]> 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:

Simplify empty line tracker (#3797)
[etc/vim.git] / src / black / __init__.py
index 86a0b63744264c85dc23f8702768d333c047f221..923a51867b5474d5a638b356fe10c5e586cc9c90 100644 (file)
@@ -7,7 +7,7 @@ import tokenize
 import traceback
 from contextlib import contextmanager
 from dataclasses import replace
-from datetime import datetime
+from datetime import datetime, timezone
 from enum import Enum
 from json.decoder import JSONDecodeError
 from pathlib import Path
@@ -30,6 +30,7 @@ from typing import (
 import click
 from click.core import ParameterSource
 from mypy_extensions import mypyc_attr
+from pathspec import PathSpec
 from pathspec.patterns.gitwildmatch import GitWildMatchPatternError
 
 from _black_version import version as __version__
@@ -61,7 +62,7 @@ from black.handle_ipynb_magics import (
     unmask_cell,
 )
 from black.linegen import LN, LineGenerator, transform_line
-from black.lines import EmptyLineTracker, Line
+from black.lines import EmptyLineTracker, LinesBlock
 from black.mode import (
     FUTURE_FLAG_TO_FEATURE,
     VERSION_TO_FEATURES,
@@ -126,7 +127,9 @@ def read_pyproject_toml(
     otherwise.
     """
     if not value:
-        value = find_pyproject_toml(ctx.params.get("src", ()))
+        value = find_pyproject_toml(
+            ctx.params.get("src", ()), ctx.params.get("stdin_filename", None)
+        )
         if value is None:
             return None
 
@@ -154,6 +157,16 @@ def read_pyproject_toml(
             "target-version", "Config key target-version must be a list"
         )
 
+    exclude = config.get("exclude")
+    if exclude is not None and not isinstance(exclude, str):
+        raise click.BadOptionUsage("exclude", "Config key exclude must be a string")
+
+    extend_exclude = config.get("extend_exclude")
+    if extend_exclude is not None and not isinstance(extend_exclude, str):
+        raise click.BadOptionUsage(
+            "extend-exclude", "Config key extend-exclude must be a string"
+        )
+
     default_map: Dict[str, Any] = {}
     if ctx.default_map:
         default_map.update(ctx.default_map)
@@ -218,8 +231,9 @@ def validate_regex(
     callback=target_version_option_callback,
     multiple=True,
     help=(
-        "Python versions that should be supported by Black's output. [default: per-file"
-        " auto-detection]"
+        "Python versions that should be supported by Black's output. By default, Black"
+        " will try to infer this from the project metadata in pyproject.toml. If this"
+        " does not yield conclusive results, Black will use per-file auto-detection."
     ),
 )
 @click.option(
@@ -243,11 +257,17 @@ def validate_regex(
     multiple=True,
     help=(
         "When processing Jupyter Notebooks, add the given magic to the list"
-        f" of known python-magics ({', '.join(PYTHON_CELL_MAGICS)})."
+        f" of known python-magics ({', '.join(sorted(PYTHON_CELL_MAGICS))})."
         " Useful for formatting cells with custom python magics."
     ),
     default=[],
 )
+@click.option(
+    "-x",
+    "--skip-source-first-line",
+    is_flag=True,
+    help="Skip the first line of the source code.",
+)
 @click.option(
     "-S",
     "--skip-string-normalization",
@@ -354,6 +374,7 @@ def validate_regex(
 @click.option(
     "--stdin-filename",
     type=str,
+    is_eager=True,
     help=(
         "The name of the file when passing it through stdin. Useful to make "
         "sure Black will respect --force-exclude option on some "
@@ -365,7 +386,10 @@ def validate_regex(
     "--workers",
     type=click.IntRange(min=1),
     default=None,
-    help="Number of parallel workers [default: number of CPUs in the system]",
+    help=(
+        "Number of parallel workers [default: BLACK_NUM_WORKERS environment variable "
+        "or number of CPUs in the system]"
+    ),
 )
 @click.option(
     "-q",
@@ -428,6 +452,7 @@ def main(  # noqa: C901
     pyi: bool,
     ipynb: bool,
     python_cell_magics: Sequence[str],
+    skip_source_first_line: bool,
     skip_string_normalization: bool,
     skip_magic_trailing_comma: bool,
     experimental_string_processing: bool,
@@ -469,22 +494,6 @@ def main(  # noqa: C901
                 fg="blue",
             )
 
-            normalized = [
-                (source, source)
-                if source == "-"
-                else (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())
@@ -501,6 +510,9 @@ def main(  # noqa: C901
                 out("Using configuration from project root.", fg="blue")
             else:
                 out(f"Using configuration in '{config}'.", fg="blue")
+            if ctx.default_map:
+                for param, value in ctx.default_map.items():
+                    out(f"{param}: {value}")
 
     error_msg = "Oh no! 💥 💔 💥"
     if (
@@ -528,6 +540,7 @@ def main(  # noqa: C901
         line_length=line_length,
         is_pyi=pyi,
         is_ipynb=ipynb,
+        skip_source_first_line=skip_source_first_line,
         string_normalization=not skip_string_normalization,
         magic_trailing_comma=not skip_magic_trailing_comma,
         experimental_string_processing=experimental_string_processing,
@@ -617,6 +630,11 @@ def get_sources(
     sources: Set[Path] = set()
     root = ctx.obj["root"]
 
+    using_default_exclude = exclude is None
+    exclude = re_compile_maybe_verbose(DEFAULT_EXCLUDES) if exclude is None else exclude
+    gitignore: Optional[Dict[Path, PathSpec]] = None
+    root_gitignore = get_gitignore(root)
+
     for s in src:
         if s == "-" and stdin_filename:
             p = Path(stdin_filename)
@@ -626,9 +644,15 @@ def get_sources(
             is_stdin = False
 
         if is_stdin or p.is_file():
-            normalized_path = normalize_path_maybe_ignore(p, ctx.obj["root"], report)
+            normalized_path: Optional[str] = normalize_path_maybe_ignore(
+                p, ctx.obj["root"], report
+            )
             if normalized_path is None:
+                if verbose:
+                    out(f'Skipping invalid source: "{normalized_path}"', fg="red")
                 continue
+            if verbose:
+                out(f'Found input source: "{normalized_path}"', fg="blue")
 
             normalized_path = "/" + normalized_path
             # Hard-exclude any files that matches the `--force-exclude` regex.
@@ -644,17 +668,21 @@ def get_sources(
                 p = Path(f"{STDIN_PLACEHOLDER}{str(p)}")
 
             if p.suffix == ".ipynb" and not jupyter_dependencies_are_installed(
-                verbose=verbose, quiet=quiet
+                warn=verbose or not quiet
             ):
                 continue
 
             sources.add(p)
         elif p.is_dir():
-            if exclude is None:
-                exclude = re_compile_maybe_verbose(DEFAULT_EXCLUDES)
-                gitignore = get_gitignore(root)
-            else:
-                gitignore = None
+            p = root / normalize_path_maybe_ignore(p, ctx.obj["root"], report)
+            if verbose:
+                out(f'Found input source directory: "{p}"', fg="blue")
+
+            if using_default_exclude:
+                gitignore = {
+                    root: root_gitignore,
+                    p: get_gitignore(p),
+                }
             sources.update(
                 gen_python_files(
                     p.iterdir(),
@@ -670,9 +698,12 @@ def get_sources(
                 )
             )
         elif s == "-":
+            if verbose:
+                out("Found input source stdin", fg="blue")
             sources.add(p)
         else:
             err(f"invalid path: {s}")
+
     return sources
 
 
@@ -784,8 +815,11 @@ def format_file_in_place(
     elif src.suffix == ".ipynb":
         mode = replace(mode, is_ipynb=True)
 
-    then = datetime.utcfromtimestamp(src.stat().st_mtime)
+    then = datetime.fromtimestamp(src.stat().st_mtime, timezone.utc)
+    header = b""
     with open(src, "rb") as buf:
+        if mode.skip_source_first_line:
+            header = buf.readline()
         src_contents, encoding, newline = decode_bytes(buf.read())
     try:
         dst_contents = format_file_contents(src_contents, fast=fast, mode=mode)
@@ -795,14 +829,16 @@ def format_file_in_place(
         raise ValueError(
             f"File '{src}' cannot be parsed as valid Jupyter notebook."
         ) from None
+    src_contents = header.decode(encoding) + src_contents
+    dst_contents = header.decode(encoding) + dst_contents
 
     if write_back == WriteBack.YES:
         with open(src, "w", encoding=encoding, newline=newline) as f:
             f.write(dst_contents)
     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"
+        now = datetime.now(timezone.utc)
+        src_name = f"{src}\t{then}"
+        dst_name = f"{src}\t{now}"
         if mode.is_ipynb:
             diff_contents = ipynb_diff(src_contents, dst_contents, src_name, dst_name)
         else:
@@ -840,7 +876,7 @@ def format_stdin_to_stdout(
     write a diff to stdout. The `mode` argument is passed to
     :func:`format_file_contents`.
     """
-    then = datetime.utcnow()
+    then = datetime.now(timezone.utc)
 
     if content is None:
         src, encoding, newline = decode_bytes(sys.stdin.buffer.read())
@@ -865,9 +901,9 @@ def format_stdin_to_stdout(
                 dst += "\n"
             f.write(dst)
         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"
+            now = datetime.now(timezone.utc)
+            src_name = f"STDIN\t{then}"
+            dst_name = f"STDOUT\t{now}"
             d = diff(src, dst, src_name, dst_name)
             if write_back == WriteBack.COLOR_DIFF:
                 d = color_diff(d)
@@ -896,9 +932,6 @@ def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileCo
     valid by calling :func:`assert_equivalent` and :func:`assert_stable` on it.
     `mode` is passed to :func:`format_str`.
     """
-    if not src_contents.strip():
-        raise NothingChanged
-
     if mode.is_ipynb:
         dst_contents = format_ipynb_string(src_contents, fast=fast, mode=mode)
     else:
@@ -993,6 +1026,9 @@ def format_ipynb_string(src_contents: str, *, fast: bool, mode: Mode) -> FileCon
     Operate cell-by-cell, only on code cells, only for Python notebooks.
     If the ``.ipynb`` originally had a trailing newline, it'll be preserved.
     """
+    if not src_contents:
+        raise NothingChanged
+
     trailing_newline = src_contents[-1] == "\n"
     modified = False
     nb = json.loads(src_contents)
@@ -1057,31 +1093,46 @@ def format_str(src_contents: str, *, mode: Mode) -> str:
 
 def _format_str_once(src_contents: str, *, mode: Mode) -> str:
     src_node = lib2to3_parse(src_contents.lstrip(), mode.target_versions)
-    dst_contents = []
+    dst_blocks: List[LinesBlock] = []
     if mode.target_versions:
         versions = mode.target_versions
     else:
         future_imports = get_future_imports(src_node)
         versions = detect_target_versions(src_node, future_imports=future_imports)
 
-    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
+    context_manager_features = {
+        feature
+        for feature in {Feature.PARENTHESIZED_CONTEXT_MANAGERS}
+        if supports_feature(versions, feature)
+    }
+    normalize_fmt_off(src_node)
+    lines = LineGenerator(mode=mode, features=context_manager_features)
+    elt = EmptyLineTracker(mode=mode)
     split_line_features = {
         feature
         for feature in {Feature.TRAILING_COMMA_IN_CALL, Feature.TRAILING_COMMA_IN_DEF}
         if supports_feature(versions, feature)
     }
+    block: Optional[LinesBlock] = None
     for current_line in lines.visit(src_node):
-        dst_contents.append(str(empty_line) * after)
-        before, after = elt.maybe_empty_lines(current_line)
-        dst_contents.append(str(empty_line) * before)
+        block = elt.maybe_empty_lines(current_line)
+        dst_blocks.append(block)
         for line in transform_line(
             current_line, mode=mode, features=split_line_features
         ):
-            dst_contents.append(str(line))
+            block.content_lines.append(str(line))
+    if dst_blocks:
+        dst_blocks[-1].after = 0
+    dst_contents = []
+    for block in dst_blocks:
+        dst_contents.extend(block.all_lines())
+    if not dst_contents:
+        # Use decode_bytes to retrieve the correct source newline (CRLF or LF),
+        # and check if normalized_content has more than one line
+        normalized_content, _, newline = decode_bytes(src_contents.encode("utf-8"))
+        if "\n" in normalized_content:
+            return newline
+        return ""
     return "".join(dst_contents)
 
 
@@ -1117,6 +1168,10 @@ def get_features_used(  # noqa: C901
     - relaxed decorator syntax;
     - usage of __future__ flags (annotations);
     - print / exec statements;
+    - parenthesized context managers;
+    - match statements;
+    - except* clause;
+    - variadic generics;
     """
     features: Set[Feature] = set()
     if future_imports:
@@ -1192,6 +1247,23 @@ def get_features_used(  # noqa: C901
         ):
             features.add(Feature.ANN_ASSIGN_EXTENDED_RHS)
 
+        elif (
+            n.type == syms.with_stmt
+            and len(n.children) > 2
+            and n.children[1].type == syms.atom
+        ):
+            atom_children = n.children[1].children
+            if (
+                len(atom_children) == 3
+                and atom_children[0].type == token.LPAR
+                and atom_children[1].type == syms.testlist_gexp
+                and atom_children[2].type == token.RPAR
+            ):
+                features.add(Feature.PARENTHESIZED_CONTEXT_MANAGERS)
+
+        elif n.type == syms.match_stmt:
+            features.add(Feature.PATTERN_MATCHING)
+
         elif (
             n.type == syms.except_clause
             and len(n.children) >= 2
@@ -1211,6 +1283,9 @@ def get_features_used(  # noqa: C901
         ):
             features.add(Feature.VARIADIC_GENERICS)
 
+        elif n.type in (syms.type_stmt, syms.typeparams):
+            features.add(Feature.TYPE_PARAMS)
+
     return features
 
 
@@ -1335,47 +1410,14 @@ def nullcontext() -> Iterator[None]:
     yield
 
 
-def patch_click() -> None:
-    """Make Click not crash on Python 3.6 with LANG=C.
-
-    On certain misconfigured environments, Python 3 selects the ASCII encoding as the
-    default which restricts paths that it can access during the lifetime of the
-    application.  Click refuses to work in this scenario by raising a RuntimeError.
-
-    In case of Black the likelihood that non-ASCII characters are going to be used in
-    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:
-        # Removed in Click 8.1.0 and newer; we keep this around for users who have
-        # older versions installed.
-        from click import _unicodefun  # type: ignore
-    except ImportError:
-        pass
-    else:
-        modules.append(_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"):
-            module._verify_python_env = lambda: None  # type: ignore
-
-
 def patched_main() -> None:
-    if sys.platform == "win32" and getattr(sys, "frozen", False):
+    # PyInstaller patches multiprocessing to need freeze_support() even in non-Windows
+    # environments so just assume we always need to call it if frozen.
+    if getattr(sys, "frozen", False):
         from multiprocessing import freeze_support
 
         freeze_support()
 
-    patch_click()
     main()