X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/0753d99519b0c90f0f9f280b73783b537900dc16..82e150a13ac78a1f24de27977d8a40883bc3d6e7:/src/black/__init__.py?ds=sidebyside diff --git a/src/black/__init__.py b/src/black/__init__.py index ad4ee1a..e9d3c1b 100644 --- a/src/black/__init__.py +++ b/src/black/__init__.py @@ -10,7 +10,7 @@ from multiprocessing import Manager, freeze_support import os from pathlib import Path from pathspec.patterns.gitwildmatch import GitWildMatchPatternError -import regex as re +import re import signal import sys import tokenize @@ -24,28 +24,37 @@ from typing import ( MutableMapping, Optional, Pattern, + Sequence, Set, Sized, Tuple, Union, ) -from dataclasses import replace import click +from click.core import ParameterSource +from dataclasses import replace +from mypy_extensions import mypyc_attr from black.const import DEFAULT_LINE_LENGTH, DEFAULT_INCLUDES, DEFAULT_EXCLUDES from black.const import STDIN_PLACEHOLDER from black.nodes import STARS, syms, is_simple_decorator_expression +from black.nodes import is_string_token from black.lines import Line, EmptyLineTracker from black.linegen import transform_line, LineGenerator, LN from black.comments import normalize_fmt_off -from black.mode import Mode, TargetVersion +from black.mode import FUTURE_FLAG_TO_FEATURE, Mode, TargetVersion from black.mode import Feature, supports_feature, VERSION_TO_FEATURES from black.cache import read_cache, write_cache, get_cache_info, filter_cached, Cache from black.concurrency import cancel, shutdown, maybe_install_uvloop from black.output import dump_to_file, ipynb_diff, diff, color_diff, out, err from black.report import Report, Changed, NothingChanged -from black.files import find_project_root, find_pyproject_toml, parse_pyproject_toml +from black.files import ( + find_project_root, + find_pyproject_toml, + parse_pyproject_toml, + find_user_pyproject_toml, +) from black.files import gen_python_files, get_gitignore, normalize_path_maybe_ignore from black.files import wrap_stream_for_windows from black.parsing import InvalidInput # noqa F401 @@ -56,6 +65,7 @@ from black.handle_ipynb_magics import ( remove_trailing_semicolon, put_trailing_semicolon_back, TRANSFORMED_MAGICS, + PYTHON_CELL_MAGICS, jupyter_dependencies_are_installed, ) @@ -66,6 +76,8 @@ from blib2to3.pgen2 import token from _black_version import version as __version__ +COMPILED = Path(__file__).suffix in (".pyd", ".so") + # types FileContent = str Encoding = str @@ -173,11 +185,16 @@ def validate_regex( ) -> Optional[Pattern[str]]: 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") from None + except re.error as e: + raise click.BadParameter(f"Not a valid regular expression: {e}") from None -@click.command(context_settings=dict(help_option_names=["-h", "--help"])) +@click.command( + context_settings={"help_option_names": ["-h", "--help"]}, + # While Click does set this field automatically using the docstring, mypyc + # (annoyingly) strips 'em so we need to set it here too. + help="The uncompromising code formatter.", +) @click.option("-c", "--code", type=str, help="Format the code passed in as a string.") @click.option( "-l", @@ -214,6 +231,16 @@ def validate_regex( "(useful when piping source on standard input)." ), ) +@click.option( + "--python-cell-magics", + multiple=True, + help=( + "When processing Jupyter Notebooks, add the given magic to the list" + f" of known python-magics ({', '.join(PYTHON_CELL_MAGICS)})." + " Useful for formatting cells with custom python magics." + ), + default=[], +) @click.option( "-S", "--skip-string-normalization", @@ -230,9 +257,14 @@ def validate_regex( "--experimental-string-processing", is_flag=True, hidden=True, + help="(DEPRECATED and now included in --preview) Normalize string literals.", +) +@click.option( + "--preview", + is_flag=True, help=( - "Experimental option that performs more normalization on string literals." - " Currently disabled because it leads to some crashes." + "Enable potentially disruptive style changes that may be added to Black's main" + " functionality in the next major release." ), ) @click.option( @@ -264,7 +296,8 @@ def validate_regex( type=str, help=( "Require a specific version of Black to be running (useful for unifying results" - " across many environments e.g. with a pyproject.toml file)." + " across many environments e.g. with a pyproject.toml file). It can be" + " either a major version number or an exact version." ), ) @click.option( @@ -346,7 +379,10 @@ def validate_regex( " due to exclusion patterns." ), ) -@click.version_option(version=__version__) +@click.version_option( + version=__version__, + message=f"%(prog)s, %(version)s (compiled: {'yes' if COMPILED else 'no'})", +) @click.argument( "src", nargs=-1, @@ -371,7 +407,7 @@ def validate_regex( help="Read configuration from FILE path.", ) @click.pass_context -def main( +def main( # noqa: C901 ctx: click.Context, code: Optional[str], line_length: int, @@ -382,12 +418,14 @@ def main( fast: bool, pyi: bool, ipynb: bool, + python_cell_magics: Sequence[str], skip_string_normalization: bool, skip_magic_trailing_comma: bool, experimental_string_processing: bool, + preview: bool, quiet: bool, verbose: bool, - required_version: str, + required_version: Optional[str], include: Pattern[str], exclude: Optional[Pattern[str]], extend_exclude: Optional[Pattern[str]], @@ -398,11 +436,65 @@ def main( config: Optional[str], ) -> None: """The uncompromising code formatter.""" - if config and verbose: - out(f"Using configuration from {config}.", bold=False, fg="blue") + ctx.ensure_object(dict) + + if src and code is not None: + out( + main.get_usage(ctx) + + "\n\n'SRC' and 'code' cannot be passed simultaneously." + ) + ctx.exit(1) + if not src and code is None: + out(main.get_usage(ctx) + "\n\nOne of 'SRC' or 'code' is required.") + ctx.exit(1) + + root, method = find_project_root(src) if code is None else (None, None) + ctx.obj["root"] = root + + if verbose: + if root: + out( + f"Identified `{root}` as project root containing a {method}.", + fg="blue", + ) + + normalized = [ + (normalize_path_maybe_ignore(Path(source), root), source) + for source in src + ] + srcs_string = ", ".join( + [ + f'"{_norm}"' + if _norm + else f'\033[31m"{source} (skipping - invalid)"\033[34m' + for _norm, source in normalized + ] + ) + out(f"Sources to be formatted: {srcs_string}", fg="blue") + + if config: + config_source = ctx.get_parameter_source("config") + user_level_config = str(find_user_pyproject_toml()) + if config == user_level_config: + out( + "Using configuration from user-level config at " + f"'{user_level_config}'.", + fg="blue", + ) + elif config_source in ( + ParameterSource.DEFAULT, + ParameterSource.DEFAULT_MAP, + ): + out("Using configuration from project root.", fg="blue") + else: + out(f"Using configuration in '{config}'.", fg="blue") error_msg = "Oh no! 💥 💔 💥" - if required_version and required_version != __version__: + if ( + required_version + and required_version != __version__ + and required_version != __version__.split(".")[0] + ): err( f"{error_msg} The required version `{required_version}` does not match" f" the running version `{__version__}`!" @@ -426,6 +518,8 @@ def main( string_normalization=not skip_string_normalization, magic_trailing_comma=not skip_magic_trailing_comma, experimental_string_processing=experimental_string_processing, + preview=preview, + python_cell_magics=set(python_cell_magics), ) if code is not None: @@ -483,6 +577,8 @@ def main( ) if verbose or not quiet: + if code is None and (verbose or report.change_count or report.failure_count): + out() out(error_msg if report.return_code else "All done! ✨ 🍰 ✨") if code is None: click.echo(str(report), err=True) @@ -503,14 +599,11 @@ def get_sources( stdin_filename: Optional[str], ) -> Set[Path]: """Compute the set of files to be formatted.""" - - root = find_project_root(src) sources: Set[Path] = set() - path_empty(src, "No Path provided. Nothing to do 😴", quiet, verbose, ctx) if exclude is None: exclude = re_compile_maybe_verbose(DEFAULT_EXCLUDES) - gitignore = get_gitignore(root) + gitignore = get_gitignore(ctx.obj["root"]) else: gitignore = None @@ -523,7 +616,7 @@ def get_sources( is_stdin = False if is_stdin or p.is_file(): - normalized_path = normalize_path_maybe_ignore(p, root, report) + normalized_path = normalize_path_maybe_ignore(p, ctx.obj["root"], report) if normalized_path is None: continue @@ -550,7 +643,7 @@ def get_sources( sources.update( gen_python_files( p.iterdir(), - root, + ctx.obj["root"], include, exclude, extend_exclude, @@ -604,6 +697,9 @@ def reformat_code( report.failed(path, str(exc)) +# diff-shades depends on being to monkeypatch this function to operate. I know it's +# not ideal, but this shouldn't cause any issues ... hopefully. ~ichard26 +@mypyc_attr(patchable=True) def reformat_one( src: Path, fast: bool, write_back: WriteBack, mode: Mode, report: "Report" ) -> None: @@ -655,6 +751,9 @@ def reformat_one( report.failed(src, str(exc)) +# diff-shades depends on being to monkeypatch this function to operate. I know it's +# not ideal, but this shouldn't cause any issues ... hopefully. ~ichard26 +@mypyc_attr(patchable=True) def reformat_many( sources: Set[Path], fast: bool, @@ -669,10 +768,11 @@ def reformat_many( worker_count = workers if workers is not None else DEFAULT_WORKERS if sys.platform == "win32": # Work around https://bugs.python.org/issue26903 + assert worker_count is not None worker_count = min(worker_count, 60) try: executor = ProcessPoolExecutor(max_workers=worker_count) - except (ImportError, OSError): + except (ImportError, NotImplementedError, 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 ThreadPoolExecutor with just a single worker (more workers would not do us @@ -891,17 +991,7 @@ def check_stability_and_equivalence( content differently. """ assert_equivalent(src_contents, dst_contents) - - # Forced second pass to work around optional trailing commas (becoming - # forced trailing commas on pass 2) interacting differently with optional - # parentheses. Admittedly ugly. - dst_contents_pass2 = format_str(dst_contents, mode=mode) - if dst_contents != dst_contents_pass2: - dst_contents = dst_contents_pass2 - assert_equivalent(src_contents, dst_contents, pass_num=2) - assert_stable(src_contents, dst_contents, mode=mode) - # Note: no need to explicitly call `assert_stable` if `dst_contents` was - # the same as `dst_contents_pass2`. + assert_stable(src_contents, dst_contents, mode=mode) def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileContent: @@ -927,8 +1017,10 @@ def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileCo return dst_contents -def validate_cell(src: str) -> None: - """Check that cell does not already contain TransformerManager transformations. +def validate_cell(src: str, mode: Mode) -> None: + """Check that cell does not already contain TransformerManager transformations, + or non-Python cell magics, which might cause tokenizer_rt to break because of + indentations. If a cell contains ``!ls``, then it'll be transformed to ``get_ipython().system('ls')``. However, if the cell originally contained @@ -944,6 +1036,11 @@ def validate_cell(src: str) -> None: """ if any(transformed_magic in src for transformed_magic in TRANSFORMED_MAGICS): raise NothingChanged + if ( + src[:2] == "%%" + and src.split()[0][2:] not in PYTHON_CELL_MAGICS | mode.python_cell_magics + ): + raise NothingChanged def format_cell(src: str, *, fast: bool, mode: Mode) -> str: @@ -962,7 +1059,7 @@ def format_cell(src: str, *, fast: bool, mode: Mode) -> str: could potentially be automagics or multi-line magics, which are currently not supported. """ - validate_cell(src) + validate_cell(src, mode) src_without_trailing_semicolon, has_trailing_semicolon = remove_trailing_semicolon( src ) @@ -1024,7 +1121,7 @@ def format_ipynb_string(src_contents: str, *, fast: bool, mode: Mode) -> FileCon raise NothingChanged -def format_str(src_contents: str, *, mode: Mode) -> FileContent: +def format_str(src_contents: str, *, mode: Mode) -> str: """Reformat a string and return new contents. `mode` determines formatting options, such as how many characters per line are @@ -1054,28 +1151,26 @@ def format_str(src_contents: str, *, mode: Mode) -> FileContent: hey """ + dst_contents = _format_str_once(src_contents, mode=mode) + # Forced second pass to work around optional trailing commas (becoming + # forced trailing commas on pass 2) interacting differently with optional + # parentheses. Admittedly ugly. + if src_contents != dst_contents: + return _format_str_once(dst_contents, mode=mode) + return dst_contents + + +def _format_str_once(src_contents: str, *, mode: Mode) -> str: src_node = lib2to3_parse(src_contents.lstrip(), mode.target_versions) dst_contents = [] future_imports = get_future_imports(src_node) if mode.target_versions: versions = mode.target_versions else: - versions = detect_target_versions(src_node) + versions = detect_target_versions(src_node, future_imports=future_imports) - # TODO: fully drop support and this code hopefully in January 2022 :D - if TargetVersion.PY27 in mode.target_versions or versions == {TargetVersion.PY27}: - msg = ( - "DEPRECATION: Python 2 support will be removed in the first stable release " - "expected in January 2022." - ) - err(msg, fg="yellow", bold=True) - - normalize_fmt_off(src_node) - lines = LineGenerator( - mode=mode, - remove_u_prefix="unicode_literals" in future_imports - or supports_feature(versions, Feature.UNICODE_LITERALS), - ) + normalize_fmt_off(src_node, preview=mode.preview) + lines = LineGenerator(mode=mode) elt = EmptyLineTracker(is_pyi=mode.is_pyi) empty_line = Line(mode=mode) after = 0 @@ -1112,7 +1207,9 @@ def decode_bytes(src: bytes) -> Tuple[FileContent, Encoding, NewLine]: return tiow.read(), encoding, newline -def get_features_used(node: Node) -> Set[Feature]: # noqa: C901 +def get_features_used( # noqa: C901 + node: Node, *, future_imports: Optional[Set[str]] = None +) -> Set[Feature]: """Return a set of (relatively) new Python features used in this file. Currently looking for: @@ -1122,12 +1219,20 @@ def get_features_used(node: Node) -> Set[Feature]: # noqa: C901 - positional only arguments in function signatures and lambdas; - assignment expression; - relaxed decorator syntax; + - usage of __future__ flags (annotations); - print / exec statements; """ features: Set[Feature] = set() + if future_imports: + features |= { + FUTURE_FLAG_TO_FEATURE[future_import] + for future_import in future_imports + if future_import in FUTURE_FLAG_TO_FEATURE + } + for n in node.pre_order(): - if n.type == token.STRING: - value_head = n.value[:2] # type: ignore + if is_string_token(n): + value_head = n.value[:2] if value_head in {'f"', 'F"', "f'", "F'", "rf", "fr", "RF", "FR"}: features.add(Feature.F_STRINGS) @@ -1135,14 +1240,6 @@ def get_features_used(node: Node) -> Set[Feature]: # noqa: C901 assert isinstance(n, Leaf) if "_" in n.value: features.add(Feature.NUMERIC_UNDERSCORES) - elif n.value.endswith(("L", "l")): - # Python 2: 10L - features.add(Feature.LONG_INT_LITERAL) - elif len(n.value) >= 2 and n.value[0] == "0" and n.value[1].isdigit(): - # Python 2: 0123; 00123; ... - if not all(char == "0" for char in n.value): - # although we don't want to match 0000 or similar - features.add(Feature.OCTAL_INT_LITERAL) elif n.type == token.SLASH: if n.parent and n.parent.type in { @@ -1180,38 +1277,29 @@ def get_features_used(node: Node) -> Set[Feature]: # noqa: C901 if argch.type in STARS: features.add(feature) - # Python 2 only features (for its deprecation) except for integers, see above - elif n.type == syms.print_stmt: - features.add(Feature.PRINT_STMT) - elif n.type == syms.exec_stmt: - features.add(Feature.EXEC_STMT) - elif n.type == syms.tfpdef: - # def set_position((x, y), value): - # ... - features.add(Feature.AUTOMATIC_PARAMETER_UNPACKING) - elif n.type == syms.except_clause: - # try: - # ... - # except Exception, err: - # ... - if len(n.children) >= 4: - if n.children[-2].type == token.COMMA: - features.add(Feature.COMMA_STYLE_EXCEPT) - elif n.type == syms.raise_stmt: - # raise Exception, "msg" - if len(n.children) >= 4: - if n.children[-2].type == token.COMMA: - features.add(Feature.COMMA_STYLE_RAISE) - elif n.type == token.BACKQUOTE: - # `i'm surprised this ever existed` - features.add(Feature.BACKQUOTE_REPR) + elif ( + n.type in {syms.return_stmt, syms.yield_expr} + and len(n.children) >= 2 + and n.children[1].type == syms.testlist_star_expr + and any(child.type == syms.star_expr for child in n.children[1].children) + ): + features.add(Feature.UNPACKING_ON_FLOW) + + elif ( + n.type == syms.annassign + and len(n.children) >= 4 + and n.children[3].type == syms.testlist_star_expr + ): + features.add(Feature.ANN_ASSIGN_EXTENDED_RHS) return features -def detect_target_versions(node: Node) -> Set[TargetVersion]: +def detect_target_versions( + node: Node, *, future_imports: Optional[Set[str]] = None +) -> Set[TargetVersion]: """Detect the version to target based on the nodes used.""" - features = get_features_used(node) + features = get_features_used(node, future_imports=future_imports) return { version for version in TargetVersion if features <= VERSION_TO_FEATURES[version] } @@ -1267,13 +1355,16 @@ def get_future_imports(node: Node) -> Set[str]: return imports -def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None: +def assert_equivalent(src: str, dst: str) -> None: """Raise AssertionError if `src` and `dst` aren't equivalent.""" try: src_ast = parse_ast(src) except Exception as exc: raise AssertionError( - "cannot use --safe with this file; failed to parse source file." + "cannot use --safe with this file; failed to parse source file AST: " + f"{exc}\n" + "This could be caused by running Black with an older Python version " + "that does not support new syntax used in your source file." ) from exc try: @@ -1281,7 +1372,7 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None: except Exception as exc: log = dump_to_file("".join(traceback.format_tb(exc.__traceback__)), dst) raise AssertionError( - f"INTERNAL ERROR: Black produced invalid code on pass {pass_num}: {exc}. " + f"INTERNAL ERROR: Black produced invalid code: {exc}. " "Please report a bug on https://github.com/psf/black/issues. " f"This invalid output might be helpful: {log}" ) from None @@ -1292,14 +1383,17 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None: log = dump_to_file(diff(src_ast_str, dst_ast_str, "src", "dst")) raise AssertionError( "INTERNAL ERROR: Black produced code that is not equivalent to the" - f" source on pass {pass_num}. Please report a bug on " + " source. Please report a bug on " f"https://github.com/psf/black/issues. This diff might be helpful: {log}" ) from None def assert_stable(src: str, dst: str, mode: Mode) -> None: """Raise AssertionError if `dst` reformats differently the second time.""" - newdst = format_str(dst, mode=mode) + # We shouldn't call format_str() here, because that formats the string + # twice and may hide a bug where we bounce back and forth between two + # versions. + newdst = _format_str_once(dst, mode=mode) if dst != newdst: log = dump_to_file( str(mode), @@ -1333,13 +1427,21 @@ def patch_click() -> None: file paths is minimal since it's Python source code. Moreover, this crash was spurious on Python 3.7 thanks to PEP 538 and PEP 540. """ + modules: List[Any] = [] try: from click import core + except ImportError: + pass + else: + modules.append(core) + try: from click import _unicodefun - except ModuleNotFoundError: - return + except ImportError: + pass + else: + modules.append(_unicodefun) - for module in (core, _unicodefun): + for module in modules: if hasattr(module, "_verify_python3_env"): module._verify_python3_env = lambda: None # type: ignore if hasattr(module, "_verify_python_env"):