X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/07b1b2f3dd0e4f13dd5df96eb9188f3d02f5726e..4cb338e3dd1fd0a1b1725c764ed21d5aeeb0ac69:/black.py diff --git a/black.py b/black.py index 0aba2c5..730c64d 100644 --- a/black.py +++ b/black.py @@ -2,7 +2,7 @@ import asyncio import pickle from asyncio.base_events import BaseEventLoop from concurrent.futures import Executor, ProcessPoolExecutor -from enum import Enum +from enum import Enum, Flag from functools import partial, wraps import keyword import logging @@ -44,8 +44,12 @@ from blib2to3.pgen2 import driver, token from blib2to3.pgen2.parse import ParseError -__version__ = "18.5b0" +__version__ = "18.5b1" DEFAULT_LINE_LENGTH = 88 +DEFAULT_EXCLUDES = ( + r"/(\.git|\.hg|\.mypy_cache|\.tox|\.venv|_build|buck-out|build|dist)/" +) +DEFAULT_INCLUDES = r"\.pyi?$" CACHE_DIR = Path(user_cache_dir("black", version=__version__)) @@ -122,6 +126,13 @@ class Changed(Enum): YES = 2 +class FileMode(Flag): + AUTO_DETECT = 0 + PYTHON36 = 1 + PYI = 2 + NO_STRING_NORMALIZATION = 4 + + @click.command() @click.option( "-l", @@ -131,6 +142,29 @@ class Changed(Enum): help="How many character per line to allow.", show_default=True, ) +@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. [default: per-file auto-detection]" + ), +) +@click.option( + "--pyi", + is_flag=True, + help=( + "Format all input files like typing stubs regardless of file extension " + "(useful when piping source on standard input)." + ), +) +@click.option( + "-S", + "--skip-string-normalization", + is_flag=True, + help="Don't normalize string quotes or prefixes.", +) @click.option( "--check", is_flag=True, @@ -151,25 +185,37 @@ class Changed(Enum): help="If --fast given, skip temporary sanity checks. [default: --safe]", ) @click.option( - "-q", - "--quiet", - is_flag=True, + "--include", + type=str, + default=DEFAULT_INCLUDES, help=( - "Don't emit non-error messages to stderr. Errors are still emitted, " - "silence those with 2>/dev/null." + "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." ), + show_default=True, ) @click.option( - "--pyi", - is_flag=True, - help="Force pyi (stub file) formatting, regardless of file extension.", + "--exclude", + type=str, + default=DEFAULT_EXCLUDES, + 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). " + "Exclusions are calculated first, inclusions later." + ), + show_default=True, ) @click.option( - "--py36", + "-q", + "--quiet", is_flag=True, help=( - "Force Python 3.6 mode, even if file doesn't currently use " - "Python 3.6-only syntax." + "Don't emit non-error messages to stderr. Errors are still emitted, " + "silence those with 2>/dev/null." ), ) @click.version_option(version=__version__) @@ -189,15 +235,31 @@ def main( fast: bool, pyi: bool, py36: bool, + skip_string_normalization: bool, quiet: bool, + include: str, + exclude: str, src: List[str], ) -> None: """The uncompromising code formatter.""" sources: List[Path] = [] + try: + include_regex = re.compile(include) + except re.error: + err(f"Invalid regular expression for include given: {include!r}") + ctx.exit(2) + try: + exclude_regex = re.compile(exclude) + except re.error: + err(f"Invalid regular expression for exclude given: {exclude!r}") + ctx.exit(2) + root = find_project_root(src) for s in src: p = Path(s) if p.is_dir(): - sources.extend(gen_python_files_in_dir(p)) + sources.extend( + gen_python_files_in_dir(p, root, include_regex, exclude_regex) + ) elif p.is_file(): # if a file was explicitly given, we don't care about its extension sources.append(p) @@ -212,6 +274,13 @@ def main( write_back = WriteBack.DIFF else: write_back = WriteBack.YES + mode = FileMode.AUTO_DETECT + if py36: + mode |= FileMode.PYTHON36 + if pyi: + mode |= FileMode.PYI + if skip_string_normalization: + mode |= FileMode.NO_STRING_NORMALIZATION report = Report(check=check, quiet=quiet) if len(sources) == 0: out("No paths given. Nothing to do 😴") @@ -223,9 +292,8 @@ def main( src=sources[0], line_length=line_length, fast=fast, - pyi=pyi, - py36=py36, write_back=write_back, + mode=mode, report=report, ) else: @@ -237,9 +305,8 @@ def main( sources=sources, line_length=line_length, fast=fast, - pyi=pyi, - py36=py36, write_back=write_back, + mode=mode, report=report, loop=loop, executor=executor, @@ -257,9 +324,8 @@ def reformat_one( src: Path, line_length: int, fast: bool, - pyi: bool, - py36: bool, write_back: WriteBack, + mode: FileMode, report: "Report", ) -> None: """Reformat a single file under `src` without spawning child processes. @@ -272,31 +338,26 @@ def reformat_one( changed = Changed.NO if not src.is_file() and str(src) == "-": if format_stdin_to_stdout( - line_length=line_length, - fast=fast, - is_pyi=pyi, - force_py36=py36, - write_back=write_back, + line_length=line_length, fast=fast, write_back=write_back, mode=mode ): changed = Changed.YES else: cache: Cache = {} if write_back != WriteBack.DIFF: - cache = read_cache(line_length, pyi, py36) - src = src.resolve() - if src in cache and cache[src] == get_cache_info(src): + cache = read_cache(line_length, mode) + res_src = src.resolve() + if res_src in cache and cache[res_src] == get_cache_info(res_src): changed = Changed.CACHED if changed is not Changed.CACHED and format_file_in_place( src, line_length=line_length, fast=fast, - force_pyi=pyi, - force_py36=py36, write_back=write_back, + mode=mode, ): changed = Changed.YES if write_back == WriteBack.YES and changed is not Changed.NO: - write_cache(cache, [src], line_length, pyi, py36) + write_cache(cache, [src], line_length, mode) report.done(src, changed) except Exception as exc: report.failed(src, str(exc)) @@ -306,9 +367,8 @@ async def schedule_formatting( sources: List[Path], line_length: int, fast: bool, - pyi: bool, - py36: bool, write_back: WriteBack, + mode: FileMode, report: "Report", loop: BaseEventLoop, executor: Executor, @@ -322,7 +382,7 @@ async def schedule_formatting( """ cache: Cache = {} if write_back != WriteBack.DIFF: - cache = read_cache(line_length, pyi, py36) + cache = read_cache(line_length, mode) sources, cached = filter_cached(cache, sources) for src in cached: report.done(src, Changed.CACHED) @@ -342,9 +402,8 @@ async def schedule_formatting( src, line_length, fast, - pyi, - py36, write_back, + mode, lock, ): src for src in sorted(sources) @@ -370,16 +429,15 @@ async def schedule_formatting( if cancelled: await asyncio.gather(*cancelled, loop=loop, return_exceptions=True) if write_back == WriteBack.YES and formatted: - write_cache(cache, formatted, line_length, pyi, py36) + write_cache(cache, formatted, line_length, mode) def format_file_in_place( src: Path, line_length: int, fast: bool, - force_pyi: bool = False, - force_py36: bool = False, write_back: WriteBack = WriteBack.NO, + mode: FileMode = FileMode.AUTO_DETECT, lock: Any = None, # multiprocessing.Manager().Lock() is some crazy proxy ) -> bool: """Format file under `src` path. Return True if changed. @@ -387,17 +445,13 @@ def format_file_in_place( If `write_back` is True, write reformatted code back to stdout. `line_length` and `fast` options are passed to :func:`format_file_contents`. """ - is_pyi = force_pyi or src.suffix == ".pyi" - + if src.suffix == ".pyi": + mode |= FileMode.PYI with tokenize.open(src) as src_buffer: src_contents = src_buffer.read() try: dst_contents = format_file_contents( - src_contents, - line_length=line_length, - fast=fast, - is_pyi=is_pyi, - force_py36=force_py36, + src_contents, line_length=line_length, fast=fast, mode=mode ) except NothingChanged: return False @@ -422,9 +476,8 @@ def format_file_in_place( def format_stdin_to_stdout( line_length: int, fast: bool, - is_pyi: bool = False, - force_py36: bool = False, write_back: WriteBack = WriteBack.NO, + mode: FileMode = FileMode.AUTO_DETECT, ) -> bool: """Format file on stdin. Return True if changed. @@ -435,13 +488,7 @@ def format_stdin_to_stdout( src = sys.stdin.read() dst = src try: - dst = format_file_contents( - src, - line_length=line_length, - fast=fast, - is_pyi=is_pyi, - force_py36=force_py36, - ) + dst = format_file_contents(src, line_length=line_length, fast=fast, mode=mode) return True except NothingChanged: @@ -461,8 +508,7 @@ def format_file_contents( *, line_length: int, fast: bool, - is_pyi: bool = False, - force_py36: bool = False, + mode: FileMode = FileMode.AUTO_DETECT, ) -> FileContent: """Reformat contents a file and return new contents. @@ -473,30 +519,18 @@ def format_file_contents( if src_contents.strip() == "": raise NothingChanged - dst_contents = format_str( - src_contents, line_length=line_length, is_pyi=is_pyi, force_py36=force_py36 - ) + dst_contents = format_str(src_contents, line_length=line_length, mode=mode) if src_contents == dst_contents: raise NothingChanged if not fast: assert_equivalent(src_contents, dst_contents) - assert_stable( - src_contents, - dst_contents, - line_length=line_length, - is_pyi=is_pyi, - force_py36=force_py36, - ) + assert_stable(src_contents, dst_contents, line_length=line_length, mode=mode) return dst_contents def format_str( - src_contents: str, - line_length: int, - *, - is_pyi: bool = False, - force_py36: bool = False, + src_contents: str, line_length: int, *, mode: FileMode = FileMode.AUTO_DETECT ) -> FileContent: """Reformat a string and return new contents. @@ -505,11 +539,15 @@ def format_str( src_node = lib2to3_parse(src_contents) dst_contents = "" future_imports = get_future_imports(src_node) - elt = EmptyLineTracker(is_pyi=is_pyi) - py36 = force_py36 or is_python36(src_node) + is_pyi = bool(mode & FileMode.PYI) + py36 = bool(mode & FileMode.PYTHON36) or is_python36(src_node) + normalize_strings = not bool(mode & FileMode.NO_STRING_NORMALIZATION) lines = LineGenerator( - remove_u_prefix=py36 or "unicode_literals" in future_imports, is_pyi=is_pyi + remove_u_prefix=py36 or "unicode_literals" in future_imports, + is_pyi=is_pyi, + normalize_strings=normalize_strings, ) + elt = EmptyLineTracker(is_pyi=is_pyi) empty_line = Line() after = 0 for current_line in lines.visit(src_node): @@ -967,27 +1005,6 @@ class Line: and second_leaf.value == "def" ) - @property - def is_flow_control(self) -> bool: - """Is this line a flow control statement? - - Those are `return`, `raise`, `break`, and `continue`. - """ - return ( - bool(self) - and self.leaves[0].type == token.NAME - and self.leaves[0].value in FLOW_CONTROL - ) - - @property - def is_yield(self) -> bool: - """Is this line a yield statement?""" - return ( - bool(self) - and self.leaves[0].type == token.NAME - and self.leaves[0].value == "yield" - ) - @property def is_class_paren_empty(self) -> bool: """Is this a class with no base classes but using parentheses? @@ -1004,6 +1021,15 @@ class Line: and self.leaves[3].value == ")" ) + @property + def is_triple_quoted_string(self) -> bool: + """Is the line a triple quoted string?""" + return ( + bool(self) + and self.leaves[0].type == token.STRING + and self.leaves[0].value.startswith(('"""', "'''")) + ) + def contains_standalone_comments(self, depth_limit: int = sys.maxsize) -> bool: """If so, needs to be split before emitting.""" for leaf in self.leaves: @@ -1211,6 +1237,7 @@ class EmptyLineTracker: the prefix of the first leaf consists of optional newlines. Those newlines are consumed by `maybe_empty_lines()` and included in the computation. """ + is_pyi: bool = False previous_line: Optional[Line] = None previous_after: int = 0 @@ -1220,8 +1247,7 @@ class EmptyLineTracker: """Return the number of extra empty lines before and after the `current_line`. This is for separating `def`, `async def` and `class` with extra empty - lines (two on module-level), as well as providing an extra empty line - after flow control keywords to make them more prominent. + lines (two on module-level). """ if isinstance(current_line, UnformattedLines): return 0, 0 @@ -1262,6 +1288,11 @@ class EmptyLineTracker: if self.previous_line.is_decorator: return 0, 0 + if self.previous_line.depth < current_line.depth and ( + self.previous_line.is_class or self.previous_line.is_def + ): + return 0, 0 + if ( self.previous_line.is_comment and self.previous_line.depth == current_line.depth @@ -1293,6 +1324,13 @@ class EmptyLineTracker: ): return (before or 1), 0 + if ( + self.previous_line + and self.previous_line.is_class + and current_line.is_triple_quoted_string + ): + return before, 1 + return before, 0 @@ -1303,7 +1341,9 @@ class LineGenerator(Visitor[Line]): Note: destroys the tree it's visiting by mutating prefixes of its leaves in ways that will no longer stringify to valid Python code on the tree. """ + is_pyi: bool = False + normalize_strings: bool = True current_line: Line = Factory(Line) remove_u_prefix: bool = False @@ -1372,7 +1412,7 @@ class LineGenerator(Visitor[Line]): else: normalize_prefix(node, inside_brackets=any_open_brackets) - if node.type == token.STRING: + if self.normalize_strings and node.type == token.STRING: normalize_string_prefix(node, remove_u_prefix=self.remove_u_prefix) normalize_string_quotes(node) if node.type not in WHITESPACE: @@ -2754,38 +2794,63 @@ def get_future_imports(node: Node) -> Set[str]: return imports -PYTHON_EXTENSIONS = {".py", ".pyi"} -BLACKLISTED_DIRECTORIES = { - "build", - "buck-out", - "dist", - "_build", - ".git", - ".hg", - ".mypy_cache", - ".tox", - ".venv", -} - - -def gen_python_files_in_dir(path: Path) -> Iterator[Path]: - """Generate all files under `path` which aren't under BLACKLISTED_DIRECTORIES - and have one of the PYTHON_EXTENSIONS. +def gen_python_files_in_dir( + path: Path, root: Path, include: Pattern[str], exclude: Pattern[str] +) -> Iterator[Path]: + """Generate all files under `path` whose paths are not excluded by the + `exclude` regex, but are included by the `include` regex. """ + assert root.is_absolute(), f"INTERNAL ERROR: `root` must be absolute but is {root}" for child in path.iterdir(): + normalized_path = child.resolve().relative_to(root).as_posix() if child.is_dir(): - if child.name in BLACKLISTED_DIRECTORIES: - continue + normalized_path += "/" + exclude_match = exclude.search(normalized_path) + if exclude_match and exclude_match.group(0): + continue + + if child.is_dir(): + yield from gen_python_files_in_dir(child, root, include, exclude) + + elif child.is_file(): + include_match = include.search(normalized_path) + if include_match: + yield child + + +def find_project_root(srcs: List[str]) -> Path: + """Return a directory containing .git, .hg, or pyproject.toml. + + That directory can be one of the directories passed in `srcs` or their + common parent. + + If no directory in the tree contains a marker that would specify it's the + project root, the root of the file system is returned. + """ + if not srcs: + return Path("/").resolve() - yield from gen_python_files_in_dir(child) + common_base = min(Path(src).resolve() for src in srcs) + if common_base.is_dir(): + # Append a fake file so `parents` below returns `common_base_dir`, too. + common_base /= "fake-file" + for directory in common_base.parents: + if (directory / ".git").is_dir(): + return directory - elif child.is_file() and child.suffix in PYTHON_EXTENSIONS: - yield child + if (directory / ".hg").is_dir(): + return directory + + if (directory / "pyproject.toml").is_file(): + return directory + + return directory @dataclass class Report: """Provides a reformatting counter. Can be rendered with `str(report)`.""" + check: bool = False quiet: bool = False change_count: int = 0 @@ -2926,12 +2991,10 @@ def assert_equivalent(src: str, dst: str) -> None: def assert_stable( - src: str, dst: str, line_length: int, is_pyi: bool = False, force_py36: bool = False + src: str, dst: str, line_length: int, mode: FileMode = FileMode.AUTO_DETECT ) -> None: """Raise AssertionError if `dst` reformats differently the second time.""" - newdst = format_str( - dst, line_length=line_length, is_pyi=is_pyi, force_py36=force_py36 - ) + newdst = format_str(dst, line_length=line_length, mode=mode) if dst != newdst: log = dump_to_file( diff(src, dst, "source", "first pass"), @@ -3142,19 +3205,21 @@ def can_omit_invisible_parens(line: Line, line_length: int) -> bool: return False -def get_cache_file(line_length: int, pyi: bool = False, py36: bool = False) -> Path: +def get_cache_file(line_length: int, mode: FileMode) -> Path: + pyi = bool(mode & FileMode.PYI) + py36 = bool(mode & FileMode.PYTHON36) return ( CACHE_DIR / f"cache.{line_length}{'.pyi' if pyi else ''}{'.py36' if py36 else ''}.pickle" ) -def read_cache(line_length: int, pyi: bool = False, py36: bool = False) -> Cache: +def read_cache(line_length: int, mode: FileMode) -> Cache: """Read the cache if it exists and is well formed. If it is not well formed, the call to write_cache later should resolve the issue. """ - cache_file = get_cache_file(line_length, pyi, py36) + cache_file = get_cache_file(line_length, mode) if not cache_file.exists(): return {} @@ -3192,14 +3257,10 @@ def filter_cached( def write_cache( - cache: Cache, - sources: List[Path], - line_length: int, - pyi: bool = False, - py36: bool = False, + cache: Cache, sources: List[Path], line_length: int, mode: FileMode ) -> None: """Update the cache file.""" - cache_file = get_cache_file(line_length, pyi, py36) + cache_file = get_cache_file(line_length, mode) try: if not CACHE_DIR.exists(): CACHE_DIR.mkdir(parents=True)