X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/a4c11a75e12300abfbe4c36854e450d42bdd1ee7..8e38b6626be681adbb367820ce65e05254664629:/black.py diff --git a/black.py b/black.py index 7b905ee..3ab4bc7 100644 --- a/black.py +++ b/black.py @@ -2,7 +2,7 @@ import ast import asyncio from abc import ABC, abstractmethod from collections import defaultdict -from concurrent.futures import Executor, ProcessPoolExecutor +from concurrent.futures import Executor, ThreadPoolExecutor, ProcessPoolExecutor from contextlib import contextmanager from datetime import datetime from enum import Enum @@ -34,11 +34,13 @@ from typing import ( Pattern, Sequence, Set, + Sized, Tuple, Type, TypeVar, Union, cast, + TYPE_CHECKING, ) from typing_extensions import Final from mypy_extensions import mypyc_attr @@ -59,6 +61,9 @@ from blib2to3.pgen2.parse import ParseError from _black_version import version as __version__ +if TYPE_CHECKING: + import colorama # noqa: F401 + DEFAULT_LINE_LENGTH = 88 DEFAULT_EXCLUDES = r"/(\.eggs|\.git|\.hg|\.mypy_cache|\.nox|\.tox|\.venv|\.svn|_build|buck-out|build|dist)/" # noqa: B950 DEFAULT_INCLUDES = r"\.pyi?$" @@ -140,12 +145,18 @@ class WriteBack(Enum): YES = 1 DIFF = 2 CHECK = 3 + COLOR_DIFF = 4 @classmethod - def from_configuration(cls, *, check: bool, diff: bool) -> "WriteBack": + def from_configuration( + cls, *, check: bool, diff: bool, color: bool = False + ) -> "WriteBack": if check and not diff: return cls.CHECK + if diff and color: + return cls.COLOR_DIFF + return cls.DIFF if diff else cls.YES @@ -343,15 +354,6 @@ def target_version_option_callback( " auto-detection]" ), ) -@click.option( - "--py36", - is_flag=True, - help=( - "Allow using Python 3.6-only syntax on all input files. This will put trailing" - " commas in function signatures and calls also after *args and **kwargs." - " Deprecated; use --target-version instead. [default: per-file auto-detection]" - ), -) @click.option( "--pyi", is_flag=True, @@ -380,6 +382,11 @@ def target_version_option_callback( is_flag=True, help="Don't write the files back, just output a diff for each file on stdout.", ) +@click.option( + "--color/--no-color", + is_flag=True, + help="Show colored diff. Only applies when `--diff` is given.", +) @click.option( "--fast/--safe", is_flag=True, @@ -409,6 +416,14 @@ def target_version_option_callback( ), show_default=True, ) +@click.option( + "--force-exclude", + type=str, + help=( + "Like --exclude, but files and directories matching this regex will be " + "excluded even when they are passed explicitly as arguments" + ), +) @click.option( "-q", "--quiet", @@ -458,31 +473,22 @@ def main( target_version: List[TargetVersion], check: bool, diff: bool, + color: bool, fast: bool, pyi: bool, - py36: bool, skip_string_normalization: bool, quiet: bool, verbose: bool, include: str, exclude: str, + force_exclude: Optional[str], src: Tuple[str, ...], config: Optional[str], ) -> None: """The uncompromising code formatter.""" - write_back = WriteBack.from_configuration(check=check, diff=diff) + write_back = WriteBack.from_configuration(check=check, diff=diff, color=color) if target_version: - if py36: - err("Cannot use both --target-version and --py36") - ctx.exit(2) - else: - versions = set(target_version) - elif py36: - err( - "--py36 is deprecated and will be removed in a future version. Use" - " --target-version py36 instead." - ) - versions = PY36_VERSIONS + versions = set(target_version) else: # We'll autodetect later. versions = set() @@ -497,6 +503,57 @@ def main( if code is not None: print(format_str(code, mode=mode)) ctx.exit(0) + report = Report(check=check, diff=diff, quiet=quiet, verbose=verbose) + sources = get_sources( + ctx=ctx, + src=src, + quiet=quiet, + verbose=verbose, + include=include, + exclude=exclude, + force_exclude=force_exclude, + report=report, + ) + + path_empty( + sources, + "No Python files are present to be formatted. Nothing to do 😴", + quiet, + verbose, + ctx, + ) + + if len(sources) == 1: + reformat_one( + src=sources.pop(), + fast=fast, + write_back=write_back, + mode=mode, + report=report, + ) + else: + reformat_many( + sources=sources, fast=fast, write_back=write_back, mode=mode, report=report + ) + + if verbose or not quiet: + out("Oh no! 💥 💔 💥" if report.return_code else "All done! ✨ 🍰 ✨") + click.secho(str(report), err=True) + ctx.exit(report.return_code) + + +def get_sources( + *, + ctx: click.Context, + src: Tuple[str, ...], + quiet: bool, + verbose: bool, + include: str, + exclude: str, + force_exclude: Optional[str], + report: "Report", +) -> Set[Path]: + """Compute the set of files to be formatted.""" try: include_regex = re_compile_maybe_verbose(include) except re.error: @@ -507,56 +564,56 @@ def main( except re.error: err(f"Invalid regular expression for exclude given: {exclude!r}") ctx.exit(2) - report = Report(check=check, diff=diff, quiet=quiet, verbose=verbose) + 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() - path_empty(src, quiet, verbose, ctx) + path_empty(src, "No Path provided. Nothing to do 😴", quiet, verbose, ctx) + exclude_regexes = [exclude_regex] + if force_exclude_regex is not None: + exclude_regexes.append(force_exclude_regex) + for s in src: p = Path(s) if p.is_dir(): sources.update( - gen_python_files_in_dir( - p, root, include_regex, exclude_regex, report, get_gitignore(root) + gen_python_files( + p.iterdir(), + root, + include_regex, + exclude_regexes, + report, + get_gitignore(root), ) ) - elif p.is_file() or s == "-": - # if a file was explicitly given, we don't care about its extension + elif s == "-": sources.add(p) + elif p.is_file(): + sources.update( + gen_python_files( + [p], root, None, exclude_regexes, report, get_gitignore(root) + ) + ) else: err(f"invalid path: {s}") - if len(sources) == 0: - if verbose or not quiet: - out("No Python files are present to be formatted. Nothing to do 😴") - ctx.exit(0) - - if len(sources) == 1: - reformat_one( - src=sources.pop(), - fast=fast, - write_back=write_back, - mode=mode, - report=report, - ) - else: - reformat_many( - sources=sources, fast=fast, write_back=write_back, mode=mode, report=report - ) - - if verbose or not quiet: - out("Oh no! 💥 💔 💥" if report.return_code else "All done! ✨ 🍰 ✨") - click.secho(str(report), err=True) - ctx.exit(report.return_code) + return sources def path_empty( - src: Tuple[str, ...], quiet: bool, verbose: bool, ctx: click.Context + src: Sized, msg: str, quiet: bool, verbose: bool, ctx: click.Context ) -> None: """ Exit if there is no `src` provided for formatting """ - if not src: + if len(src) == 0: if verbose or not quiet: - out("No Path provided. Nothing to do 😴") + out(msg) ctx.exit(0) @@ -597,12 +654,21 @@ def reformat_many( sources: Set[Path], fast: bool, write_back: WriteBack, mode: Mode, report: "Report" ) -> None: """Reformat multiple files using a ProcessPoolExecutor.""" + executor: Executor loop = asyncio.get_event_loop() worker_count = os.cpu_count() if sys.platform == "win32": # Work around https://bugs.python.org/issue26903 worker_count = min(worker_count, 61) - executor = ProcessPoolExecutor(max_workers=worker_count) + try: + executor = ProcessPoolExecutor(max_workers=worker_count) + except OSError: + # we arrive here if the underlying system does not support multi-processing + # like in AWS Lambda, in which case we gracefully fallback to + # a ThreadPollExecutor with just a single worker (more workers would not do us + # any good due to the Global Interpreter Lock) + executor = ThreadPoolExecutor(max_workers=1) + try: loop.run_until_complete( schedule_formatting( @@ -617,7 +683,8 @@ def reformat_many( ) finally: shutdown(loop) - executor.shutdown() + if executor is not None: + executor.shutdown() async def schedule_formatting( @@ -718,12 +785,15 @@ def format_file_in_place( if write_back == WriteBack.YES: with open(src, "w", encoding=encoding, newline=newline) as f: f.write(dst_contents) - elif write_back == WriteBack.DIFF: + elif write_back in (WriteBack.DIFF, WriteBack.COLOR_DIFF): now = datetime.utcnow() src_name = f"{src}\t{then} +0000" dst_name = f"{src}\t{now} +0000" diff_contents = diff(src_contents, dst_contents, src_name, dst_name) + if write_back == write_back.COLOR_DIFF: + diff_contents = color_diff(diff_contents) + with lock or nullcontext(): f = io.TextIOWrapper( sys.stdout.buffer, @@ -731,12 +801,57 @@ def format_file_in_place( newline=newline, write_through=True, ) + f = wrap_stream_for_windows(f) f.write(diff_contents) f.detach() return True +def color_diff(contents: str) -> str: + """Inject the ANSI color codes to the diff.""" + lines = contents.split("\n") + for i, line in enumerate(lines): + if line.startswith("+++") or line.startswith("---"): + line = "\033[1;37m" + line + "\033[0m" # bold white, reset + if line.startswith("@@"): + line = "\033[36m" + line + "\033[0m" # cyan, reset + if line.startswith("+"): + line = "\033[32m" + line + "\033[0m" # green, reset + elif line.startswith("-"): + line = "\033[31m" + line + "\033[0m" # red, reset + lines[i] = line + return "\n".join(lines) + + +def wrap_stream_for_windows( + f: io.TextIOWrapper, +) -> Union[io.TextIOWrapper, "colorama.AnsiToWin32.AnsiToWin32"]: + """ + Wrap the stream in colorama's wrap_stream so colors are shown on Windows. + + If `colorama` is not found, then no change is made. If `colorama` does + exist, then it handles the logic to determine whether or not to change + things. + """ + try: + from colorama import initialise + + # We set `strip=False` so that we can don't have to modify + # test_express_diff_with_color. + f = initialise.wrap_stream( + f, convert=None, strip=False, autoreset=False, wrap=True + ) + + # wrap_stream returns a `colorama.AnsiToWin32.AnsiToWin32` object + # which does not have a `detach()` method. So we fake one. + f.detach = lambda *args, **kwargs: None # type: ignore + except ImportError: + pass + + return f + + def format_stdin_to_stdout( fast: bool, *, write_back: WriteBack = WriteBack.NO, mode: Mode ) -> bool: @@ -762,11 +877,15 @@ def format_stdin_to_stdout( ) if write_back == WriteBack.YES: f.write(dst) - elif write_back == WriteBack.DIFF: + elif write_back in (WriteBack.DIFF, WriteBack.COLOR_DIFF): now = datetime.utcnow() src_name = f"STDIN\t{then} +0000" dst_name = f"STDOUT\t{now} +0000" - f.write(diff(src, dst, src_name, dst_name)) + d = diff(src, dst, src_name, dst_name) + if write_back == WriteBack.COLOR_DIFF: + d = color_diff(d) + f = wrap_stream_for_windows(f) + f.write(d) f.detach() @@ -2544,6 +2663,7 @@ def transform_line( is_line_short_enough(line, line_length=line_length, line_str=line_str) or line.contains_unsplittable_type_ignore() ) + and not (line.contains_standalone_comments() and line.inside_brackets) ): # Only apply basic string preprocessing, since lines shouldn't be split here. transformers = [string_merge, string_paren_strip] @@ -5155,7 +5275,7 @@ def generate_ignored_nodes(leaf: Leaf) -> Iterator[LN]: """ container: Optional[LN] = container_of(leaf) while container is not None and container.type != token.ENDMARKER: - if fmt_on(container): + if is_fmt_on(container): return # fix for fmt: on in children @@ -5169,17 +5289,21 @@ def generate_ignored_nodes(leaf: Leaf) -> Iterator[LN]: container = container.next_sibling -def fmt_on(container: LN) -> bool: - is_fmt_on = False +def is_fmt_on(container: LN) -> bool: + """Determine whether formatting is switched on within a container. + Determined by whether the last `# fmt:` comment is `on` or `off`. + """ + fmt_on = False for comment in list_comments(container.prefix, is_endmarker=False): if comment.value in FMT_ON: - is_fmt_on = True + fmt_on = True elif comment.value in FMT_OFF: - is_fmt_on = False - return is_fmt_on + fmt_on = False + return fmt_on def contains_fmt_on_at_column(container: LN, column: int) -> bool: + """Determine if children at a given column have formatting switched on.""" for child in container.children: if ( isinstance(child, Node) @@ -5187,13 +5311,14 @@ def contains_fmt_on_at_column(container: LN, column: int) -> bool: or isinstance(child, Leaf) and child.column == column ): - if fmt_on(child): + if is_fmt_on(child): return True return False def first_leaf_column(node: Node) -> Optional[int]: + """Returns the column of the first leaf child of a node.""" for child in node.children: if isinstance(child, Leaf): return child.column @@ -5624,11 +5749,11 @@ def get_gitignore(root: Path) -> PathSpec: return PathSpec.from_lines("gitwildmatch", lines) -def gen_python_files_in_dir( - path: Path, +def gen_python_files( + paths: Iterable[Path], root: Path, - include: Pattern[str], - exclude: Pattern[str], + include: Optional[Pattern[str]], + exclude_regexes: Iterable[Pattern[str]], report: "Report", gitignore: PathSpec, ) -> Iterator[Path]: @@ -5640,19 +5765,13 @@ def gen_python_files_in_dir( `report` is where output about exclusions goes. """ assert root.is_absolute(), f"INTERNAL ERROR: `root` must be absolute but is {root}" - for child in path.iterdir(): - # First ignore files matching .gitignore - if gitignore.match_file(child.as_posix()): - report.path_ignored(child, "matches the .gitignore file content") - continue - + for child in paths: # Then ignore with `exclude` option. try: - normalized_path = "/" + child.resolve().relative_to(root).as_posix() + normalized_path = child.resolve().relative_to(root).as_posix() except OSError as e: report.path_ignored(child, f"cannot be read because {e}") continue - except ValueError: if child.is_symlink(): report.path_ignored( @@ -5662,21 +5781,32 @@ def gen_python_files_in_dir( raise + # First ignore files matching .gitignore + if gitignore.match_file(normalized_path): + report.path_ignored(child, "matches the .gitignore file content") + continue + + normalized_path = "/" + normalized_path if child.is_dir(): normalized_path += "/" - exclude_match = exclude.search(normalized_path) - if exclude_match and exclude_match.group(0): - report.path_ignored(child, "matches the --exclude regular expression") + is_excluded = False + for exclude in exclude_regexes: + exclude_match = exclude.search(normalized_path) if exclude else None + if exclude_match and exclude_match.group(0): + report.path_ignored(child, "matches the --exclude regular expression") + is_excluded = True + break + if is_excluded: continue if child.is_dir(): - yield from gen_python_files_in_dir( - child, root, include, exclude, report, gitignore + yield from gen_python_files( + child.iterdir(), root, include, exclude_regexes, report, gitignore ) elif child.is_file(): - include_match = include.search(normalized_path) + include_match = include.search(normalized_path) if include else True if include_match: yield child