pre-commit-hooks/pre_commit_hooks/catch_dotenv.py

#!/usr/bin/env python
from __future__ import annotations

import argparse
import os
import re
import tempfile
from collections.abc import Sequence
from typing import Iterable

# --- Defaults / Constants ---
DEFAULT_ENV_FILE = ".env"          # Canonical env file name
DEFAULT_GITIGNORE_FILE = ".gitignore"
DEFAULT_EXAMPLE_ENV_FILE = ".env.example"
GITIGNORE_BANNER = "# Added by pre-commit hook to prevent committing secrets"

_KEY_REGEX = re.compile(r"^\s*(?:export\s+)?([A-Za-z_][A-Za-z0-9_]*)\s*=")


def _atomic_write(path: str, data: str) -> None:
    """Write text to path atomically (best-effort)."""
    # Using same directory for atomic os.replace semantics on POSIX.
    fd, tmp_path = tempfile.mkstemp(dir=os.path.dirname(path) or ".")
    try:
        with os.fdopen(fd, "w", encoding="utf-8", newline="") as tmp_f:
            tmp_f.write(data)
        os.replace(tmp_path, path)
    finally:  # Clean up if replace failed
        if os.path.exists(tmp_path):  # pragma: no cover (rare failure case)
            try:
                os.remove(tmp_path)
            except OSError:  # pragma: no cover
                pass


def ensure_env_in_gitignore(env_file: str, gitignore_file: str, banner: str) -> bool:
    """Normalize `.gitignore` so it contains exactly one banner + env line at end.

    Returns True if the file was created or its contents changed, False otherwise.
    Strategy: read existing lines, strip trailing blanks, remove any prior occurrences of
    the banner or env_file (even if duplicated), then append a single blank line,
    banner, and env_file. Produces an idempotent final layout.
    """
    try:
        if os.path.exists(gitignore_file):
            with open(gitignore_file, "r", encoding="utf-8") as f:
                lines = f.read().splitlines()
        else:
            lines = []
    except OSError as exc:
        print(f"ERROR: unable to read '{gitignore_file}': {exc}")
        return False

    original = list(lines)

    # Trim trailing blank lines
    while lines and not lines[-1].strip():
        lines.pop()

    # Remove existing occurrences (exact match after strip)
    filtered: list[str] = [ln for ln in lines if ln.strip() not in {env_file, banner}]

    if filtered and filtered[-1].strip():
        filtered.append("")  # ensure single blank before banner
    elif not filtered:  # empty file -> still separate section visually
        filtered.append("")

    filtered.append(banner)
    filtered.append(env_file)

    new_content = "\n".join(filtered) + "\n"
    if original == filtered:
        return False
    try:
        _atomic_write(gitignore_file, new_content)
        return True
    except OSError as exc:  # pragma: no cover
        print(f"ERROR: unable to write '{gitignore_file}': {exc}")
        return False


def create_example_env(src_env: str, example_file: str) -> bool:
    """Write example file containing only variable keys from real env file.

    Returns True if file written (or updated), False on read/write error.
    Lines accepted: optional 'export ' prefix then KEY=...; ignores comments & duplicates.
    """
    try:
        with open(src_env, "r", encoding="utf-8") as f_env:
            lines = f_env.readlines()
    except OSError as exc:
        print(f"ERROR: unable to read '{src_env}': {exc}")
        return False

    seen: set[str] = set()
    keys: list[str] = []
    for line in lines:
        stripped = line.strip()
        if not stripped or stripped.startswith('#'):
            continue
        m = _KEY_REGEX.match(stripped)
        if not m:
            continue
        key = m.group(1)
        if key not in seen:
            seen.add(key)
            keys.append(key)

    header = [
        '# Generated by catch-dotenv hook.',
        '# Variable names only – fill in sample values as needed.',
        '',
    ]
    body = [f"{k}=" for k in keys]
    try:
        _atomic_write(example_file, "\n".join(header + body) + "\n")
        return True
    except OSError as exc:  # pragma: no cover
        print(f"ERROR: unable to write '{example_file}': {exc}")
        return False


def _has_env(filenames: Iterable[str], env_file: str) -> bool:
    """Return True if any staged path refers to a target env file by basename."""
    return any(os.path.basename(name) == env_file for name in filenames)


def _find_repo_root(start: str = '.') -> str:
    """Ascend from start until a directory containing '.git' is found.

    Falls back to absolute path of start if no parent contains '.git'. This mirrors
    typical pre-commit execution (already at repo root) but makes behavior stable
    when hook is invoked from a subdirectory (e.g. for direct ad‑hoc testing).
    """
    cur = os.path.abspath(start)
    prev = None
    while cur != prev:
        if os.path.isdir(os.path.join(cur, '.git')):
            return cur
        prev, cur = cur, os.path.abspath(os.path.join(cur, os.pardir))
    return os.path.abspath(start)


def _print_failure(env_file: str, gitignore_file: str, example_created: bool, gitignore_modified: bool) -> None:
    parts: list[str] = [f"Blocked committing '{env_file}'."]
    if gitignore_modified:
        parts.append(f"Added to '{gitignore_file}'.")
    if example_created:
        parts.append("Example file generated.")
    parts.append(f"Remove '{env_file}' from the commit and commit again.")
    print(" ".join(parts))


def main(argv: Sequence[str] | None = None) -> int:
    """Main function for the pre-commit hook."""
    parser = argparse.ArgumentParser(description="Block committing environment files (.env).")
    parser.add_argument('filenames', nargs='*', help='Staged filenames (supplied by pre-commit).')
    parser.add_argument('--create-example', action='store_true', help='Generate example env file (.env.example).')
    args = parser.parse_args(argv)
    env_file = DEFAULT_ENV_FILE
    # Resolve repository root (directory containing .git) so writes happen there
    repo_root = _find_repo_root('.')
    gitignore_file = os.path.join(repo_root, DEFAULT_GITIGNORE_FILE)
    example_file = os.path.join(repo_root, DEFAULT_EXAMPLE_ENV_FILE)
    env_abspath = os.path.join(repo_root, env_file)

    if not _has_env(args.filenames, env_file):
        return 0

    gitignore_modified = ensure_env_in_gitignore(env_file, gitignore_file, GITIGNORE_BANNER)
    example_created = False
    if args.create_example:
        # Source env is always looked up relative to repo root
        if os.path.exists(env_abspath):
            example_created = create_example_env(env_abspath, example_file)

    _print_failure(env_file, gitignore_file, example_created, gitignore_modified)
    return 1  # Block commit


if __name__ == '__main__':
    raise SystemExit(main())