Create the 2024 stable style

CHANGES.md

@@ -6,22 +6,54 @@
 
 <!-- Include any especially major or disruptive changes here -->
 
+This release introduces the new 2024 stable style (#4106), stabilizing the following
+changes:
+
+- Add parentheses around `if`-`else` expressions (#2278)
+- Dummy class and function implementations consisting only of `...` are formatted more
+  compactly (#3796)
+- If an assignment statement is too long, we now prefer splitting on the right-hand side
+  (#3368)
+- Hex codes in Unicode escape sequences are now standardized to lowercase (#2916)
+- Allow empty first lines at the beginning of most blocks (#3967, #4061)
+- Add parentheses around long type annotations (#3899)
+- Standardize on a single newline after module docstrings (#3932)
+- Fix incorrect magic trailing comma handling in return types (#3916)
+- Remove blank lines before class docstrings (#3692)
+- Wrap multiple context managers in parentheses if combined in a single `with` statement
+  (#3489)
+- Fix bug in line length calculations for power operations (#3942)
+- Add trailing commas to collection literals even if there's a comment after the last
+  entry (#3393)
+- When using `--skip-magic-trailing-comma` or `-C`, trailing commas are stripped from
+  subscript expressions with more than 1 element (#3209)
+- Add extra blank lines in stubs in a few cases (#3564, #3862)
+- Accept raw strings as docstrings (#3947)
+- Split long lines in case blocks (#4024)
+- Stop removing spaces from walrus operators within subscripts (#3823)
+- Fix incorrect formatting of certain async statements (#3609)
+- Allow combining `# fmt: skip` with other comments (#3959)
+
 ### Stable style
 
 <!-- Changes that affect Black's stable style -->
 
-### Preview style
-
-<!-- Changes that affect Black's preview style -->
+Several bug fixes were made in features that are moved to the stable style in this
+release:
 
 - Fix comment handling when parenthesising conditional expressions (#4134)
-- Format module docstrings the same as class and function docstrings (#4095)
 - Fix bug where spaces were not added around parenthesized walruses in subscripts,
   unlike other binary operators (#4109)
 - Remove empty lines before docstrings in async functions (#4132)
 - Address a missing case in the change to allow empty lines at the beginning of all
   blocks, except immediately before a docstring (#4130)
 - For stubs, fix logic to enforce empty line after nested classes with bodies (#4141)
+
+### Preview style
+
+<!-- Changes that affect Black's preview style -->
+
+- Format module docstrings the same as class and function docstrings (#4095)
 - Fix crash when using a walrus in a dictionary (#4155)
 - Fix unnecessary parentheses when wrapping long dicts (#4135)
 

src/black/comments.py

@@ -3,7 +3,7 @@
 from functools import lru_cache
 from typing import Collection, Final, Iterator, List, Optional, Tuple, Union
 
-from black.mode import Mode, Preview
+from black.mode import Mode
 from black.nodes import (
     CLOSING_BRACKETS,
     STANDALONE_COMMENT,
@@ -390,22 +390,18 @@ def _contains_fmt_skip_comment(comment_line: str, mode: Mode) -> bool:
       # noqa:XXX # fmt:skip # a nice line  <-- multiple comments (Preview)
       # pylint:XXX; fmt:skip               <-- list of comments (; separated, Preview)
     """
-    semantic_comment_blocks = (
-        [
-            comment_line,
-            *[
-                _COMMENT_PREFIX + comment.strip()
-                for comment in comment_line.split(_COMMENT_PREFIX)[1:]
-            ],
-            *[
-                _COMMENT_PREFIX + comment.strip()
-                for comment in comment_line.strip(_COMMENT_PREFIX).split(
-                    _COMMENT_LIST_SEPARATOR
-                )
-            ],
-        ]
-        if Preview.single_line_format_skip_with_multiple_comments in mode
-        else [comment_line]
-    )
+    semantic_comment_blocks = [
+        comment_line,
+        *[
+            _COMMENT_PREFIX + comment.strip()
+            for comment in comment_line.split(_COMMENT_PREFIX)[1:]
+        ],
+        *[
+            _COMMENT_PREFIX + comment.strip()
+            for comment in comment_line.strip(_COMMENT_PREFIX).split(
+                _COMMENT_LIST_SEPARATOR
+            )
+        ],
+    ]
 
     return any(comment in FMT_SKIP for comment in semantic_comment_blocks)

src/black/linegen.py

@@ -115,10 +115,8 @@ def line(self, indent: int = 0) -> Iterator[Line]:
             self.current_line.depth += indent
             return  # Line is empty, don't emit. Creating a new one unnecessary.
 
-        if (
-            Preview.improved_async_statements_handling in self.mode
-            and len(self.current_line.leaves) == 1
-            and is_async_stmt_or_funcdef(self.current_line.leaves[0])
+        if len(self.current_line.leaves) == 1 and is_async_stmt_or_funcdef(
+            self.current_line.leaves[0]
         ):
             # Special case for async def/for/with statements. `visit_async_stmt`
             # adds an `ASYNC` leaf then visits the child def/for/with statement
@@ -164,20 +162,19 @@ def visit_default(self, node: LN) -> Iterator[Line]:
     def visit_test(self, node: Node) -> Iterator[Line]:
         """Visit an `x if y else z` test"""
 
-        if Preview.parenthesize_conditional_expressions in self.mode:
-            already_parenthesized = (
-                node.prev_sibling and node.prev_sibling.type == token.LPAR
-            )
+        already_parenthesized = (
+            node.prev_sibling and node.prev_sibling.type == token.LPAR
+        )
 
-            if not already_parenthesized:
-                # Similar to logic in wrap_in_parentheses
-                lpar = Leaf(token.LPAR, "")
-                rpar = Leaf(token.RPAR, "")
-                prefix = node.prefix
-                node.prefix = ""
-                lpar.prefix = prefix
-                node.insert_child(0, lpar)
-                node.append_child(rpar)
+        if not already_parenthesized:
+            # Similar to logic in wrap_in_parentheses
+            lpar = Leaf(token.LPAR, "")
+            rpar = Leaf(token.RPAR, "")
+            prefix = node.prefix
+            node.prefix = ""
+            lpar.prefix = prefix
+            node.insert_child(0, lpar)
+            node.append_child(rpar)
 
         yield from self.visit_default(node)
 
@@ -292,9 +289,7 @@ def visit_match_case(self, node: Node) -> Iterator[Line]:
 
     def visit_suite(self, node: Node) -> Iterator[Line]:
         """Visit a suite."""
-        if (
-            self.mode.is_pyi or Preview.dummy_implementations in self.mode
-        ) and is_stub_suite(node, self.mode):
+        if is_stub_suite(node):
             yield from self.visit(node.children[2])
         else:
             yield from self.visit_default(node)
@@ -308,23 +303,15 @@ def visit_simple_stmt(self, node: Node) -> Iterator[Line]:
             prev_type = child.type
 
         if node.parent and node.parent.type in STATEMENT:
-            if Preview.dummy_implementations in self.mode:
-                condition = is_parent_function_or_class(node)
-            else:
-                condition = self.mode.is_pyi
-            if condition and is_stub_body(node):
+            if is_parent_function_or_class(node) and is_stub_body(node):
                 yield from self.visit_default(node)
             else:
                 yield from self.line(+1)
                 yield from self.visit_default(node)
                 yield from self.line(-1)
 
         else:
-            if (
-                not (self.mode.is_pyi or Preview.dummy_implementations in self.mode)
-                or not node.parent
-                or not is_stub_suite(node.parent, self.mode)
-            ):
+            if not node.parent or not is_stub_suite(node.parent):
                 yield from self.line()
             yield from self.visit_default(node)
 
@@ -342,11 +329,7 @@ def visit_async_stmt(self, node: Node) -> Iterator[Line]:
                 break
 
         internal_stmt = next(children)
-        if Preview.improved_async_statements_handling in self.mode:
-            yield from self.visit(internal_stmt)
-        else:
-            for child in internal_stmt.children:
-                yield from self.visit(child)
+        yield from self.visit(internal_stmt)
 
     def visit_decorators(self, node: Node) -> Iterator[Line]:
         """Visit decorators."""
@@ -420,10 +403,9 @@ def foo(a: int, b: float = 7): ...
 
         def foo(a: (int), b: (float) = 7): ...
         """
-        if Preview.parenthesize_long_type_hints in self.mode:
-            assert len(node.children) == 3
-            if maybe_make_parens_invisible_in_atom(node.children[2], parent=node):
-                wrap_in_parentheses(node, node.children[2], visible=False)
+        assert len(node.children) == 3
+        if maybe_make_parens_invisible_in_atom(node.children[2], parent=node):
+            wrap_in_parentheses(node, node.children[2], visible=False)
 
         yield from self.visit_default(node)
 
@@ -529,13 +511,7 @@ def __post_init__(self) -> None:
         self.visit_with_stmt = partial(v, keywords={"with"}, parens={"with"})
         self.visit_classdef = partial(v, keywords={"class"}, parens=Ø)
 
-        # When this is moved out of preview, add ":" directly to ASSIGNMENTS in nodes.py
-        if Preview.parenthesize_long_type_hints in self.mode:
-            assignments = ASSIGNMENTS | {":"}
-        else:
-            assignments = ASSIGNMENTS
-        self.visit_expr_stmt = partial(v, keywords=Ø, parens=assignments)
-
+        self.visit_expr_stmt = partial(v, keywords=Ø, parens=ASSIGNMENTS)
         self.visit_return_stmt = partial(v, keywords={"return"}, parens={"return"})
         self.visit_import_from = partial(v, keywords=Ø, parens={"import"})
         self.visit_del_stmt = partial(v, keywords=Ø, parens={"del"})
@@ -576,9 +552,7 @@ def transform_line(
     # We need the line string when power operators are hugging to determine if we should
     # split the line. Default to line_str, if no power operator are present on the line.
     line_str_hugging_power_ops = (
-        (_hugging_power_ops_line_to_string(line, features, mode) or line_str)
-        if Preview.fix_power_op_line_length in mode
-        else line_str
+        _hugging_power_ops_line_to_string(line, features, mode) or line_str
     )
 
     ll = mode.line_length
@@ -688,9 +662,6 @@ def should_split_funcdef_with_rhs(line: Line, mode: Mode) -> bool:
     """If a funcdef has a magic trailing comma in the return type, then we should first
     split the line with rhs to respect the comma.
     """
-    if Preview.respect_magic_trailing_comma_in_return_type not in mode:
-        return False
-
     return_type_leaves: List[Leaf] = []
     in_return_type = False
 
@@ -919,9 +890,6 @@ def _maybe_split_omitting_optional_parens(
         try:
             # The RHSResult Omitting Optional Parens.
             rhs_oop = _first_right_hand_split(line, omit=omit)
-            prefer_splitting_rhs_mode = (
-                Preview.prefer_splitting_right_hand_side_of_assignments in line.mode
-            )
             is_split_right_after_equal = (
                 len(rhs.head.leaves) >= 2 and rhs.head.leaves[-2].type == token.EQUAL
             )
@@ -937,8 +905,7 @@ def _maybe_split_omitting_optional_parens(
             )
             if (
                 not (
-                    prefer_splitting_rhs_mode
-                    and is_split_right_after_equal
+                    is_split_right_after_equal
                     and rhs_head_contains_brackets
                     and rhs_head_short_enough
                     and rhs_head_explode_blocked_by_magic_trailing_comma
@@ -1224,11 +1191,7 @@ def append_to_line(leaf: Leaf) -> Iterator[Line]:
                     trailing_comma_safe and Feature.TRAILING_COMMA_IN_CALL in features
                 )
 
-        if (
-            Preview.add_trailing_comma_consistently in mode
-            and last_leaf.type == STANDALONE_COMMENT
-            and leaf_idx == last_non_comment_leaf
-        ):
+        if last_leaf.type == STANDALONE_COMMENT and leaf_idx == last_non_comment_leaf:
             current_line = _safe_add_trailing_comma(
                 trailing_comma_safe, delimiter_priority, current_line
             )
@@ -1315,11 +1278,7 @@ def normalize_invisible_parens(  # noqa: C901
 
         # Fixes a bug where invisible parens are not properly wrapped around
         # case blocks.
-        if (
-            isinstance(child, Node)
-            and child.type == syms.case_block
-            and Preview.long_case_block_line_splitting in mode
-        ):
+        if isinstance(child, Node) and child.type == syms.case_block:
             normalize_invisible_parens(
                 child, parens_after={"case"}, mode=mode, features=features
             )
@@ -1374,7 +1333,6 @@ def normalize_invisible_parens(  # noqa: C901
                 and child.next_sibling is not None
                 and child.next_sibling.type == token.COLON
                 and child.value == "case"
-                and Preview.long_case_block_line_splitting in mode
             ):
                 # A special patch for "case case:" scenario, the second occurrence
                 # of case will be not parsed as a Python keyword.
@@ -1448,7 +1406,6 @@ def _maybe_wrap_cms_in_parens(
     """
     if (
         Feature.PARENTHESIZED_CONTEXT_MANAGERS not in features
-        or Preview.wrap_multiple_context_managers_in_parens not in mode
         or len(node.children) <= 2
         # If it's an atom, it's already wrapped in parens.
         or node.children[1].type == syms.atom