]> git.madduck.net Git - etc/vim.git/blobdiff - src/black/nodes.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 / nodes.py
index 8f2e15b2cc3013926bc0636c05995749645f8b20..edd201a21e9f2311d3482c1183f1d100e1f073ef 100644 (file)
@@ -3,34 +3,24 @@ blib2to3 Node/Leaf transformation-related utility functions.
 """
 
 import sys
-from typing import (
-    Collection,
-    Generic,
-    Iterator,
-    List,
-    Optional,
-    Set,
-    Tuple,
-    TypeVar,
-    Union,
-)
-
-if sys.version_info < (3, 8):
-    from typing_extensions import Final
+from typing import Final, Generic, Iterator, List, Optional, Set, Tuple, TypeVar, Union
+
+if sys.version_info >= (3, 10):
+    from typing import TypeGuard
 else:
-    from typing import Final
+    from typing_extensions import TypeGuard
 
-# lib2to3 fork
-from blib2to3.pytree import Node, Leaf, type_repr
-from blib2to3 import pygram
-from blib2to3.pgen2 import token
+from mypy_extensions import mypyc_attr
 
 from black.cache import CACHE_DIR
+from black.mode import Mode, Preview
 from black.strings import has_triple_quotes
-
+from blib2to3 import pygram
+from blib2to3.pgen2 import token
+from blib2to3.pytree import NL, Leaf, Node, type_repr
 
 pygram.initialize(CACHE_DIR)
-syms = pygram.python_symbols
+syms: Final = pygram.python_symbols
 
 
 # types
@@ -50,6 +40,8 @@ STATEMENT: Final = {
     syms.with_stmt,
     syms.funcdef,
     syms.classdef,
+    syms.match_stmt,
+    syms.case_block,
 }
 STANDALONE_COMMENT: Final = 153
 token.tok_name[STANDALONE_COMMENT] = "STANDALONE_COMMENT"
@@ -93,6 +85,8 @@ UNPACKING_PARENTS: Final = {
     syms.listmaker,
     syms.testlist_gexp,
     syms.testlist_star_expr,
+    syms.subject_expr,
+    syms.pattern,
 }
 TEST_DESCENDANTS: Final = {
     syms.test,
@@ -111,6 +105,7 @@ TEST_DESCENDANTS: Final = {
     syms.term,
     syms.power,
 }
+TYPED_NAMES: Final = {syms.tname, syms.tname_star}
 ASSIGNMENTS: Final = {
     "=",
     "+=",
@@ -128,16 +123,21 @@ ASSIGNMENTS: Final = {
     "//=",
 }
 
-IMPLICIT_TUPLE = {syms.testlist, syms.testlist_star_expr, syms.exprlist}
-BRACKET = {token.LPAR: token.RPAR, token.LSQB: token.RSQB, token.LBRACE: token.RBRACE}
-OPENING_BRACKETS = set(BRACKET.keys())
-CLOSING_BRACKETS = set(BRACKET.values())
-BRACKETS = OPENING_BRACKETS | CLOSING_BRACKETS
-ALWAYS_NO_SPACE = CLOSING_BRACKETS | {token.COMMA, STANDALONE_COMMENT}
+IMPLICIT_TUPLE: Final = {syms.testlist, syms.testlist_star_expr, syms.exprlist}
+BRACKET: Final = {
+    token.LPAR: token.RPAR,
+    token.LSQB: token.RSQB,
+    token.LBRACE: token.RBRACE,
+}
+OPENING_BRACKETS: Final = set(BRACKET.keys())
+CLOSING_BRACKETS: Final = set(BRACKET.values())
+BRACKETS: Final = OPENING_BRACKETS | CLOSING_BRACKETS
+ALWAYS_NO_SPACE: Final = CLOSING_BRACKETS | {token.COMMA, STANDALONE_COMMENT}
 
 RARROW = 55
 
 
+@mypyc_attr(allow_interpreted_subclasses=True)
 class Visitor(Generic[T]):
     """Basic lib2to3 visitor that yields things of type `T` on `visit()`."""
 
@@ -172,15 +172,15 @@ class Visitor(Generic[T]):
                 yield from self.visit(child)
 
 
-def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
+def whitespace(leaf: Leaf, *, complex_subscript: bool, mode: Mode) -> str:  # noqa: C901
     """Return whitespace prefix if needed for the given `leaf`.
 
     `complex_subscript` signals whether the given leaf is part of a subscription
     which has non-trivial arguments, like arithmetic expressions or function calls.
     """
-    NO = ""
-    SPACE = " "
-    DOUBLESPACE = "  "
+    NO: Final[str] = ""
+    SPACE: Final[str] = " "
+    DOUBLESPACE: Final[str] = "  "
     t = leaf.type
     p = leaf.parent
     v = leaf.value
@@ -229,6 +229,14 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
                     # that, too.
                     return prevp.prefix
 
+        elif (
+            prevp.type == token.STAR
+            and parent_type(prevp) == syms.star_expr
+            and parent_type(prevp.parent) == syms.subscriptlist
+        ):
+            # No space between typevar tuples.
+            return NO
+
         elif prevp.type in VARARGS_SPECIALS:
             if is_vararg(prevp, within=VARARGS_PARENTS | UNPACKING_PARENTS):
                 return NO
@@ -244,16 +252,6 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
         ):
             return NO
 
-        elif (
-            prevp.type == token.RIGHTSHIFT
-            and prevp.parent
-            and prevp.parent.type == syms.shift_expr
-            and prevp.prev_sibling
-            and prevp.prev_sibling.type == token.NAME
-            and prevp.prev_sibling.value == "print"  # type: ignore
-        ):
-            # Python 2 print chevron
-            return NO
         elif prevp.type == token.AT and p.parent and p.parent.type == syms.decorator:
             # no space in decorators
             return NO
@@ -277,7 +275,7 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
             return NO
 
         if t == token.EQUAL:
-            if prev.type != syms.tname:
+            if prev.type not in TYPED_NAMES:
                 return NO
 
         elif prev.type == token.EQUAL:
@@ -288,7 +286,7 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
         elif prev.type != token.COMMA:
             return NO
 
-    elif p.type == syms.tname:
+    elif p.type in TYPED_NAMES:
         # type names
         if not prev:
             prevp = preceding_leaf(p)
@@ -301,12 +299,7 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
             return NO
 
         if not prev:
-            if t == token.DOT:
-                prevp = preceding_leaf(p)
-                if not prevp or prevp.type != token.NUMBER:
-                    return NO
-
-            elif t == token.LSQB:
+            if t == token.DOT or t == token.LSQB:
                 return NO
 
         elif prev.type != token.COMMA:
@@ -353,6 +346,11 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
 
             return NO
 
+        elif Preview.walrus_subscript in mode and (
+            t == token.COLONEQUAL or prev.type == token.COLONEQUAL
+        ):
+            return SPACE
+
         elif not complex_subscript:
             return NO
 
@@ -402,6 +400,10 @@ def whitespace(leaf: Leaf, *, complex_subscript: bool) -> str:  # noqa: C901
     elif p.type == syms.sliceop:
         return NO
 
+    elif p.type == syms.except_clause:
+        if t == token.STAR:
+            return NO
+
     return SPACE
 
 
@@ -439,27 +441,6 @@ def prev_siblings_are(node: Optional[LN], tokens: List[Optional[NodeType]]) -> b
     return prev_siblings_are(node.prev_sibling, tokens[:-1])
 
 
-def last_two_except(leaves: List[Leaf], omit: Collection[LeafID]) -> Tuple[Leaf, Leaf]:
-    """Return (penultimate, last) leaves skipping brackets in `omit` and contents."""
-    stop_after = None
-    last = None
-    for leaf in reversed(leaves):
-        if stop_after:
-            if leaf is stop_after:
-                stop_after = None
-            continue
-
-        if last:
-            return leaf, last
-
-        if id(leaf) in omit:
-            stop_after = leaf.opening_bracket
-        else:
-            last = leaf
-    else:
-        raise LookupError("Last two leaves were also skipped")
-
-
 def parent_type(node: Optional[LN]) -> Optional[NodeType]:
     """
     Returns:
@@ -523,23 +504,24 @@ def container_of(leaf: Leaf) -> LN:
     return container
 
 
-def first_leaf_column(node: Node) -> Optional[int]:
-    """Returns the column of the first leaf child of a node."""
-    for child in node.children:
-        if isinstance(child, Leaf):
-            return child.column
-    return None
+def first_leaf_of(node: LN) -> Optional[Leaf]:
+    """Returns the first leaf of the node tree."""
+    if isinstance(node, Leaf):
+        return node
+    if node.children:
+        return first_leaf_of(node.children[0])
+    else:
+        return None
 
 
-def first_child_is_arith(node: Node) -> bool:
-    """Whether first child is an arithmetic or a binary arithmetic expression"""
-    expr_types = {
+def is_arith_like(node: LN) -> bool:
+    """Whether node is an arithmetic or a binary arithmetic expression"""
+    return node.type in {
         syms.arith_expr,
         syms.shift_expr,
         syms.xor_expr,
         syms.and_expr,
     }
-    return bool(node.children and node.children[0].type in expr_types)
 
 
 def is_docstring(leaf: Leaf) -> bool:
@@ -583,9 +565,25 @@ def is_one_tuple(node: LN) -> bool:
     )
 
 
-def is_one_tuple_between(opening: Leaf, closing: Leaf, leaves: List[Leaf]) -> bool:
-    """Return True if content between `opening` and `closing` looks like a one-tuple."""
-    if opening.type != token.LPAR and closing.type != token.RPAR:
+def is_tuple_containing_walrus(node: LN) -> bool:
+    """Return True if `node` holds a tuple that contains a walrus operator."""
+    if node.type != syms.atom:
+        return False
+    gexp = unwrap_singleton_parenthesis(node)
+    if gexp is None or gexp.type != syms.testlist_gexp:
+        return False
+
+    return any(child.type == syms.namedexpr_test for child in gexp.children)
+
+
+def is_one_sequence_between(
+    opening: Leaf,
+    closing: Leaf,
+    leaves: List[Leaf],
+    brackets: Tuple[int, int] = (token.LPAR, token.RPAR),
+) -> bool:
+    """Return True if content between `opening` and `closing` is a one-sequence."""
+    if (opening.type, closing.type) != brackets:
         return False
 
     depth = closing.bracket_depth + 1
@@ -676,7 +674,7 @@ def is_yield(node: LN) -> bool:
     if node.type == syms.yield_expr:
         return True
 
-    if node.type == token.NAME and node.value == "yield":  # type: ignore
+    if is_name_token(node) and node.value == "yield":
         return True
 
     if node.type != syms.atom:
@@ -722,6 +720,11 @@ def is_multiline_string(leaf: Leaf) -> bool:
 
 def is_stub_suite(node: Node) -> bool:
     """Return True if `node` is a suite with a stub body."""
+
+    # If there is a comment, we want to keep it.
+    if node.prefix.strip():
+        return False
+
     if (
         len(node.children) != 4
         or node.children[0].type != token.NEWLINE
@@ -730,6 +733,9 @@ def is_stub_suite(node: Node) -> bool:
     ):
         return False
 
+    if node.children[3].prefix.strip():
+        return False
+
     return is_stub_body(node.children[2])
 
 
@@ -743,7 +749,8 @@ def is_stub_body(node: LN) -> bool:
 
     child = node.children[0]
     return (
-        child.type == syms.atom
+        not child.prefix.strip()
+        and child.type == syms.atom
         and len(child.children) == 3
         and all(leaf == Leaf(token.DOT, ".") for leaf in child.children)
     )
@@ -793,12 +800,54 @@ def is_import(leaf: Leaf) -> bool:
     )
 
 
-def is_type_comment(leaf: Leaf, suffix: str = "") -> bool:
-    """Return True if the given leaf is a special comment.
-    Only returns true for type comments for now."""
+def is_with_or_async_with_stmt(leaf: Leaf) -> bool:
+    """Return True if the given leaf starts a with or async with statement."""
+    return bool(
+        leaf.type == token.NAME
+        and leaf.value == "with"
+        and leaf.parent
+        and leaf.parent.type == syms.with_stmt
+    ) or bool(
+        leaf.type == token.ASYNC
+        and leaf.next_sibling
+        and leaf.next_sibling.type == syms.with_stmt
+    )
+
+
+def is_async_stmt_or_funcdef(leaf: Leaf) -> bool:
+    """Return True if the given leaf starts an async def/for/with statement.
+
+    Note that `async def` can be either an `async_stmt` or `async_funcdef`,
+    the latter is used when it has decorators.
+    """
+    return bool(
+        leaf.type == token.ASYNC
+        and leaf.parent
+        and leaf.parent.type in {syms.async_stmt, syms.async_funcdef}
+    )
+
+
+def is_type_comment(leaf: Leaf) -> bool:
+    """Return True if the given leaf is a type comment. This function should only
+    be used for general type comments (excluding ignore annotations, which should
+    use `is_type_ignore_comment`). Note that general type comments are no longer
+    used in modern version of Python, this function may be deprecated in the future."""
     t = leaf.type
     v = leaf.value
-    return t in {token.COMMENT, STANDALONE_COMMENT} and v.startswith("# type:" + suffix)
+    return t in {token.COMMENT, STANDALONE_COMMENT} and v.startswith("# type:")
+
+
+def is_type_ignore_comment(leaf: Leaf) -> bool:
+    """Return True if the given leaf is a type comment with ignore annotation."""
+    t = leaf.type
+    v = leaf.value
+    return t in {token.COMMENT, STANDALONE_COMMENT} and is_type_ignore_comment_string(v)
+
+
+def is_type_ignore_comment_string(value: str) -> bool:
+    """Return True if the given string match with type comment with
+    ignore annotation."""
+    return value.startswith("# type: ignore")
 
 
 def wrap_in_parentheses(parent: Node, child: LN, *, visible: bool = True) -> None:
@@ -843,3 +892,35 @@ def ensure_visible(leaf: Leaf) -> None:
         leaf.value = "("
     elif leaf.type == token.RPAR:
         leaf.value = ")"
+
+
+def is_name_token(nl: NL) -> TypeGuard[Leaf]:
+    return nl.type == token.NAME
+
+
+def is_lpar_token(nl: NL) -> TypeGuard[Leaf]:
+    return nl.type == token.LPAR
+
+
+def is_rpar_token(nl: NL) -> TypeGuard[Leaf]:
+    return nl.type == token.RPAR
+
+
+def is_string_token(nl: NL) -> TypeGuard[Leaf]:
+    return nl.type == token.STRING
+
+
+def is_number_token(nl: NL) -> TypeGuard[Leaf]:
+    return nl.type == token.NUMBER
+
+
+def is_part_of_annotation(leaf: Leaf) -> bool:
+    """Returns whether this leaf is part of type annotations."""
+    ancestor = leaf.parent
+    while ancestor is not None:
+        if ancestor.prev_sibling and ancestor.prev_sibling.type == token.RARROW:
+            return True
+        if ancestor.parent and ancestor.parent.type == syms.tname:
+            return True
+        ancestor = ancestor.parent
+    return False