Merge pull request #771 from datafold/feature/pre-commit-automation

Add pre-commit automation for faster, safer development

Author: dlawin

.github/workflows/formatter.yml

@@ -7,20 +7,16 @@ on:
 
 jobs:
   linter_name:
-    name: runner / black
+    name: runner / ruff
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
-      - name: Check files using the black formatter
-        uses: rickstaa/action-black@v1
-        id: black_formatter
+      - uses: actions/checkout@v3
+      - name: Check files using the ruff formatter
+        uses: chartboost/ruff-action@v1
+        id: ruff_formatter
         with:
-          black_args: ". -l 120"
-      - name: Annotate diff changes using reviewdog
-        if: steps.black_formatter.outputs.is_formatted == 'true'
-        uses: reviewdog/action-suggester@v1
+          args: format
+      - name: Auto commit ruff formatting
+        uses: stefanzweifel/git-auto-commit-action@v5
         with:
-          tool_name: blackfmt
-      - name: Fail if there are formatting suggestions
-        if: steps.black_formatter.outputs.is_formatted == 'true'
-        run: exit 1
+          commit_message: 'style fixes by ruff'

.pre-commit-config.yaml

@@ -0,0 +1,11 @@
+default_language_version:
+  python: python3
+
+repos:
+# Run the Ruff linter.
+- repo: https://github.com/astral-sh/ruff-pre-commit
+  # Ruff version.
+  rev: v0.1.2
+  hooks:
+    # Run the Ruff formatter.
+    - id: ruff-format

CONTRIBUTING.md

@@ -54,6 +54,8 @@ Once inside, you can install the dependencies.
 
 - Option 2: Run `pip install -e .` to install them, and data-diff, in the global context.
 
+- Run `pre-commit install` to automatically format your code before committing.
+
 At the bare minimum, you need MySQL to run the tests.
 
 You can create a local MySQL instance using `docker-compose up mysql`. The URI for it will be `mysql://mysql:Password1@localhost/mysql`. If you're using a different server, make sure to update `TEST_MYSQL_CONN_STRING` in `tests/common.py`. For your convenience, we recommend creating `tests/local_settings.py`, and to override the value there.

data_diff/databases/redshift.py

@@ -126,9 +126,7 @@ def select_view_columns(self, path: DbPath) -> str:
 
         return """select * from pg_get_cols('{}.{}')
                 cols(view_schema name, view_name name, col_name name, col_type varchar, col_num int)
-            """.format(
-            schema, table
-        )
+            """.format(schema, table)
 
     def query_pg_get_cols(self, path: DbPath) -> Dict[str, tuple]:
         rows = self.query(self.select_view_columns(path), list)

data_diff/queries/ast_classes.py

@@ -337,6 +337,7 @@ def else_(self, then: Expr) -> Self:
 @attrs.define(frozen=True, eq=False)
 class QB_When:
     "Partial case-when, used for query-building"
+
     casewhen: CaseWhen
     when: Expr
 

poetry.lock

@@ -61,6 +61,7 @@ clickhouse-driver = "*"
 vertica-python = "*"
 duckdb = "^0.7.0"
 dbt-core = "^1.0.0"
+ruff = "^0.1.4"
 # google-cloud-bigquery = "*"
 # databricks-sql-connector = "*"
 
@@ -80,6 +81,9 @@ clickhouse = ["clickhouse-driver"]
 vertica = ["vertica-python"]
 duckdb = ["duckdb"]
 
+[tool.poetry.group.dev.dependencies]
+pre-commit = "^3.5.0"
+
 [build-system]
 requires = ["poetry-core>=1.0.0"]
 build-backend = "poetry.core.masonry.api"

pyproject.toml

@@ -0,0 +1,2 @@
+# Allow lines to be as long as 120.
+line-length = 120

ruff.toml

@@ -135,6 +135,7 @@ def str_to_checksum(str: str):
 
 class DiffTestCase(unittest.TestCase):
     "Sets up two tables for diffing"
+
     db_cls = None
     src_schema = None
     dst_schema = None