]> git.madduck.net Git - etc/vim.git/blobdiff - src/black/__init__.py

madduck's git repository

Every one of the projects in this repository is available at the canonical URL git://git.madduck.net/madduck/pub/<projectpath> — see each project's metadata for the exact URL.

All patches and comments are welcome. Please squash your changes to logical commits before using git-format-patch and git-send-email to patches@git.madduck.net. If you'd read over the Git project's submission guidelines and adhered to them, I'd be especially grateful.

SSH access, as well as push access can be individually arranged.

If you use my repositories frequently, consider adding the following snippet to ~/.gitconfig and using the third clone URL listed for each project:

[url "git://git.madduck.net/madduck/"]
  insteadOf = madduck:

Bump sphinx from 4.3.2 to 4.4.0 in /docs (#2776)
[etc/vim.git] / src / black / __init__.py
index 60f4fa34e9ddf28ad097be35704b2d3e3790f3a1..fa918ce293172239fb87db2e0bfb600e836494d0 100644 (file)
@@ -10,7 +10,7 @@ from multiprocessing import Manager, freeze_support
 import os
 from pathlib import Path
 from pathspec.patterns.gitwildmatch import GitWildMatchPatternError
 import os
 from pathlib import Path
 from pathspec.patterns.gitwildmatch import GitWildMatchPatternError
-import regex as re
+import re
 import signal
 import sys
 import tokenize
 import signal
 import sys
 import tokenize
@@ -30,16 +30,19 @@ from typing import (
     Union,
 )
 
     Union,
 )
 
-from dataclasses import replace
 import click
 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.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.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.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
@@ -56,6 +59,7 @@ from black.handle_ipynb_magics import (
     remove_trailing_semicolon,
     put_trailing_semicolon_back,
     TRANSFORMED_MAGICS,
     remove_trailing_semicolon,
     put_trailing_semicolon_back,
     TRANSFORMED_MAGICS,
+    PYTHON_CELL_MAGICS,
     jupyter_dependencies_are_installed,
 )
 
     jupyter_dependencies_are_installed,
 )
 
@@ -66,6 +70,8 @@ from blib2to3.pgen2 import token
 
 from _black_version import version as __version__
 
 
 from _black_version import version as __version__
 
+COMPILED = Path(__file__).suffix in (".pyd", ".so")
+
 # types
 FileContent = str
 Encoding = str
 # types
 FileContent = str
 Encoding = str
@@ -95,6 +101,8 @@ class WriteBack(Enum):
 # Legacy name, left for integrations.
 FileMode = Mode
 
 # Legacy name, left for integrations.
 FileMode = Mode
 
+DEFAULT_WORKERS = os.cpu_count()
+
 
 def read_pyproject_toml(
     ctx: click.Context, param: click.Parameter, value: Optional[str]
 
 def read_pyproject_toml(
     ctx: click.Context, param: click.Parameter, value: Optional[str]
@@ -114,7 +122,7 @@ def read_pyproject_toml(
     except (OSError, ValueError) as e:
         raise click.FileError(
             filename=value, hint=f"Error reading configuration file: {e}"
     except (OSError, ValueError) as e:
         raise click.FileError(
             filename=value, hint=f"Error reading configuration file: {e}"
-        )
+        ) from None
 
     if not config:
         return None
 
     if not config:
         return None
@@ -168,14 +176,19 @@ def validate_regex(
     ctx: click.Context,
     param: click.Parameter,
     value: Optional[str],
     ctx: click.Context,
     param: click.Parameter,
     value: Optional[str],
-) -> Optional[Pattern]:
+) -> Optional[Pattern[str]]:
     try:
         return re_compile_maybe_verbose(value) if value is not None else None
     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")
+    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",
 @click.option("-c", "--code", type=str, help="Format the code passed in as a string.")
 @click.option(
     "-l",
@@ -318,6 +331,14 @@ def validate_regex(
         "editors that rely on using stdin."
     ),
 )
         "editors that rely on using stdin."
     ),
 )
+@click.option(
+    "-W",
+    "--workers",
+    type=click.IntRange(min=1),
+    default=DEFAULT_WORKERS,
+    show_default=True,
+    help="Number of parallel workers",
+)
 @click.option(
     "-q",
     "--quiet",
 @click.option(
     "-q",
     "--quiet",
@@ -336,7 +357,10 @@ def validate_regex(
         " due to exclusion patterns."
     ),
 )
         " 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,
 @click.argument(
     "src",
     nargs=-1,
@@ -377,18 +401,48 @@ def main(
     experimental_string_processing: bool,
     quiet: bool,
     verbose: bool,
     experimental_string_processing: bool,
     quiet: bool,
     verbose: bool,
-    required_version: str,
-    include: Pattern,
-    exclude: Optional[Pattern],
-    extend_exclude: Optional[Pattern],
-    force_exclude: Optional[Pattern],
+    required_version: Optional[str],
+    include: Pattern[str],
+    exclude: Optional[Pattern[str]],
+    extend_exclude: Optional[Pattern[str]],
+    force_exclude: Optional[Pattern[str]],
     stdin_filename: Optional[str],
     stdin_filename: Optional[str],
+    workers: int,
     src: Tuple[str, ...],
     config: Optional[str],
 ) -> None:
     """The uncompromising code formatter."""
     src: Tuple[str, ...],
     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)
+    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")
+            if 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__:
 
     error_msg = "Oh no! 💥 💔 💥"
     if required_version and required_version != __version__:
@@ -468,9 +522,12 @@ def main(
                 write_back=write_back,
                 mode=mode,
                 report=report,
                 write_back=write_back,
                 mode=mode,
                 report=report,
+                workers=workers,
             )
 
     if verbose or not quiet:
             )
 
     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)
         out(error_msg if report.return_code else "All done! ✨ 🍰 ✨")
         if code is None:
             click.echo(str(report), err=True)
@@ -491,14 +548,12 @@ def get_sources(
     stdin_filename: Optional[str],
 ) -> Set[Path]:
     """Compute the set of files to be formatted."""
     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)
     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
 
     else:
         gitignore = None
 
@@ -511,7 +566,7 @@ def get_sources(
             is_stdin = False
 
         if is_stdin or p.is_file():
             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
 
             if normalized_path is None:
                 continue
 
@@ -538,7 +593,7 @@ def get_sources(
             sources.update(
                 gen_python_files(
                     p.iterdir(),
             sources.update(
                 gen_python_files(
                     p.iterdir(),
-                    root,
+                    ctx.obj["root"],
                     include,
                     exclude,
                     extend_exclude,
                     include,
                     exclude,
                     extend_exclude,
@@ -643,19 +698,28 @@ def reformat_one(
         report.failed(src, str(exc))
 
 
         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(
 def reformat_many(
-    sources: Set[Path], fast: bool, write_back: WriteBack, mode: Mode, report: "Report"
+    sources: Set[Path],
+    fast: bool,
+    write_back: WriteBack,
+    mode: Mode,
+    report: "Report",
+    workers: Optional[int],
 ) -> None:
     """Reformat multiple files using a ProcessPoolExecutor."""
     executor: Executor
     loop = asyncio.get_event_loop()
 ) -> None:
     """Reformat multiple files using a ProcessPoolExecutor."""
     executor: Executor
     loop = asyncio.get_event_loop()
-    worker_count = os.cpu_count()
+    worker_count = workers if workers is not None else DEFAULT_WORKERS
     if sys.platform == "win32":
         # Work around https://bugs.python.org/issue26903
     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)
         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
         # 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
@@ -746,7 +810,10 @@ async def schedule_formatting(
                     sources_to_cache.append(src)
                 report.done(src, changed)
     if cancelled:
                     sources_to_cache.append(src)
                 report.done(src, changed)
     if cancelled:
-        await asyncio.gather(*cancelled, loop=loop, return_exceptions=True)
+        if sys.version_info >= (3, 7):
+            await asyncio.gather(*cancelled, return_exceptions=True)
+        else:
+            await asyncio.gather(*cancelled, loop=loop, return_exceptions=True)
     if sources_to_cache:
         write_cache(cache, sources_to_cache, mode)
 
     if sources_to_cache:
         write_cache(cache, sources_to_cache, mode)
 
@@ -777,7 +844,9 @@ def format_file_in_place(
     except NothingChanged:
         return False
     except JSONDecodeError:
     except NothingChanged:
         return False
     except JSONDecodeError:
-        raise ValueError(f"File '{src}' cannot be parsed as valid Jupyter notebook.")
+        raise ValueError(
+            f"File '{src}' cannot be parsed as valid Jupyter notebook."
+        ) from None
 
     if write_back == WriteBack.YES:
         with open(src, "w", encoding=encoding, newline=newline) as f:
 
     if write_back == WriteBack.YES:
         with open(src, "w", encoding=encoding, newline=newline) as f:
@@ -906,7 +975,9 @@ def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileCo
 
 
 def validate_cell(src: str) -> None:
 
 
 def validate_cell(src: str) -> None:
-    """Check that cell does not already contain TransformerManager transformations.
+    """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
 
     If a cell contains ``!ls``, then it'll be transformed to
     ``get_ipython().system('ls')``. However, if the cell originally contained
@@ -922,6 +993,8 @@ def validate_cell(src: str) -> None:
     """
     if any(transformed_magic in src for transformed_magic in TRANSFORMED_MAGICS):
         raise NothingChanged
     """
     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:
+        raise NothingChanged
 
 
 def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
 
 
 def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
@@ -947,7 +1020,7 @@ def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
     try:
         masked_src, replacements = mask_cell(src_without_trailing_semicolon)
     except SyntaxError:
     try:
         masked_src, replacements = mask_cell(src_without_trailing_semicolon)
     except SyntaxError:
-        raise NothingChanged
+        raise NothingChanged from None
     masked_dst = format_str(masked_src, mode=mode)
     if not fast:
         check_stability_and_equivalence(masked_src, masked_dst, mode=mode)
     masked_dst = format_str(masked_src, mode=mode)
     if not fast:
         check_stability_and_equivalence(masked_src, masked_dst, mode=mode)
@@ -957,7 +1030,7 @@ def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
     )
     dst = dst.rstrip("\n")
     if dst == src:
     )
     dst = dst.rstrip("\n")
     if dst == src:
-        raise NothingChanged
+        raise NothingChanged from None
     return dst
 
 
     return dst
 
 
@@ -970,14 +1043,14 @@ def validate_metadata(nb: MutableMapping[str, Any]) -> None:
     """
     language = nb.get("metadata", {}).get("language_info", {}).get("name", None)
     if language is not None and language != "python":
     """
     language = nb.get("metadata", {}).get("language_info", {}).get("name", None)
     if language is not None and language != "python":
-        raise NothingChanged
+        raise NothingChanged from None
 
 
 def format_ipynb_string(src_contents: str, *, fast: bool, mode: Mode) -> FileContent:
     """Format Jupyter notebook.
 
     Operate cell-by-cell, only on code cells, only for Python notebooks.
 
 
 def format_ipynb_string(src_contents: str, *, fast: bool, mode: Mode) -> FileContent:
     """Format Jupyter notebook.
 
     Operate cell-by-cell, only on code cells, only for Python notebooks.
-    If the ``.ipynb`` originally had a trailing newline, it'll be preseved.
+    If the ``.ipynb`` originally had a trailing newline, it'll be preserved.
     """
     trailing_newline = src_contents[-1] == "\n"
     modified = False
     """
     trailing_newline = src_contents[-1] == "\n"
     modified = False
@@ -1038,13 +1111,10 @@ def format_str(src_contents: str, *, mode: Mode) -> FileContent:
     if mode.target_versions:
         versions = mode.target_versions
     else:
     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)
+
     normalize_fmt_off(src_node)
     normalize_fmt_off(src_node)
-    lines = LineGenerator(
-        mode=mode,
-        remove_u_prefix="unicode_literals" in future_imports
-        or supports_feature(versions, Feature.UNICODE_LITERALS),
-    )
+    lines = LineGenerator(mode=mode)
     elt = EmptyLineTracker(is_pyi=mode.is_pyi)
     empty_line = Line(mode=mode)
     after = 0
     elt = EmptyLineTracker(is_pyi=mode.is_pyi)
     empty_line = Line(mode=mode)
     after = 0
@@ -1081,7 +1151,9 @@ def decode_bytes(src: bytes) -> Tuple[FileContent, Encoding, NewLine]:
         return tiow.read(), encoding, newline
 
 
         return tiow.read(), encoding, newline
 
 
-def get_features_used(node: Node) -> Set[Feature]:
+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:
     """Return a set of (relatively) new Python features used in this file.
 
     Currently looking for:
@@ -1091,20 +1163,34 @@ def get_features_used(node: Node) -> Set[Feature]:
     - positional only arguments in function signatures and lambdas;
     - assignment expression;
     - relaxed decorator syntax;
     - 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()
     """
     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():
     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)
 
         elif n.type == token.NUMBER:
             if value_head in {'f"', 'F"', "f'", "F'", "rf", "fr", "RF", "FR"}:
                 features.add(Feature.F_STRINGS)
 
         elif n.type == token.NUMBER:
-            if "_" in n.value:  # type: ignore
+            assert isinstance(n, Leaf)
+            if "_" in n.value:
                 features.add(Feature.NUMERIC_UNDERSCORES)
 
         elif n.type == token.SLASH:
                 features.add(Feature.NUMERIC_UNDERSCORES)
 
         elif n.type == token.SLASH:
-            if n.parent and n.parent.type in {syms.typedargslist, syms.arglist}:
+            if n.parent and n.parent.type in {
+                syms.typedargslist,
+                syms.arglist,
+                syms.varargslist,
+            }:
                 features.add(Feature.POS_ONLY_ARGUMENTS)
 
         elif n.type == token.COLONEQUAL:
                 features.add(Feature.POS_ONLY_ARGUMENTS)
 
         elif n.type == token.COLONEQUAL:
@@ -1135,12 +1221,29 @@ def get_features_used(node: Node) -> Set[Feature]:
                         if argch.type in STARS:
                             features.add(feature)
 
                         if argch.type in STARS:
                             features.add(feature)
 
+        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
 
 
     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."""
     """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]
     }
     return {
         version for version in TargetVersion if features <= VERSION_TO_FEATURES[version]
     }
@@ -1202,9 +1305,8 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None:
         src_ast = parse_ast(src)
     except Exception as exc:
         raise AssertionError(
         src_ast = parse_ast(src)
     except Exception as exc:
         raise AssertionError(
-            "cannot use --safe with this file; failed to parse source file.  AST"
-            f" error message: {exc}"
-        )
+            f"cannot use --safe with this file; failed to parse source file: {exc}"
+        ) from exc
 
     try:
         dst_ast = parse_ast(dst)
 
     try:
         dst_ast = parse_ast(dst)
@@ -1265,7 +1367,7 @@ def patch_click() -> None:
     """
     try:
         from click import core
     """
     try:
         from click import core
-        from click import _unicodefun  # type: ignore
+        from click import _unicodefun
     except ModuleNotFoundError:
         return
 
     except ModuleNotFoundError:
         return