ci: Check F ruff rules in docstrings

.pre-commit-config.yaml

@@ -33,13 +33,14 @@ repos:
             (?x)^(
                 tests/.*|
                 # TODO: gradually enable
-                narwhals/series\.py$|
+                narwhals/series\.py|
                 # TODO: gradually enable
-                narwhals/dataframe\.py$|
+                narwhals/dataframe\.py|
                 # TODO: gradually enable
-                narwhals/dependencies\.py$|
+                narwhals/dependencies\.py|
                 # private, so less urgent to document too well
-                narwhals/_.*
+                narwhals/_.*|
+                ^utils/.*
             )$
 - repo: local
   hooks:
@@ -55,6 +56,11 @@ repos:
       language: python
       files: ^narwhals/
       exclude: ^narwhals/dependencies\.py
+    - id: check-docstrings-execute
+      name: check docstrings execute
+      entry: python utils/check_docstrings.py
+      language: python
+      files: ^narwhals/
 - repo: https://github.com/kynan/nbstripout
   rev: 0.8.0
   hooks:

narwhals/stable/v1/__init__.py

@@ -1773,15 +1773,15 @@ def narwhalify(
         Instead of writing
 
         >>> import narwhals as nw
-        >>> def func(df):
+        >>> def agnostic_group_by_sum(df):
         ...     df = nw.from_native(df, pass_through=True)
         ...     df = df.group_by("a").agg(nw.col("b").sum())
         ...     return nw.to_native(df)
 
         you can just write
 
         >>> @nw.narwhalify
-        ... def func(df):
+        ... def agnostic_group_by_sum(df):
         ...     return df.group_by("a").agg(nw.col("b").sum())
     """
     pass_through = validate_strict_and_pass_though(

narwhals/translate.py

@@ -816,15 +816,15 @@ def narwhalify(
         Instead of writing
 
         >>> import narwhals as nw
-        >>> def func(df):
+        >>> def agnostic_group_by_sum(df):
         ...     df = nw.from_native(df, pass_through=True)
         ...     df = df.group_by("a").agg(nw.col("b").sum())
         ...     return nw.to_native(df)
 
         you can just write
 
         >>> @nw.narwhalify
-        ... def func(df):
+        ... def agnostic_group_by_sum(df):
         ...     return df.group_by("a").agg(nw.col("b").sum())
     """
     from narwhals.utils import validate_strict_and_pass_though

requirements-dev.txt

@@ -9,6 +9,7 @@ pytest
 pytest-cov
 pytest-randomly
 pytest-env
+ruff
 hypothesis
 hypothesis[numpy]
 scikit-learn

utils/check_docstrings.py

@@ -0,0 +1,110 @@
+from __future__ import annotations
+
+import ast
+import doctest
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+
+def extract_docstring_examples(files: list[Path]) -> list[tuple[Path, str, str]]:
+    """Extract examples from docstrings in Python files."""
+    examples: list[tuple[Path, str, str]] = []
+
+    for file in files:
+        with open(file, encoding="utf-8") as f:
+            tree = ast.parse(f.read())
+
+        for node in ast.walk(tree):
+            if isinstance(node, (ast.FunctionDef, ast.ClassDef)):
+                docstring = ast.get_docstring(node)
+                if docstring:
+                    parsed_examples = doctest.DocTestParser().get_examples(docstring)
+                    example_code = "\n".join(
+                        example.source for example in parsed_examples
+                    )
+                    if example_code.strip():
+                        examples.append((file, node.name, example_code))
+
+    return examples
+
+
+def create_temp_files(examples: list[tuple[Path, str, str]]) -> list[tuple[Path, str]]:
+    """Create temporary files for all examples and return their paths."""
+    temp_files: list[tuple[Path, str]] = []
+
+    for file, name, example in examples:
+        temp_file = tempfile.NamedTemporaryFile(mode="w", suffix=".py", delete=False)  # noqa: SIM115
+        temp_file.write(example)
+        temp_file_path = temp_file.name
+        temp_file.close()
+        temp_files.append((Path(temp_file_path), f"{file}:{name}"))
+
+    return temp_files
+
+
+def run_ruff_on_temp_files(temp_files: list[tuple[Path, str]]) -> list[str]:
+    """Run ruff on all temporary files and collect error messages."""
+    temp_file_paths = [str(temp_file[0]) for temp_file in temp_files]
+
+    result = subprocess.run(  # noqa: S603
+        [  # noqa: S607
+            "python",
+            "-m",
+            "ruff",
+            "check",
+            "--select=F",
+            "--ignore=F811",
+            *temp_file_paths,
+        ],
+        capture_output=True,
+        text=True,
+        check=False,
+    )
+
+    if result.returncode == 0:
+        return []  # No issues found
+    return result.stdout.splitlines()  # Return ruff errors as a list of lines
+
+
+def report_errors(errors: list[str], temp_files: list[tuple[Path, str]]) -> None:
+    """Map errors back to original examples and report them."""
+    if not errors:
+        return
+
+    print("❌ Ruff issues found in examples:\n")  # noqa: T201
+    for line in errors:
+        for temp_file, original_context in temp_files:
+            if str(temp_file) in line:
+                print(f"{original_context}{line.replace(str(temp_file), '')}")  # noqa: T201
+                break
+
+
+def cleanup_temp_files(temp_files: list[tuple[Path, str]]) -> None:
+    """Remove all temporary files."""
+    for temp_file, _ in temp_files:
+        temp_file.unlink()
+
+
+def main(python_files: list[str]) -> None:
+    docstring_examples = extract_docstring_examples(python_files)
+
+    if not docstring_examples:
+        sys.exit(0)
+
+    temp_files = create_temp_files(docstring_examples)
+
+    try:
+        errors = run_ruff_on_temp_files(temp_files)
+        report_errors(errors, temp_files)
+    finally:
+        cleanup_temp_files(temp_files)
+
+    if errors:
+        sys.exit(1)
+    sys.exit(0)
+
+
+if __name__ == "__main__":
+    main(sys.argv[1:])