_AND_SPS") %} + {% set sql %} + CREATE schema if NOT EXISTS silver; +{{ create_udf_bulk_rest_api_v2() }} + + {% endset %} + {% do run_query(sql) %} + {% endif %} +{% endmacro %} diff --git a/macros/custom_naming_macros.sql b/macros/custom_naming_macros.sql new file mode 100644 index 0000000..cb6fc87 --- /dev/null +++ b/macros/custom_naming_macros.sql @@ -0,0 +1,23 @@ +{% macro generate_schema_name( + custom_schema_name = none, + node = none + ) -%} + {% set node_name = node.name %} + {% set split_name = node_name.split('__') %} + {{ split_name [0] | trim }} +{%- endmacro %} + +{% macro generate_alias_name( + custom_alias_name = none, + node = none + ) -%} + {% set node_name = node.name %} + {% set split_name = node_name.split('__') %} + {{ split_name [1] | trim }} +{%- endmacro %} + +{% macro generate_tmp_view_name(model_name) -%} + {% set node_name = model_name.name %} + {% set split_name = node_name.split('__') %} + {{ target.database ~ '.' ~ split_name[0] ~ '.' ~ split_name [1] ~ '__dbt_tmp' | trim }} +{%- endmacro %} diff --git a/macros/custom_query_tag.sql b/macros/custom_query_tag.sql new file mode 100644 index 0000000..809e1bb --- /dev/null +++ b/macros/custom_query_tag.sql @@ -0,0 +1,11 @@ +{% macro set_query_tag() -%} + {% set new_json = {"repo":project_name, "object":this.table, "profile":target.profile_name, "env":target.name, "existing_tag":get_current_query_tag() } %} +{% set new_query_tag = tojson(new_json) | as_text %} + {% if new_query_tag %} + {% set original_query_tag = get_current_query_tag() %} + {{ log("Setting query_tag to '" ~ new_query_tag ~ "'. Will reset to '" ~ original_query_tag ~ "' after materialization.") }} + {% do run_query("alter session set query_tag = '{}'".format(new_query_tag)) %} + {{ return(original_query_tag)}} + {% endif %} + {{ return(none)}} +{% endmacro %} \ No newline at end of file diff --git a/macros/dbt/get_merge_sql.sql b/macros/dbt/get_merge_sql.sql new file mode 100644 index 0000000..e1a8be2 --- /dev/null +++ b/macros/dbt/get_merge_sql.sql @@ -0,0 +1,16 @@ +{% macro get_merge_sql( + target, + source, + unique_key, + dest_columns, + incremental_predicates + ) -%} + {% set merge_sql = fsc_utils.get_merge_sql( + target, + source, + unique_key, + dest_columns, + incremental_predicates + ) %} + {{ return(merge_sql) }} +{% endmacro %} diff --git a/macros/dbt/get_tmp_relation_type.sql b/macros/dbt/get_tmp_relation_type.sql new file mode 100644 index 0000000..3bb7438 --- /dev/null +++ b/macros/dbt/get_tmp_relation_type.sql @@ -0,0 +1,8 @@ +{% macro dbt_snowflake_get_tmp_relation_type( + strategy, + unique_key, + language + ) %} + -- always table + {{ return('table') }} +{% endmacro %} diff --git a/macros/run_sp_create_prod_clone.sql b/macros/run_sp_create_prod_clone.sql new file mode 100644 index 0000000..b4a3d45 --- /dev/null +++ b/macros/run_sp_create_prod_clone.sql @@ -0,0 +1,10 @@ +{% macro run_sp_create_prod_clone() %} + {% set clone_query %} + call movement._internal.create_prod_clone( + 'movement', + 'movement_dev', + 'internal_dev' + ); +{% endset %} + {% do run_query(clone_query) %} +{% endmacro %} diff --git a/macros/sp_create_prod_clone.sql b/macros/sp_create_prod_clone.sql new file mode 100644 index 0000000..20ee897 --- /dev/null +++ b/macros/sp_create_prod_clone.sql @@ -0,0 +1,44 @@ +{% macro sp_create_prod_clone(target_schema) -%} + +create or replace procedure {{ target_schema }}.create_prod_clone(source_db_name string, destination_db_name string, role_name string) +returns boolean +language javascript +execute as caller +as +$$ + snowflake.execute({sqlText: `BEGIN TRANSACTION;`}); + try { + snowflake.execute({sqlText: `CREATE OR REPLACE DATABASE ${DESTINATION_DB_NAME} CLONE ${SOURCE_DB_NAME}`}); + snowflake.execute({sqlText: `DROP SCHEMA IF EXISTS ${DESTINATION_DB_NAME}._INTERNAL`}); /* this only needs to be in prod */ + + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL SCHEMAS IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL FUNCTIONS IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL PROCEDURES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL VIEWS IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL STAGES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON ALL TABLES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON FUTURE FUNCTIONS IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME};`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON FUTURE PROCEDURES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME};`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON FUTURE VIEWS IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME};`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON FUTURE STAGES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME};`}); + snowflake.execute({sqlText: `GRANT OWNERSHIP ON FUTURE TABLES IN DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME};`}); + + snowflake.execute({sqlText: `GRANT OWNERSHIP ON DATABASE ${DESTINATION_DB_NAME} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}) + + var existing_tags = snowflake.execute({sqlText: `SHOW TAGS IN DATABASE ${DESTINATION_DB_NAME};`}); + while (existing_tags.next()) { + var schema = existing_tags.getColumnValue(4); + var tag_name = existing_tags.getColumnValue(2) + snowflake.execute({sqlText: `GRANT OWNERSHIP ON TAG ${DESTINATION_DB_NAME}.${schema}.${tag_name} TO ROLE ${ROLE_NAME} COPY CURRENT GRANTS;`}); + } + + snowflake.execute({sqlText: `COMMIT;`}); + } catch (err) { + snowflake.execute({sqlText: `ROLLBACK;`}); + throw(err); + } + + return true +$$ + +{%- endmacro %} \ No newline at end of file diff --git a/macros/streamline/api_integrations.sql b/macros/streamline/api_integrations.sql new file mode 100644 index 0000000..2b1ee6a --- /dev/null +++ b/macros/streamline/api_integrations.sql @@ -0,0 +1,21 @@ +{% macro create_aws_movement_api() %} + {{ log( + "Creating integration for target:" ~ target + ) }} + + {% if target.name == "prod" %} + {% set sql %} + CREATE api integration IF NOT EXISTS aws_movement_api_prod_v2 api_provider = aws_api_gateway api_aws_role_arn = 'arn:aws:iam::924682671219:role/movement-api-prod-rolesnowflakeudfsAF733095-uYaV9o2uJLDU' api_allowed_prefixes = ( + 'https://d0t060jjxf.execute-api.us-east-1.amazonaws.com/prod/' + ) enabled = TRUE; +{% endset %} + {% do run_query(sql) %} + {% else %} + {% set sql %} + CREATE api integration IF NOT EXISTS aws_movement_api_stg_v2 api_provider = aws_api_gateway api_aws_role_arn = 'arn:aws:iam::704693948482:role/movement-api-stg-rolesnowflakeudfsAF733095-XfjrcIZRSZkR' api_allowed_prefixes = ( + 'https://qjj5rutl05.execute-api.us-east-1.amazonaws.com/stg/' + ) enabled = TRUE; +{% endset %} + {% do run_query(sql) %} + {% endif %} +{% endmacro %} diff --git a/macros/streamline/models.sql b/macros/streamline/models.sql new file mode 100644 index 0000000..503e680 --- /dev/null +++ b/macros/streamline/models.sql @@ -0,0 +1,70 @@ +{% macro streamline_external_table_query_v2( + model, + partition_function + ) %} + + {% set days = var("BRONZE_LOOKBACK_DAYS")%} + + WITH meta AS ( + SELECT + last_modified AS inserted_timestamp, + file_name, + {{ partition_function }} AS partition_key + FROM + TABLE( + information_schema.external_table_file_registration_history( + start_time => DATEADD('day', -ABS({{days}}), CURRENT_TIMESTAMP()), + table_name => '{{ source( "bronze_streamline", model) }}') + ) A + ) + SELECT + s.*, + b.file_name, + inserted_timestamp + FROM + {{ source( + "bronze_streamline", + model + ) }} + s + JOIN meta b + ON b.file_name = metadata$filename + AND b.partition_key = s.partition_key + WHERE + b.partition_key = s.partition_key + AND DATA :error IS NULL +{% endmacro %} + +{% macro streamline_external_table_FR_query_v2( + model, + partition_function + ) %} + WITH meta AS ( + SELECT + registered_on AS inserted_timestamp, + file_name, + {{ partition_function }} AS partition_key + FROM + TABLE( + information_schema.external_table_files( + table_name => '{{ source( "bronze_streamline", model) }}' + ) + ) A + ) +SELECT + s.*, + b.file_name, + inserted_timestamp +FROM + {{ source( + "bronze_streamline", + model + ) }} + s + JOIN meta b + ON b.file_name = metadata$filename + AND b.partition_key = s.partition_key +WHERE + b.partition_key = s.partition_key + AND DATA :error IS NULL +{% endmacro %} diff --git a/macros/streamline/streamline_udfs.sql b/macros/streamline/streamline_udfs.sql new file mode 100644 index 0000000..998cf6a --- /dev/null +++ b/macros/streamline/streamline_udfs.sql @@ -0,0 +1,10 @@ +{% macro create_udf_bulk_rest_api_v2() %} + CREATE + OR REPLACE EXTERNAL FUNCTION streamline.udf_bulk_rest_api_v2( + json OBJECT + ) returns ARRAY api_integration = {% if target.name == "prod" %} + aws_movement_api_prod_v2 AS 'https://d0t060jjxf.execute-api.us-east-1.amazonaws.com/prod/udf_bulk_rest_api' + {% else %} + aws_movement_api_stg_v2 AS 'https://qjj5rutl05.execute-api.us-east-1.amazonaws.com/stg/udf_bulk_rest_api' + {%- endif %}; +{% endmacro %} diff --git a/macros/tags/add_database_or_schema_tags.sql b/macros/tags/add_database_or_schema_tags.sql new file mode 100644 index 0000000..eb00e45 --- /dev/null +++ b/macros/tags/add_database_or_schema_tags.sql @@ -0,0 +1,6 @@ +{% macro add_database_or_schema_tags() %} + {{ set_database_tag_value( + 'BLOCKCHAIN_NAME', + 'movement' + ) }} +{% endmacro %} diff --git a/macros/tags/snowflake_tagging.sql b/macros/tags/snowflake_tagging.sql new file mode 100644 index 0000000..bc25e69 --- /dev/null +++ b/macros/tags/snowflake_tagging.sql @@ -0,0 +1,127 @@ +{% macro apply_meta_as_tags(results) %} + {% if var("UPDATE_SNOWFLAKE_TAGS") %} + {{ log('apply_meta_as_tags', info=False) }} + {{ log(results, info=False) }} + {% if execute %} + + {%- set tags_by_schema = {} -%} + {% for res in results -%} + {% if res.node.meta.database_tags %} + + {%- set model_database = res.node.database -%} + {%- set model_schema = res.node.schema -%} + {%- set model_schema_full = model_database+'.'+model_schema -%} + {%- set model_alias = res.node.alias -%} + + {% if model_schema_full not in tags_by_schema.keys() %} + {{ log('need to fetch tags for schema '+model_schema_full, info=False) }} + {%- call statement('main', fetch_result=True) -%} + show tags in {{model_database}}.{{model_schema}} + {%- endcall -%} + {%- set _ = tags_by_schema.update({model_schema_full: load_result('main')['table'].columns.get('name').values()|list}) -%} + {{ log('Added tags to cache', info=False) }} + {% else %} + {{ log('already have tag info for schema', info=False) }} + {% endif %} + + {%- set current_tags_in_schema = tags_by_schema[model_schema_full] -%} + {{ log('current_tags_in_schema:', info=False) }} + {{ log(current_tags_in_schema, info=False) }} + {{ log("========== Processing tags for "+model_schema_full+"."+model_alias+" ==========", info=False) }} + + {% set line -%} + node: {{ res.node.unique_id }}; status: {{ res.status }} (message: {{ res.message }}) + node full: {{ res.node}} + meta: {{ res.node.meta}} + materialized: {{ res.node.config.materialized }} + {%- endset %} + {{ log(line, info=False) }} + + {%- call statement('main', fetch_result=True) -%} + select LEVEL,UPPER(TAG_NAME) as TAG_NAME,TAG_VALUE from table(information_schema.tag_references_all_columns('{{model_schema}}.{{model_alias}}', 'table')) + {%- endcall -%} + {%- set existing_tags_for_table = load_result('main')['data'] -%} + {{ log('Existing tags for table:', info=False) }} + {{ log(existing_tags_for_table, info=False) }} + + {{ log('--', info=False) }} + {% for table_tag in res.node.meta.database_tags.table %} + + {{ create_tag_if_missing(current_tags_in_schema,table_tag|upper) }} + {% set desired_tag_value = res.node.meta.database_tags.table[table_tag] %} + + {{set_table_tag_value_if_different(model_schema,model_alias,table_tag,desired_tag_value,existing_tags_for_table)}} + {% endfor %} + {{ log("========== Finished processing tags for "+model_alias+" ==========", info=False) }} + {% endif %} + {% endfor %} + {% endif %} + {% endif %} +{% endmacro %} + + +{% macro create_tag_if_missing(all_tag_names,table_tag) %} + {% if table_tag not in all_tag_names %} + {{ log('Creating missing tag '+table_tag, info=False) }} + {%- call statement('main', fetch_result=True) -%} + create tag if not exists silver.{{table_tag}} + {%- endcall -%} + {{ log(load_result('main').data, info=False) }} + {% else %} + {{ log('Tag already exists: '+table_tag, info=False) }} + {% endif %} +{% endmacro %} + +{% macro set_table_tag_value_if_different(model_schema,table_name,tag_name,desired_tag_value,existing_tags) %} + {{ log('Ensuring tag '+tag_name+' has value '+desired_tag_value+' at table level', info=False) }} + {%- set existing_tag_for_table = existing_tags|selectattr('0','equalto','TABLE')|selectattr('1','equalto',tag_name|upper)|list -%} + {{ log('Filtered tags for table:', info=False) }} + {{ log(existing_tag_for_table[0], info=False) }} + {% if existing_tag_for_table|length > 0 and existing_tag_for_table[0][2]==desired_tag_value %} + {{ log('Correct tag value already exists', info=False) }} + {% else %} + {{ log('Setting tag value for '+tag_name+' to value '+desired_tag_value, info=False) }} + {%- call statement('main', fetch_result=True) -%} + alter table {{model_schema}}.{{table_name}} set tag {{tag_name}} = '{{desired_tag_value}}' + {%- endcall -%} + {{ log(load_result('main').data, info=False) }} + {% endif %} +{% endmacro %} + +{% macro set_column_tag_value_if_different(table_name,column_name,tag_name,desired_tag_value,existing_tags) %} + {{ log('Ensuring tag '+tag_name+' has value '+desired_tag_value+' at column level', info=False) }} + {%- set existing_tag_for_column = existing_tags|selectattr('0','equalto','COLUMN')|selectattr('1','equalto',tag_name|upper)|list -%} + {{ log('Filtered tags for column:', info=False) }} + {{ log(existing_tag_for_column[0], info=False) }} + {% if existing_tag_for_column|length > 0 and existing_tag_for_column[0][2]==desired_tag_value %} + {{ log('Correct tag value already exists', info=False) }} + {% else %} + {{ log('Setting tag value for '+tag_name+' to value '+desired_tag_value, info=False) }} + {%- call statement('main', fetch_result=True) -%} + alter table {{table_name}} modify column {{column_name}} set tag {{tag_name}} = '{{desired_tag_value}}' + {%- endcall -%} + {{ log(load_result('main').data, info=False) }} + {% endif %} +{% endmacro %} + +{% macro set_database_tag_value(tag_name,tag_value) %} + {% set query %} + create tag if not exists silver.{{tag_name}} + {% endset %} + {% do run_query(query) %} + {% set query %} + alter database {{target.database}} set tag {{target.database}}.silver.{{tag_name}} = '{{tag_value}}' + {% endset %} + {% do run_query(query) %} +{% endmacro %} + +{% macro set_schema_tag_value(target_schema,tag_name,tag_value) %} + {% set query %} + create tag if not exists silver.{{tag_name}} + {% endset %} + {% do run_query(query) %} + {% set query %} + alter schema {{target.database}}.{{target_schema}} set tag {{target.database}}.silver.{{tag_name}} = '{{tag_value}}' + {% endset %} + {% do run_query(query) %} +{% endmacro %} \ No newline at end of file diff --git a/macros/tests/compare_model_subset.sql b/macros/tests/compare_model_subset.sql new file mode 100644 index 0000000..18aa624 --- /dev/null +++ b/macros/tests/compare_model_subset.sql @@ -0,0 +1,29 @@ +{% test compare_model_subset(model, compare_model, compare_columns, model_condition) %} + +{% set compare_cols_csv = compare_columns | join(', ') %} + +with a as ( + select {{compare_cols_csv}} from {{ model }} + {{ model_condition }} +), +b as ( + select {{compare_cols_csv}} from {{ compare_model }} +), +a_minus_b as ( + select * from a + except + select * from b +), +b_minus_a as ( + select * from b + except + select * from a +), +unioned as ( + select 'in_actual_not_in_expected' as which_diff, a_minus_b.* from a_minus_b + union all + select 'in_expected_not_in_actual' as which_diff, b_minus_a.* from b_minus_a +) +select * from unioned + +{% endtest %} \ No newline at end of file diff --git a/macros/tests/sequence_gaps.sql b/macros/tests/sequence_gaps.sql new file mode 100644 index 0000000..84a9aa9 --- /dev/null +++ b/macros/tests/sequence_gaps.sql @@ -0,0 +1,37 @@ +{% macro sequence_gaps( + table, + partition_by, + column + ) %} + {%- set partition_sql = partition_by | join(", ") -%} + {%- set previous_column = "prev_" ~ column -%} + WITH source AS ( + SELECT + {{ partition_sql + "," if partition_sql }} + {{ column }}, + LAG( + {{ column }}, + 1 + ) over ( + {{ "PARTITION BY " ~ partition_sql if partition_sql }} + ORDER BY + {{ column }} ASC + ) AS {{ previous_column }} + FROM + {{ table }} + WHERE + block_timestamp::date <= current_date - 1 + ) +SELECT + {{ partition_sql + "," if partition_sql }} + {{ previous_column }}, + {{ column }}, + {{ column }} - {{ previous_column }} + - 1 AS gap +FROM + source +WHERE + {{ column }} - {{ previous_column }} <> 1 +ORDER BY + gap DESC +{% endmacro %} diff --git a/macros/tests/tx_gaps.sql b/macros/tests/tx_gaps.sql new file mode 100644 index 0000000..821a4d7 --- /dev/null +++ b/macros/tests/tx_gaps.sql @@ -0,0 +1,33 @@ +{% macro tx_gaps( + model + ) %} + WITH block_base AS ( + SELECT + block_number, + tx_count + FROM + {{ ref('silver__blocks') }} + ), + model_name AS ( + SELECT + block_number, + COUNT( + DISTINCT tx_id + ) AS model_tx_count + FROM + {{ model }} + GROUP BY + block_number + ) +SELECT + block_base.block_number, + tx_count, + model_name.block_number, + model_tx_count +FROM + block_base + LEFT JOIN model_name + ON block_base.block_number = model_name.block_number +WHERE + tx_count <> model_tx_count +{% endmacro %} diff --git a/macros/utils.sql b/macros/utils.sql new file mode 100644 index 0000000..2e3b3cb --- /dev/null +++ b/macros/utils.sql @@ -0,0 +1,42 @@ +{% macro if_data_call_wait() %} + {% if var( + "STREAMLINE_INVOKE_STREAMS" + ) %} + {% set query %} + SELECT + 1 + WHERE + EXISTS( + SELECT + 1 + FROM + {{ model.schema ~ "." ~ model.alias }} + LIMIT + 1 + ) {% endset %} + {% if execute %} + {% set results = run_query( + query + ) %} + {% if results %} + {{ log( + "Waiting...", + info = True + ) }} + + {% set wait_query %} + SELECT + system$wait( + {{ var( + "WAIT", + 600 + ) }} + ) {% endset %} + {% do run_query(wait_query) %} + {% else %} + SELECT + NULL; + {% endif %} + {% endif %} + {% endif %} +{% endmacro %} diff --git a/package-lock.yml b/package-lock.yml new file mode 100644 index 0000000..fc976d2 --- /dev/null +++ b/package-lock.yml @@ -0,0 +1,16 @@ +packages: +- package: calogica/dbt_expectations + version: 0.8.5 +- package: dbt-labs/dbt_utils + version: 1.0.0 +- git: https://github.com/FlipsideCrypto/fsc-utils.git + revision: eb33ac727af26ebc8a8cc9711d4a6ebc3790a107 +- package: get-select/dbt_snowflake_query_tags + version: 2.5.0 +- package: Snowflake-Labs/dbt_constraints + version: 0.6.3 +- package: calogica/dbt_date + version: 0.7.2 +- git: https://github.com/FlipsideCrypto/livequery-models.git + revision: b024188be4e9c6bc00ed77797ebdc92d351d620e +sha1_hash: d3219b9c206b5988189dcdafae0ec22ca9b4056c diff --git a/packages.yml b/packages.yml new file mode 100644 index 0000000..8001507 --- /dev/null +++ b/packages.yml @@ -0,0 +1,11 @@ +packages: + - package: calogica/dbt_expectations + version: [">=0.4.0", "<0.9.0"] + - package: dbt-labs/dbt_utils + version: 1.0.0 + - git: https://github.com/FlipsideCrypto/fsc-utils.git + revision: v1.32.0 + - package: get-select/dbt_snowflake_query_tags + version: [">=2.0.0", "<3.0.0"] + - package: Snowflake-Labs/dbt_constraints + version: [">=0.4.0"] \ No newline at end of file diff --git a/profiles.yml b/profiles.yml new file mode 100644 index 0000000..e287ff0 --- /dev/null +++ b/profiles.yml @@ -0,0 +1,29 @@ +movement: + target: dev + outputs: + dev: + type: snowflake + account: "{{ env_var('ACCOUNT') }}" + user: "{{ env_var('USER') }}" + password: "{{ env_var('PASSWORD') }}" + role: "{{ env_var('ROLE') }}" + schema: "{{ env_var('SCHEMA') }}" + region: "{{ env_var('REGION') }}" + database: "{{ env_var('DATABASE') }}" + warehouse: "{{ env_var('WAREHOUSE') }}" + threads: 8 + client_session_keep_alive: False + prod: + type: snowflake + account: "{{ env_var('ACCOUNT') }}" + user: "{{ env_var('USER') }}" + password: "{{ env_var('PASSWORD') }}" + role: "{{ env_var('ROLE') }}" + schema: "{{ env_var('SCHEMA') }}" + region: "{{ env_var('REGION') }}" + database: "{{ env_var('DATABASE') }}" + warehouse: "{{ env_var('WAREHOUSE') }}" + threads: 8 + client_session_keep_alive: False + config: + send_anonymous_usage_stats: False \ No newline at end of file diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 0000000..b6ec923 --- /dev/null +++ b/requirements.txt @@ -0,0 +1,3 @@ +dbt-core>=1.9,<1.10 +dbt-snowflake>=1.9,<1.10 +protobuf==4.25.3 diff --git a/snapshots/.gitkeep b/snapshots/.gitkeep new file mode 100644 index 0000000..e69de29