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 io
import keyword
import logging
from multiprocessing import Manager
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__))
YES = 1
DIFF = 2
+ @classmethod
+ def from_configuration(cls, *, check: bool, diff: bool) -> "WriteBack":
+ if check and not diff:
+ return cls.NO
+
+ return cls.DIFF if diff else cls.YES
+
class Changed(Enum):
NO = 0
YES = 2
+class FileMode(Flag):
+ AUTO_DETECT = 0
+ PYTHON36 = 1
+ PYI = 2
+ NO_STRING_NORMALIZATION = 4
+
+ @classmethod
+ def from_configuration(
+ cls, *, py36: bool, pyi: bool, skip_string_normalization: bool
+ ) -> "FileMode":
+ mode = cls.AUTO_DETECT
+ if py36:
+ mode |= cls.PYTHON36
+ if pyi:
+ mode |= cls.PYI
+ if skip_string_normalization:
+ mode |= cls.NO_STRING_NORMALIZATION
+ return mode
+
+
@click.command()
@click.option(
"-l",
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,
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",
+ "--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(
+ "-q",
+ "--quiet",
is_flag=True,
help=(
- "Consider all input files typing stubs regardless of file extension "
- "(useful when piping source on standard input)."
+ "Don't emit non-error messages to stderr. Errors are still emitted, "
+ "silence those with 2>/dev/null."
),
)
@click.option(
- "--py36",
+ "-v",
+ "--verbose",
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]"
+ "Also emit messages to stderr about files that were not changed or were "
+ "ignored due to --exclude=."
),
)
@click.version_option(version=__version__)
fast: bool,
pyi: bool,
py36: bool,
+ skip_string_normalization: bool,
quiet: bool,
+ verbose: bool,
+ include: str,
+ exclude: str,
src: List[str],
) -> None:
"""The uncompromising code formatter."""
- sources: List[Path] = []
+ write_back = WriteBack.from_configuration(check=check, diff=diff)
+ mode = FileMode.from_configuration(
+ py36=py36, pyi=pyi, skip_string_normalization=skip_string_normalization
+ )
+ report = Report(check=check, quiet=quiet, verbose=verbose)
+ sources: Set[Path] = set()
+ 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))
- elif p.is_file():
+ sources.update(
+ gen_python_files_in_dir(p, root, include_regex, exclude_regex, report)
+ )
+ elif p.is_file() or s == "-":
# if a file was explicitly given, we don't care about its extension
- sources.append(p)
- elif s == "-":
- sources.append(Path("-"))
+ sources.add(p)
else:
err(f"invalid path: {s}")
-
- if check and not diff:
- write_back = WriteBack.NO
- elif diff:
- write_back = WriteBack.DIFF
- else:
- write_back = WriteBack.YES
- report = Report(check=check, quiet=quiet)
if len(sources) == 0:
- out("No paths given. Nothing to do 😴")
+ if verbose or not quiet:
+ out("No paths given. Nothing to do 😴")
ctx.exit(0)
return
elif len(sources) == 1:
reformat_one(
- src=sources[0],
+ src=sources.pop(),
line_length=line_length,
fast=fast,
- pyi=pyi,
- py36=py36,
write_back=write_back,
+ mode=mode,
report=report,
)
else:
sources=sources,
line_length=line_length,
fast=fast,
- pyi=pyi,
- py36=py36,
write_back=write_back,
+ mode=mode,
report=report,
loop=loop,
executor=executor,
)
finally:
shutdown(loop)
- if not quiet:
+ if verbose or not quiet:
out("All done! ✨ 🍰 ✨")
click.echo(str(report))
ctx.exit(report.return_code)
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.
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))
async def schedule_formatting(
- sources: List[Path],
+ sources: Set[Path],
line_length: int,
fast: bool,
- pyi: bool,
- py36: bool,
write_back: WriteBack,
+ mode: FileMode,
report: "Report",
loop: BaseEventLoop,
executor: Executor,
"""
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:
+ for src in sorted(cached):
report.done(src, Changed.CACHED)
cancelled = []
formatted = []
src,
line_length,
fast,
- pyi,
- py36,
write_back,
+ mode,
lock,
): src
for src in sorted(sources)
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.
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()
+ with open(src, "rb") as buf:
+ newline, encoding, src_contents = prepare_input(buf.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
if write_back == write_back.YES:
- with open(src, "w", encoding=src_buffer.encoding) as f:
+ with open(src, "w", encoding=encoding, newline=newline) as f:
f.write(dst_contents)
elif write_back == write_back.DIFF:
src_name = f"{src} (original)"
if lock:
lock.acquire()
try:
- sys.stdout.write(diff_contents)
+ f = io.TextIOWrapper(
+ sys.stdout.buffer,
+ encoding=encoding,
+ newline=newline,
+ write_through=True,
+ )
+ f.write(diff_contents)
+ f.detach()
finally:
if lock:
lock.release()
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.
`line_length`, `fast`, `is_pyi`, and `force_py36` arguments are passed to
:func:`format_file_contents`.
"""
- src = sys.stdin.read()
+ newline, encoding, src = prepare_input(sys.stdin.buffer.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:
finally:
if write_back == WriteBack.YES:
- sys.stdout.write(dst)
+ f = io.TextIOWrapper(
+ sys.stdout.buffer,
+ encoding=encoding,
+ newline=newline,
+ write_through=True,
+ )
+ f.write(dst)
+ f.detach()
elif write_back == WriteBack.DIFF:
src_name = "<stdin> (original)"
dst_name = "<stdin> (formatted)"
- sys.stdout.write(diff(src, dst, src_name, dst_name))
+ f = io.TextIOWrapper(
+ sys.stdout.buffer,
+ encoding=encoding,
+ newline=newline,
+ write_through=True,
+ )
+ f.write(diff(src, dst, src_name, dst_name))
+ f.detach()
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.
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.
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):
return dst_contents
+def prepare_input(src: bytes) -> Tuple[str, str, str]:
+ """Analyze `src` and return a tuple of (newline, encoding, decoded_contents)
+
+ Where `newline` is either CRLF or LF, and `decoded_contents` is decoded with
+ universal newlines (i.e. only LF).
+ """
+ srcbuf = io.BytesIO(src)
+ encoding, lines = tokenize.detect_encoding(srcbuf.readline)
+ newline = "\r\n" if b"\r\n" == lines[0][-2:] else "\n"
+ srcbuf.seek(0)
+ return newline, encoding, io.TextIOWrapper(srcbuf, encoding).read()
+
+
GRAMMARS = [
pygram.python_grammar_no_print_statement_no_exec_statement,
pygram.python_grammar_no_print_statement,
"""Given a string with source, return the lib2to3 Node."""
grammar = pygram.python_grammar_no_print_statement
if src_txt[-1] != "\n":
- nl = "\r\n" if "\r\n" in src_txt[:1024] else "\n"
- src_txt += nl
+ src_txt += "\n"
for grammar in GRAMMARS:
drv = driver.Driver(grammar, pytree.convert)
try:
"""
is_pyi: bool = False
+ normalize_strings: bool = True
current_line: Line = Factory(Line)
remove_u_prefix: bool = False
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:
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,
+ root: Path,
+ include: Pattern[str],
+ exclude: Pattern[str],
+ report: "Report",
+) -> Iterator[Path]:
+ """Generate all files under `path` whose paths are not excluded by the
+ `exclude` regex, but are included by the `include` regex.
-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.
+ `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():
+ 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):
+ report.path_ignored(child, f"matches --exclude={exclude.pattern}")
+ continue
+
+ if child.is_dir():
+ yield from gen_python_files_in_dir(child, root, include, exclude, report)
+
+ 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()
+
+ 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
+
+ if (directory / ".hg").is_dir():
+ return directory
- yield from gen_python_files_in_dir(child)
+ if (directory / "pyproject.toml").is_file():
+ return directory
- elif child.is_file() and child.suffix in PYTHON_EXTENSIONS:
- yield child
+ return directory
@dataclass
check: bool = False
quiet: bool = False
+ verbose: bool = False
change_count: int = 0
same_count: int = 0
failure_count: int = 0
"""Increment the counter for successful reformatting. Write out a message."""
if changed is Changed.YES:
reformatted = "would reformat" if self.check else "reformatted"
- if not self.quiet:
+ if self.verbose or not self.quiet:
out(f"{reformatted} {src}")
self.change_count += 1
else:
- if not self.quiet:
+ if self.verbose:
if changed is Changed.NO:
msg = f"{src} already well formatted, good job."
else:
err(f"error: cannot format {src}: {message}")
self.failure_count += 1
+ def path_ignored(self, path: Path, message: str) -> None:
+ if self.verbose:
+ out(f"{path} ignored: {message}", bold=False)
+
@property
def return_code(self) -> int:
"""Return the exit code that the app should use.
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"),
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 {}
return stat.st_mtime, stat.st_size
-def filter_cached(
- cache: Cache, sources: Iterable[Path]
-) -> Tuple[List[Path], List[Path]]:
- """Split a list of paths into two.
+def filter_cached(cache: Cache, sources: Iterable[Path]) -> Tuple[Set[Path], Set[Path]]:
+ """Split an iterable of paths in `sources` into two sets.
- The first list contains paths of files that modified on disk or are not in the
- cache. The other list contains paths to non-modified files.
+ The first contains paths of files that modified on disk or are not in the
+ cache. The other contains paths to non-modified files.
"""
- todo, done = [], []
+ todo, done = set(), set()
for src in sources:
src = src.resolve()
if cache.get(src) != get_cache_info(src):
- todo.append(src)
+ todo.add(src)
else:
- done.append(src)
+ done.add(src)
return todo, done
def write_cache(
- cache: Cache,
- sources: List[Path],
- line_length: int,
- pyi: bool = False,
- py36: bool = False,
+ cache: Cache, sources: Iterable[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)
projects / etc / vim.git / blobdiff