X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/858225d34dc49ca353f9e573dd82dd5845766115..ea9fefb359b86208ad20c944c3c420e6c3fb2c5e:/src/black/__init__.py?ds=inline diff --git a/src/black/__init__.py b/src/black/__init__.py index a1d16d9..0a893aa 100644 --- a/src/black/__init__.py +++ b/src/black/__init__.py @@ -48,7 +48,20 @@ from appdirs import user_cache_dir from dataclasses import dataclass, field, replace import click import toml -from typed_ast import ast3, ast27 + +try: + from typed_ast import ast3, ast27 +except ImportError: + if sys.version_info < (3, 8): + print( + "The typed_ast package is not installed.\n" + "You can install it with `python3 -m pip install typed-ast`.", + file=sys.stderr, + ) + sys.exit(1) + else: + ast3 = ast27 = ast + from pathspec import PathSpec # lib2to3 fork @@ -69,7 +82,7 @@ if TYPE_CHECKING: import colorama # noqa: F401 DEFAULT_LINE_LENGTH = 88 -DEFAULT_EXCLUDES = r"/(\.direnv|\.eggs|\.git|\.hg|\.mypy_cache|\.nox|\.tox|\.venv|\.svn|_build|buck-out|build|dist)/" # noqa: B950 +DEFAULT_EXCLUDES = r"/(\.direnv|\.eggs|\.git|\.hg|\.mypy_cache|\.nox|\.tox|\.venv|venv|\.svn|_build|buck-out|build|dist)/" # noqa: B950 DEFAULT_INCLUDES = r"\.pyi?$" CACHE_DIR = Path(user_cache_dir("black", version=__version__)) STDIN_PLACEHOLDER = "__BLACK_STDIN_FILENAME__" @@ -289,11 +302,15 @@ def supports_feature(target_versions: Set[TargetVersion], feature: Feature) -> b return all(feature in VERSION_TO_FEATURES[version] for version in target_versions) -def find_pyproject_toml(path_search_start: Iterable[str]) -> Optional[str]: +def find_pyproject_toml(path_search_start: Tuple[str, ...]) -> Optional[str]: """Find the absolute filepath to a pyproject.toml if it exists""" path_project_root = find_project_root(path_search_start) path_pyproject_toml = path_project_root / "pyproject.toml" - return str(path_pyproject_toml) if path_pyproject_toml.is_file() else None + if path_pyproject_toml.is_file(): + return str(path_pyproject_toml) + + path_user_pyproject_toml = find_user_pyproject_toml() + return str(path_user_pyproject_toml) if path_user_pyproject_toml.is_file() else None def parse_pyproject_toml(path_config: str) -> Dict[str, Any]: @@ -363,6 +380,17 @@ def target_version_option_callback( return [TargetVersion[val.upper()] for val in v] +def validate_regex( + ctx: click.Context, + param: click.Parameter, + value: Optional[str], +) -> Optional[Pattern]: + try: + return re_compile_maybe_verbose(value) if value is not None else None + except re.error: + raise click.BadParameter("Not a valid regular expression") + + @click.command(context_settings=dict(help_option_names=["-h", "--help"])) @click.option("-c", "--code", type=str, help="Format the code passed in as a string.") @click.option( @@ -417,8 +445,8 @@ def target_version_option_callback( "--check", is_flag=True, help=( - "Don't write the files back, just return the status. Return code 0 means" - " nothing would change. Return code 1 means some files would be reformatted." + "Don't write the files back, just return the status. Return code 0 means" + " nothing would change. Return code 1 means some files would be reformatted." " Return code 123 means there was an internal error." ), ) @@ -441,11 +469,12 @@ def target_version_option_callback( "--include", type=str, default=DEFAULT_INCLUDES, + callback=validate_regex, help=( "A regular expression that matches files and directories that should be" - " included on recursive searches. An empty value means all files are included" - " regardless of the name. Use forward slashes for directories on all platforms" - " (Windows, too). Exclusions are calculated first, inclusions later." + " included on recursive searches. An empty value means all files are included" + " regardless of the name. Use forward slashes for directories on all platforms" + " (Windows, too). Exclusions are calculated first, inclusions later." ), show_default=True, ) @@ -453,17 +482,28 @@ def target_version_option_callback( "--exclude", type=str, default=DEFAULT_EXCLUDES, + callback=validate_regex, help=( "A regular expression that matches files and directories that should be" - " excluded on recursive searches. An empty value means no paths are excluded." - " Use forward slashes for directories on all platforms (Windows, too). " + " excluded on recursive searches. An empty value means no paths are excluded." + " Use forward slashes for directories on all platforms (Windows, too)." " Exclusions are calculated first, inclusions later." ), show_default=True, ) +@click.option( + "--extend-exclude", + type=str, + callback=validate_regex, + help=( + "Like --exclude, but adds additional files and directories on top of the" + " excluded ones. (Useful if you simply want to add to the default)" + ), +) @click.option( "--force-exclude", type=str, + callback=validate_regex, help=( "Like --exclude, but files and directories matching this regex will be " "excluded even when they are passed explicitly as arguments." @@ -493,7 +533,7 @@ def target_version_option_callback( is_flag=True, help=( "Also emit messages to stderr about files that were not changed or were ignored" - " due to --exclude=." + " due to exclusion patterns." ), ) @click.version_option(version=__version__) @@ -535,9 +575,10 @@ def main( experimental_string_processing: bool, quiet: bool, verbose: bool, - include: str, - exclude: str, - force_exclude: Optional[str], + include: Pattern, + exclude: Pattern, + extend_exclude: Optional[Pattern], + force_exclude: Optional[Pattern], stdin_filename: Optional[str], src: Tuple[str, ...], config: Optional[str], @@ -570,6 +611,7 @@ def main( verbose=verbose, include=include, exclude=exclude, + extend_exclude=extend_exclude, force_exclude=force_exclude, report=report, stdin_filename=stdin_filename, @@ -608,30 +650,14 @@ def get_sources( src: Tuple[str, ...], quiet: bool, verbose: bool, - include: str, - exclude: str, - force_exclude: Optional[str], + include: Pattern[str], + exclude: Pattern[str], + extend_exclude: Optional[Pattern[str]], + force_exclude: Optional[Pattern[str]], report: "Report", stdin_filename: Optional[str], ) -> Set[Path]: """Compute the set of files to be formatted.""" - try: - include_regex = re_compile_maybe_verbose(include) - except re.error: - err(f"Invalid regular expression for include given: {include!r}") - ctx.exit(2) - try: - exclude_regex = re_compile_maybe_verbose(exclude) - except re.error: - err(f"Invalid regular expression for exclude given: {exclude!r}") - ctx.exit(2) - try: - force_exclude_regex = ( - re_compile_maybe_verbose(force_exclude) if force_exclude else None - ) - except re.error: - err(f"Invalid regular expression for force_exclude given: {force_exclude!r}") - ctx.exit(2) root = find_project_root(src) sources: Set[Path] = set() @@ -653,8 +679,8 @@ def get_sources( normalized_path = "/" + normalized_path # Hard-exclude any files that matches the `--force-exclude` regex. - if force_exclude_regex: - force_exclude_match = force_exclude_regex.search(normalized_path) + if force_exclude: + force_exclude_match = force_exclude.search(normalized_path) else: force_exclude_match = None if force_exclude_match and force_exclude_match.group(0): @@ -670,9 +696,10 @@ def get_sources( gen_python_files( p.iterdir(), root, - include_regex, - exclude_regex, - force_exclude_regex, + include, + exclude, + extend_exclude, + force_exclude, report, gitignore, ) @@ -757,7 +784,7 @@ def reformat_many( except (ImportError, OSError): # we arrive here if the underlying system does not support multi-processing # like in AWS Lambda or Termux, in which case we gracefully fallback to - # a ThreadPollExecutor with just a single worker (more workers would not do us + # a ThreadPoolExecutor with just a single worker (more workers would not do us # any good due to the Global Interpreter Lock) executor = ThreadPoolExecutor(max_workers=1) @@ -820,7 +847,7 @@ async def schedule_formatting( ): src for src in sorted(sources) } - pending: Iterable["asyncio.Future[bool]"] = tasks.keys() + pending = tasks.keys() try: loop.add_signal_handler(signal.SIGINT, cancel, pending) loop.add_signal_handler(signal.SIGTERM, cancel, pending) @@ -2126,16 +2153,35 @@ class LineGenerator(Visitor[Line]): # We're ignoring docstrings with backslash newline escapes because changing # indentation of those changes the AST representation of the code. prefix = get_string_prefix(leaf.value) - lead_len = len(prefix) + 3 - tail_len = -3 - indent = " " * 4 * self.current_line.depth - docstring = fix_docstring(leaf.value[lead_len:tail_len], indent) + docstring = leaf.value[len(prefix) :] # Remove the prefix + quote_char = docstring[0] + # A natural way to remove the outer quotes is to do: + # docstring = docstring.strip(quote_char) + # but that breaks on """""x""" (which is '""x'). + # So we actually need to remove the first character and the next two + # characters but only if they are the same as the first. + quote_len = 1 if docstring[1] != quote_char else 3 + docstring = docstring[quote_len:-quote_len] + + if is_multiline_string(leaf): + indent = " " * 4 * self.current_line.depth + docstring = fix_docstring(docstring, indent) + else: + docstring = docstring.strip() + if docstring: - if leaf.value[lead_len - 1] == docstring[0]: + # Add some padding if the docstring starts / ends with a quote mark. + if docstring[0] == quote_char: docstring = " " + docstring - if leaf.value[tail_len + 1] == docstring[-1]: + if docstring[-1] == quote_char: docstring = docstring + " " - leaf.value = leaf.value[0:lead_len] + docstring + leaf.value[tail_len:] + else: + # Add some padding if the docstring is empty. + docstring = " " + + # We could enforce triple quotes at this point. + quote = quote_char * quote_len + leaf.value = prefix + quote + docstring + quote yield from self.visit_default(leaf) @@ -2682,6 +2728,13 @@ def make_comment(content: str) -> str: if content[0] == "#": content = content[1:] + NON_BREAKING_SPACE = " " + if ( + content + and content[0] == NON_BREAKING_SPACE + and not content.lstrip().startswith("type:") + ): + content = " " + content[1:] # Replace NBSP by a simple space if content and content[0] not in " !:#'%": content = " " + content return "#" + content @@ -5298,15 +5351,10 @@ def normalize_numeric_literal(leaf: Leaf) -> None: def format_hex(text: str) -> str: """ - Formats a hexadecimal string like "0x12b3" - - Uses lowercase because of similarity between "B" and "8", which - can cause security issues. - see: https://github.com/psf/black/issues/1692 + Formats a hexadecimal string like "0x12B3" """ - before, after = text[:2], text[2:] - return f"{before}{after.lower()}" + return f"{before}{after.upper()}" def format_scientific_notation(text: str) -> str: @@ -6079,7 +6127,7 @@ def get_future_imports(node: Node) -> Set[str]: @lru_cache() def get_gitignore(root: Path) -> PathSpec: - """ Return a PathSpec matching gitignore content if present.""" + """Return a PathSpec matching gitignore content if present.""" gitignore = root / ".gitignore" lines: List[str] = [] if gitignore.is_file(): @@ -6112,17 +6160,27 @@ def normalize_path_maybe_ignore( return normalized_path +def path_is_excluded( + normalized_path: str, + pattern: Optional[Pattern[str]], +) -> bool: + match = pattern.search(normalized_path) if pattern else None + return bool(match and match.group(0)) + + def gen_python_files( paths: Iterable[Path], root: Path, include: Optional[Pattern[str]], exclude: Pattern[str], + extend_exclude: Optional[Pattern[str]], force_exclude: Optional[Pattern[str]], report: "Report", gitignore: PathSpec, ) -> Iterator[Path]: """Generate all files under `path` whose paths are not excluded by the - `exclude_regex` or `force_exclude` regexes, but are included by the `include` regex. + `exclude_regex`, `extend_exclude`, or `force_exclude` regexes, + but are included by the `include` regex. Symbolic links pointing outside of the `root` directory are ignored. @@ -6139,20 +6197,22 @@ def gen_python_files( report.path_ignored(child, "matches the .gitignore file content") continue - # Then ignore with `--exclude` and `--force-exclude` options. + # Then ignore with `--exclude` `--extend-exclude` and `--force-exclude` options. normalized_path = "/" + normalized_path if child.is_dir(): normalized_path += "/" - exclude_match = exclude.search(normalized_path) if exclude else None - if exclude_match and exclude_match.group(0): + if path_is_excluded(normalized_path, exclude): report.path_ignored(child, "matches the --exclude regular expression") continue - force_exclude_match = ( - force_exclude.search(normalized_path) if force_exclude else None - ) - if force_exclude_match and force_exclude_match.group(0): + if path_is_excluded(normalized_path, extend_exclude): + report.path_ignored( + child, "matches the --extend-exclude regular expression" + ) + continue + + if path_is_excluded(normalized_path, force_exclude): report.path_ignored(child, "matches the --force-exclude regular expression") continue @@ -6162,6 +6222,7 @@ def gen_python_files( root, include, exclude, + extend_exclude, force_exclude, report, gitignore, @@ -6174,7 +6235,7 @@ def gen_python_files( @lru_cache() -def find_project_root(srcs: Iterable[str]) -> Path: +def find_project_root(srcs: Tuple[str, ...]) -> Path: """Return a directory containing .git, .hg, or pyproject.toml. That directory will be a common parent of all files and directories @@ -6212,6 +6273,22 @@ def find_project_root(srcs: Iterable[str]) -> Path: return directory +@lru_cache() +def find_user_pyproject_toml() -> Path: + r"""Return the path to the top-level user configuration for black. + + This looks for ~\.black on Windows and ~/.config/black on Linux and other + Unix systems. + """ + if sys.platform == "win32": + # Windows + user_config_path = Path.home() / ".black" + else: + config_root = os.environ.get("XDG_CONFIG_HOME", "~/.config") + user_config_path = Path(config_root).expanduser() / "black" + return user_config_path.resolve() + + @dataclass class Report: """Provides a reformatting counter. Can be rendered with `str(report)`.""" @@ -6313,7 +6390,12 @@ def parse_ast(src: str) -> Union[ast.AST, ast3.AST, ast27.AST]: return ast3.parse(src, filename, feature_version=feature_version) except SyntaxError: continue - + if ast27.__name__ == "ast": + raise SyntaxError( + "The requested source code has invalid Python 3 syntax.\n" + "If you are trying to format Python 2 files please reinstall Black" + " with the 'python2' extra: `python3 -m pip install black[python2]`." + ) return ast27.parse(src) @@ -6379,12 +6461,22 @@ def _stringify_ast( # Constant strings may be indented across newlines, if they are # docstrings; fold spaces after newlines when comparing. Similarly, # trailing and leading space may be removed. + # Note that when formatting Python 2 code, at least with Windows + # line-endings, docstrings can end up here as bytes instead of + # str so make sure that we handle both cases. if ( isinstance(node, ast.Constant) and field == "value" - and isinstance(value, str) + and isinstance(value, (str, bytes)) ): - normalized = re.sub(r" *\n[ \t]*", "\n", value).strip() + lineend = "\n" if isinstance(value, str) else b"\n" + # To normalize, we strip any leading and trailing space from + # each line... + stripped = [line.strip() for line in value.splitlines()] + normalized = lineend.join(stripped) # type: ignore[attr-defined] + # ...and remove any blank lines at the beginning and end of + # the whole string + normalized = normalized.strip() else: normalized = value yield f"{' ' * (depth+2)}{normalized!r}, # {value.__class__.__name__}" @@ -6885,11 +6977,6 @@ def patched_main() -> None: def is_docstring(leaf: Leaf) -> bool: - if not is_multiline_string(leaf): - # For the purposes of docstring re-indentation, we don't need to do anything - # with single-line docstrings. - return False - if prev_siblings_are( leaf.parent, [None, token.NEWLINE, token.INDENT, syms.simple_stmt] ):