]> 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:

Allow setting custom cache directory on all platforms (#2739)
[etc/vim.git] / src / black / __init__.py
index f2efdec83b22a53d1ad00f6d1b1057dcae1f0c0f..eaf72f9c2b3f4d12ae275d01472112931af331e9 100644 (file)
@@ -24,6 +24,7 @@ from typing import (
     MutableMapping,
     Optional,
     Pattern,
+    Sequence,
     Set,
     Sized,
     Tuple,
@@ -31,12 +32,14 @@ from typing import (
 )
 
 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.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
@@ -223,6 +226,16 @@ 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(PYTHON_CELL_MAGICS)})."
+        " Useful for formatting cells with custom python magics."
+    ),
+    default=[],
+)
 @click.option(
     "-S",
     "--skip-string-normalization",
@@ -239,9 +252,14 @@ 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 will be added to Black's main"
+        " functionality in the next major release."
     ),
 )
 @click.option(
@@ -394,9 +412,11 @@ def main(
     fast: bool,
     pyi: bool,
     ipynb: bool,
+    python_cell_magics: Sequence[str],
     skip_string_normalization: bool,
     skip_magic_trailing_comma: bool,
     experimental_string_processing: bool,
+    preview: bool,
     quiet: bool,
     verbose: bool,
     required_version: Optional[str],
@@ -410,8 +430,37 @@ def main(
     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__:
@@ -438,6 +487,8 @@ def main(
         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:
@@ -495,6 +546,8 @@ def main(
             )
 
     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)
@@ -515,14 +568,12 @@ 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)
+        gitignore = get_gitignore(ctx.obj["root"])
     else:
         gitignore = None
 
@@ -535,7 +586,7 @@ def get_sources(
             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
 
@@ -562,7 +613,7 @@ def get_sources(
             sources.update(
                 gen_python_files(
                     p.iterdir(),
-                    root,
+                    ctx.obj["root"],
                     include,
                     exclude,
                     extend_exclude,
@@ -943,7 +994,7 @@ def format_file_contents(src_contents: str, *, fast: bool, mode: Mode) -> FileCo
     return dst_contents
 
 
-def validate_cell(src: str) -> None:
+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.
@@ -962,7 +1013,10 @@ 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:
+    if (
+        src[:2] == "%%"
+        and src.split()[0][2:] not in PYTHON_CELL_MAGICS | mode.python_cell_magics
+    ):
         raise NothingChanged
 
 
@@ -982,7 +1036,7 @@ 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
     )
@@ -1082,20 +1136,8 @@ def format_str(src_contents: str, *, mode: Mode) -> FileContent:
     else:
         versions = detect_target_versions(src_node, future_imports=future_imports)
 
-    # TODO: fully drop support and this code hopefully in January 2022 :D
-    if TargetVersion.PY27 in mode.target_versions or versions == {TargetVersion.PY27}:
-        msg = (
-            "DEPRECATION: Python 2 support will be removed in the first stable release "
-            "expected in January 2022."
-        )
-        err(msg, fg="yellow", bold=True)
-
     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
@@ -1156,8 +1198,8 @@ def get_features_used(  # noqa: C901
         }
 
     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)
 
@@ -1165,14 +1207,6 @@ def get_features_used(  # noqa: C901
             assert isinstance(n, Leaf)
             if "_" in n.value:
                 features.add(Feature.NUMERIC_UNDERSCORES)
-            elif n.value.endswith(("L", "l")):
-                # Python 2: 10L
-                features.add(Feature.LONG_INT_LITERAL)
-            elif len(n.value) >= 2 and n.value[0] == "0" and n.value[1].isdigit():
-                # Python 2: 0123; 00123; ...
-                if not all(char == "0" for char in n.value):
-                    # although we don't want to match 0000 or similar
-                    features.add(Feature.OCTAL_INT_LITERAL)
 
         elif n.type == token.SLASH:
             if n.parent and n.parent.type in {
@@ -1210,31 +1244,20 @@ def get_features_used(  # noqa: C901
                         if argch.type in STARS:
                             features.add(feature)
 
-        # Python 2 only features (for its deprecation) except for integers, see above
-        elif n.type == syms.print_stmt:
-            features.add(Feature.PRINT_STMT)
-        elif n.type == syms.exec_stmt:
-            features.add(Feature.EXEC_STMT)
-        elif n.type == syms.tfpdef:
-            # def set_position((x, y), value):
-            #     ...
-            features.add(Feature.AUTOMATIC_PARAMETER_UNPACKING)
-        elif n.type == syms.except_clause:
-            # try:
-            #     ...
-            # except Exception, err:
-            #     ...
-            if len(n.children) >= 4:
-                if n.children[-2].type == token.COMMA:
-                    features.add(Feature.COMMA_STYLE_EXCEPT)
-        elif n.type == syms.raise_stmt:
-            # raise Exception, "msg"
-            if len(n.children) >= 4:
-                if n.children[-2].type == token.COMMA:
-                    features.add(Feature.COMMA_STYLE_RAISE)
-        elif n.type == token.BACKQUOTE:
-            # `i'm surprised this ever existed`
-            features.add(Feature.BACKQUOTE_REPR)
+        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
 
@@ -1305,7 +1328,10 @@ def assert_equivalent(src: str, dst: str, *, pass_num: int = 1) -> None:
         src_ast = parse_ast(src)
     except Exception as exc:
         raise AssertionError(
-            f"cannot use --safe with this file; failed to parse source file: {exc}"
+            f"cannot use --safe with this file; failed to parse source file AST: "
+            f"{exc}\n"
+            f"This could be caused by running Black with an older Python version "
+            f"that does not support new syntax used in your source file."
         ) from exc
 
     try: