]> 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:

Fix cache versioning when BLACK_CACHE_DIR is set (#3937)
[etc/vim.git] / src / black / __init__.py
index 7fed1355f47752ccd6b22141f0827f568cae43f2..188a4f79f0e51f027f445525bda3b2e68fd93081 100644 (file)
@@ -1,20 +1,16 @@
-import asyncio
-from json.decoder import JSONDecodeError
-import json
-from concurrent.futures import Executor, ThreadPoolExecutor, ProcessPoolExecutor
-from contextlib import contextmanager
-from datetime import datetime
-from enum import Enum
 import io
-from multiprocessing import Manager, freeze_support
-import os
-from pathlib import Path
-from pathspec.patterns.gitwildmatch import GitWildMatchPatternError
-import regex as re
-import signal
+import json
+import platform
+import re
 import sys
 import tokenize
 import traceback
+from contextlib import contextmanager
+from dataclasses import replace
+from datetime import datetime, timezone
+from enum import Enum
+from json.decoder import JSONDecodeError
+from pathlib import Path
 from typing import (
     Any,
     Dict,
@@ -24,47 +20,68 @@ from typing import (
     MutableMapping,
     Optional,
     Pattern,
+    Sequence,
     Set,
     Sized,
     Tuple,
     Union,
 )
 
-from dataclasses import replace
 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.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.lines import Line, EmptyLineTracker
-from black.linegen import transform_line, LineGenerator, LN
+from _black_version import version as __version__
+from black.cache import Cache
 from black.comments import normalize_fmt_off
-from black.mode import 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.output import dump_to_file, ipynb_diff, diff, color_diff, out, err
-from black.report import Report, Changed, NothingChanged
-from black.files import find_project_root, find_pyproject_toml, parse_pyproject_toml
-from black.files import gen_python_files, get_gitignore, normalize_path_maybe_ignore
-from black.files import wrap_stream_for_windows
-from black.parsing import InvalidInput  # noqa F401
-from black.parsing import lib2to3_parse, parse_ast, stringify_ast
+from black.const import (
+    DEFAULT_EXCLUDES,
+    DEFAULT_INCLUDES,
+    DEFAULT_LINE_LENGTH,
+    STDIN_PLACEHOLDER,
+)
+from black.files import (
+    find_project_root,
+    find_pyproject_toml,
+    find_user_pyproject_toml,
+    gen_python_files,
+    get_gitignore,
+    normalize_path_maybe_ignore,
+    parse_pyproject_toml,
+    wrap_stream_for_windows,
+)
 from black.handle_ipynb_magics import (
-    mask_cell,
-    unmask_cell,
-    remove_trailing_semicolon,
-    put_trailing_semicolon_back,
+    PYTHON_CELL_MAGICS,
     TRANSFORMED_MAGICS,
     jupyter_dependencies_are_installed,
+    mask_cell,
+    put_trailing_semicolon_back,
+    remove_trailing_semicolon,
+    unmask_cell,
 )
-
-
-# lib2to3 fork
-from blib2to3.pytree import Node, Leaf
+from black.linegen import LN, LineGenerator, transform_line
+from black.lines import EmptyLineTracker, LinesBlock
+from black.mode import FUTURE_FLAG_TO_FEATURE, VERSION_TO_FEATURES, Feature
+from black.mode import Mode as Mode  # re-exported
+from black.mode import TargetVersion, supports_feature
+from black.nodes import (
+    STARS,
+    is_number_token,
+    is_simple_decorator_expression,
+    is_string_token,
+    syms,
+)
+from black.output import color_diff, diff, dump_to_file, err, ipynb_diff, out
+from black.parsing import InvalidInput  # noqa F401
+from black.parsing import lib2to3_parse, parse_ast, stringify_ast
+from black.report import Changed, NothingChanged, Report
+from black.trans import iter_fexpr_spans
 from blib2to3.pgen2 import token
+from blib2to3.pytree import Leaf, Node
 
-from _black_version import version as __version__
+COMPILED = Path(__file__).suffix in (".pyd", ".so")
 
 # types
 FileContent = str
@@ -105,7 +122,9 @@ 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
 
@@ -133,6 +152,16 @@ 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)
@@ -168,14 +197,19 @@ def validate_regex(
     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
-    except re.error:
-        raise click.BadParameter("Not a valid regular expression") from None
+    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",
@@ -192,8 +226,9 @@ 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(
@@ -212,6 +247,22 @@ def validate_regex(
         "(useful when piping source on standard input)."
     ),
 )
+@click.option(
+    "--python-cell-magics",
+    multiple=True,
+    help=(
+        "When processing Jupyter Notebooks, add the given magic to the list"
+        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",
@@ -228,9 +279,14 @@ def validate_regex(
     "--experimental-string-processing",
     is_flag=True,
     hidden=True,
+    help="(DEPRECATED and now included in --preview) Normalize string literals.",
+)
+@click.option(
+    "--preview",
+    is_flag=True,
     help=(
-        "Experimental option that performs more normalization on string literals."
-        " Currently disabled because it leads to some crashes."
+        "Enable potentially disruptive style changes that may be added to Black's main"
+        " functionality in the next major release."
     ),
 )
 @click.option(
@@ -262,7 +318,8 @@ def validate_regex(
     type=str,
     help=(
         "Require a specific version of Black to be running (useful for unifying results"
-        " across many environments e.g. with a pyproject.toml file)."
+        " across many environments e.g. with a pyproject.toml file). It can be"
+        " either a major version number or an exact version."
     ),
 )
 @click.option(
@@ -312,12 +369,23 @@ 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 "
         "editors that rely on using stdin."
     ),
 )
+@click.option(
+    "-W",
+    "--workers",
+    type=click.IntRange(min=1),
+    default=None,
+    help=(
+        "Number of parallel workers [default: BLACK_NUM_WORKERS environment variable "
+        "or number of CPUs in the system]"
+    ),
+)
 @click.option(
     "-q",
     "--quiet",
@@ -336,7 +404,13 @@ def validate_regex(
         " 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'})\n"
+        f"Python ({platform.python_implementation()}) {platform.python_version()}"
+    ),
+)
 @click.argument(
     "src",
     nargs=-1,
@@ -361,7 +435,7 @@ def validate_regex(
     help="Read configuration from FILE path.",
 )
 @click.pass_context
-def main(
+def main(  # noqa: C901
     ctx: click.Context,
     code: Optional[str],
     line_length: int,
@@ -372,26 +446,75 @@ def main(
     fast: bool,
     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,
+    preview: 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],
+    workers: Optional[int],
     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)
+
+    if src and code is not None:
+        out(
+            main.get_usage(ctx)
+            + "\n\n'SRC' and 'code' cannot be passed simultaneously."
+        )
+        ctx.exit(1)
+    if not src and code is None:
+        out(main.get_usage(ctx) + "\n\nOne of 'SRC' or 'code' is required.")
+        ctx.exit(1)
+
+    root, method = (
+        find_project_root(src, stdin_filename) 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",
+            )
+
+        if config:
+            config_source = ctx.get_parameter_source("config")
+            user_level_config = str(find_user_pyproject_toml())
+            if config == user_level_config:
+                out(
+                    "Using configuration from user-level config at "
+                    f"'{user_level_config}'.",
+                    fg="blue",
+                )
+            elif 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")
+            if ctx.default_map:
+                for param, value in ctx.default_map.items():
+                    out(f"{param}: {value}")
 
     error_msg = "Oh no! 💥 💔 💥"
-    if required_version and required_version != __version__:
+    if (
+        required_version
+        and required_version != __version__
+        and required_version != __version__.split(".")[0]
+    ):
         err(
             f"{error_msg} The required version `{required_version}` does not match"
             f" the running version `{__version__}`!"
@@ -412,9 +535,12 @@ def main(
         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,
+        preview=preview,
+        python_cell_magics=set(python_cell_magics),
     )
 
     if code is not None:
@@ -429,9 +555,10 @@ def main(
             content=code, fast=fast, write_back=write_back, mode=mode, report=report
         )
     else:
+        assert root is not None  # root is only None if code is not None
         try:
             sources = get_sources(
-                ctx=ctx,
+                root=root,
                 src=src,
                 quiet=quiet,
                 verbose=verbose,
@@ -462,15 +589,20 @@ def main(
                 report=report,
             )
         else:
+            from black.concurrency import reformat_many
+
             reformat_many(
                 sources=sources,
                 fast=fast,
                 write_back=write_back,
                 mode=mode,
                 report=report,
+                workers=workers,
             )
 
     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)
@@ -479,7 +611,7 @@ def main(
 
 def get_sources(
     *,
-    ctx: click.Context,
+    root: Path,
     src: Tuple[str, ...],
     quiet: bool,
     verbose: bool,
@@ -491,16 +623,12 @@ def get_sources(
     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)
-        gitignore = get_gitignore(root)
-    else:
-        gitignore = None
+    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:
@@ -511,9 +639,15 @@ def get_sources(
             is_stdin = False
 
         if is_stdin or p.is_file():
-            normalized_path = normalize_path_maybe_ignore(p, root, report)
+            normalized_path: Optional[str] = normalize_path_maybe_ignore(
+                p, 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.
@@ -529,12 +663,23 @@ 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():
+            p_relative = normalize_path_maybe_ignore(p, root, report)
+            assert p_relative is not None
+            p = root / p_relative
+            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(),
@@ -550,9 +695,12 @@ 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
 
 
@@ -592,6 +740,9 @@ def reformat_code(
         report.failed(path, 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_one(
     src: Path, fast: bool, write_back: WriteBack, mode: Mode, report: "Report"
 ) -> None:
@@ -621,12 +772,9 @@ def reformat_one(
             if format_stdin_to_stdout(fast=fast, write_back=write_back, mode=mode):
                 changed = Changed.YES
         else:
-            cache: Cache = {}
+            cache = Cache.read(mode)
             if write_back not in (WriteBack.DIFF, WriteBack.COLOR_DIFF):
-                cache = read_cache(mode)
-                res_src = src.resolve()
-                res_src_s = str(res_src)
-                if res_src_s in cache and cache[res_src_s] == get_cache_info(res_src):
+                if not cache.is_changed(src):
                     changed = Changed.CACHED
             if changed is not Changed.CACHED and format_file_in_place(
                 src, fast=fast, write_back=write_back, mode=mode
@@ -635,7 +783,7 @@ def reformat_one(
             if (write_back is WriteBack.YES and changed is not Changed.CACHED) or (
                 write_back is WriteBack.CHECK and changed is Changed.NO
             ):
-                write_cache(cache, [src], mode)
+                cache.write([src])
         report.done(src, changed)
     except Exception as exc:
         if report.verbose:
@@ -643,114 +791,6 @@ def reformat_one(
         report.failed(src, str(exc))
 
 
-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, 60)
-    try:
-        executor = ProcessPoolExecutor(max_workers=worker_count)
-    except (ImportError, 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
-        # any good due to the Global Interpreter Lock)
-        executor = ThreadPoolExecutor(max_workers=1)
-
-    try:
-        loop.run_until_complete(
-            schedule_formatting(
-                sources=sources,
-                fast=fast,
-                write_back=write_back,
-                mode=mode,
-                report=report,
-                loop=loop,
-                executor=executor,
-            )
-        )
-    finally:
-        shutdown(loop)
-        if executor is not None:
-            executor.shutdown()
-
-
-async def schedule_formatting(
-    sources: Set[Path],
-    fast: bool,
-    write_back: WriteBack,
-    mode: Mode,
-    report: "Report",
-    loop: asyncio.AbstractEventLoop,
-    executor: Executor,
-) -> None:
-    """Run formatting of `sources` in parallel using the provided `executor`.
-
-    (Use ProcessPoolExecutors for actual parallelism.)
-
-    `write_back`, `fast`, and `mode` options are passed to
-    :func:`format_file_in_place`.
-    """
-    cache: Cache = {}
-    if write_back not in (WriteBack.DIFF, WriteBack.COLOR_DIFF):
-        cache = read_cache(mode)
-        sources, cached = filter_cached(cache, sources)
-        for src in sorted(cached):
-            report.done(src, Changed.CACHED)
-    if not sources:
-        return
-
-    cancelled = []
-    sources_to_cache = []
-    lock = None
-    if write_back in (WriteBack.DIFF, WriteBack.COLOR_DIFF):
-        # For diff output, we need locks to ensure we don't interleave output
-        # from different processes.
-        manager = Manager()
-        lock = manager.Lock()
-    tasks = {
-        asyncio.ensure_future(
-            loop.run_in_executor(
-                executor, format_file_in_place, src, fast, mode, write_back, lock
-            )
-        ): src
-        for src in sorted(sources)
-    }
-    pending = tasks.keys()
-    try:
-        loop.add_signal_handler(signal.SIGINT, cancel, pending)
-        loop.add_signal_handler(signal.SIGTERM, cancel, pending)
-    except NotImplementedError:
-        # There are no good alternatives for these on Windows.
-        pass
-    while pending:
-        done, _ = await asyncio.wait(pending, return_when=asyncio.FIRST_COMPLETED)
-        for task in done:
-            src = tasks.pop(task)
-            if task.cancelled():
-                cancelled.append(task)
-            elif task.exception():
-                report.failed(src, str(task.exception()))
-            else:
-                changed = Changed.YES if task.result() else Changed.NO
-                # If the file was written back or was successfully checked as
-                # well-formatted, store this information in the cache.
-                if write_back is WriteBack.YES or (
-                    write_back is WriteBack.CHECK and changed is Changed.NO
-                ):
-                    sources_to_cache.append(src)
-                report.done(src, changed)
-    if cancelled:
-        await asyncio.gather(*cancelled, loop=loop, return_exceptions=True)
-    if sources_to_cache:
-        write_cache(cache, sources_to_cache, mode)
-
-
 def format_file_in_place(
     src: Path,
     fast: bool,
@@ -769,8 +809,11 @@ 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)
@@ -780,14 +823,16 @@ 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:
@@ -825,7 +870,7 @@ 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())
@@ -850,9 +895,9 @@ 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)
@@ -871,17 +916,7 @@ def check_stability_and_equivalence(
     content differently.
     """
     assert_equivalent(src_contents, dst_contents)
-
-    # Forced second pass to work around optional trailing commas (becoming
-    # forced trailing commas on pass 2) interacting differently with optional
-    # parentheses.  Admittedly ugly.
-    dst_contents_pass2 = format_str(dst_contents, mode=mode)
-    if dst_contents != dst_contents_pass2:
-        dst_contents = dst_contents_pass2
-        assert_equivalent(src_contents, dst_contents, pass_num=2)
-        assert_stable(src_contents, dst_contents, mode=mode)
-    # Note: no need to explicitly call `assert_stable` if `dst_contents` was
-    # the same as `dst_contents_pass2`.
+    assert_stable(src_contents, dst_contents, mode=mode)
 
 
 def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileContent:
@@ -891,9 +926,6 @@ 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:
@@ -907,8 +939,10 @@ def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileCo
     return dst_contents
 
 
-def validate_cell(src: str) -> None:
-    """Check that cell does not already contain TransformerManager transformations.
+def validate_cell(src: str, mode: Mode) -> None:
+    """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
@@ -924,6 +958,11 @@ def validate_cell(src: str) -> None:
     """
     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 | mode.python_cell_magics
+    ):
+        raise NothingChanged
 
 
 def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
@@ -942,7 +981,7 @@ def format_cell(src: str, *, fast: bool, mode: Mode) -> str:
     could potentially be automagics or multi-line magics, which
     are currently not supported.
     """
-    validate_cell(src)
+    validate_cell(src, mode)
     src_without_trailing_semicolon, has_trailing_semicolon = remove_trailing_semicolon(
         src
     )
@@ -981,6 +1020,9 @@ 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)
@@ -1004,7 +1046,7 @@ def format_ipynb_string(src_contents: str, *, fast: bool, mode: Mode) -> FileCon
         raise NothingChanged
 
 
-def format_str(src_contents: str, *, mode: Mode) -> FileContent:
+def format_str(src_contents: str, *, mode: Mode) -> str:
     """Reformat a string and return new contents.
 
     `mode` determines formatting options, such as how many characters per line are
@@ -1034,35 +1076,57 @@ def format_str(src_contents: str, *, mode: Mode) -> FileContent:
         hey
 
     """
+    dst_contents = _format_str_once(src_contents, mode=mode)
+    # Forced second pass to work around optional trailing commas (becoming
+    # forced trailing commas on pass 2) interacting differently with optional
+    # parentheses.  Admittedly ugly.
+    if src_contents != dst_contents:
+        return _format_str_once(dst_contents, mode=mode)
+    return dst_contents
+
+
+def _format_str_once(src_contents: str, *, mode: Mode) -> str:
     src_node = lib2to3_parse(src_contents.lstrip(), mode.target_versions)
-    dst_contents = []
-    future_imports = get_future_imports(src_node)
+    dst_blocks: List[LinesBlock] = []
     if mode.target_versions:
         versions = mode.target_versions
     else:
-        versions = detect_target_versions(src_node)
+        future_imports = get_future_imports(src_node)
+        versions = detect_target_versions(src_node, future_imports=future_imports)
+
+    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,
-        remove_u_prefix="unicode_literals" in future_imports
-        or supports_feature(versions, Feature.UNICODE_LITERALS),
-    )
-    elt = EmptyLineTracker(is_pyi=mode.is_pyi)
-    empty_line = Line(mode=mode)
-    after = 0
+    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)
 
 
@@ -1083,30 +1147,55 @@ def decode_bytes(src: bytes) -> Tuple[FileContent, 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:
     - f-strings;
+    - self-documenting expressions in f-strings (f"{x=}");
     - underscores in numeric literals;
     - trailing commas after * or ** in function signatures and calls;
     - positional only arguments in function signatures and lambdas;
     - assignment expression;
     - 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:
+        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():
-        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 "_" in n.value:  # type: ignore
+                if Feature.DEBUG_F_STRINGS not in features:
+                    for span_beg, span_end in iter_fexpr_spans(n.value):
+                        if n.value[span_beg : span_end - 1].rstrip().endswith("="):
+                            features.add(Feature.DEBUG_F_STRINGS)
+                            break
+
+        elif is_number_token(n):
+            if "_" in n.value:
                 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:
@@ -1137,12 +1226,68 @@ def get_features_used(node: Node) -> Set[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)
+
+        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
+            and n.children[1].type == token.STAR
+        ):
+            features.add(Feature.EXCEPT_STAR)
+
+        elif n.type in {syms.subscriptlist, syms.trailer} and any(
+            child.type == syms.star_expr for child in n.children
+        ):
+            features.add(Feature.VARIADIC_GENERICS)
+
+        elif (
+            n.type == syms.tname_star
+            and len(n.children) == 3
+            and n.children[2].type == syms.star_expr
+        ):
+            features.add(Feature.VARIADIC_GENERICS)
+
+        elif n.type in (syms.type_stmt, syms.typeparams):
+            features.add(Feature.TYPE_PARAMS)
+
     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."""
-    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]
     }
@@ -1198,13 +1343,16 @@ def get_future_imports(node: Node) -> Set[str]:
     return imports
 
 
-def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None:
+def assert_equivalent(src: str, dst: str) -> None:
     """Raise AssertionError if `src` and `dst` aren't equivalent."""
     try:
         src_ast = parse_ast(src)
     except Exception as exc:
         raise AssertionError(
-            "cannot use --safe with this file; failed to parse source file."
+            "cannot use --safe with this file; failed to parse source file AST: "
+            f"{exc}\n"
+            "This could be caused by running Black with an older Python version "
+            "that does not support new syntax used in your source file."
         ) from exc
 
     try:
@@ -1212,7 +1360,7 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None:
     except Exception as exc:
         log = dump_to_file("".join(traceback.format_tb(exc.__traceback__)), dst)
         raise AssertionError(
-            f"INTERNAL ERROR: Black produced invalid code on pass {pass_num}: {exc}. "
+            f"INTERNAL ERROR: Black produced invalid code: {exc}. "
             "Please report a bug on https://github.com/psf/black/issues.  "
             f"This invalid output might be helpful: {log}"
         ) from None
@@ -1223,14 +1371,17 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None:
         log = dump_to_file(diff(src_ast_str, dst_ast_str, "src", "dst"))
         raise AssertionError(
             "INTERNAL ERROR: Black produced code that is not equivalent to the"
-            f" source on pass {pass_num}.  Please report a bug on "
+            " source.  Please report a bug on "
             f"https://github.com/psf/black/issues.  This diff might be helpful: {log}"
         ) from None
 
 
 def assert_stable(src: str, dst: str, mode: Mode) -> None:
     """Raise AssertionError if `dst` reformats differently the second time."""
-    newdst = format_str(dst, mode=mode)
+    # We shouldn't call format_str() here, because that formats the string
+    # twice and may hide a bug where we bounce back and forth between two
+    # versions.
+    newdst = _format_str_once(dst, mode=mode)
     if dst != newdst:
         log = dump_to_file(
             str(mode),
@@ -1253,34 +1404,14 @@ 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.
-    """
-    try:
-        from click import core
-        from click import _unicodefun  # type: ignore
-    except ModuleNotFoundError:
-        return
-
-    for module in (core, _unicodefun):
-        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:
+    # 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()
 
-def patched_main() -> None:
-    maybe_install_uvloop()
-    freeze_support()
-    patch_click()
     main()