X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/b719d85ccc330170e40b2617307a7e3b2a0bab14..cb67a322423cb2b4bcfb2faa68fb769326741d6a:/black.py diff --git a/black.py b/black.py index b65693e..f2c450c 100644 --- a/black.py +++ b/black.py @@ -79,15 +79,15 @@ syms = pygram.python_symbols class NothingChanged(UserWarning): - """Raised by :func:`format_file` when reformatted code is the same as source.""" + """Raised when reformatted code is the same as source.""" class CannotSplit(Exception): - """A readable split that fits the allotted line length is impossible. + """A readable split that fits the allotted line length is impossible.""" - Raised by :func:`left_hand_split`, :func:`right_hand_split`, and - :func:`delimiter_split`. - """ + +class InvalidInput(ValueError): + """Raised when input source code fails all parse attempts.""" class WriteBack(Enum): @@ -676,7 +676,7 @@ def lib2to3_parse(src_txt: str) -> Node: faulty_line = lines[lineno - 1] except IndexError: faulty_line = "" - exc = ValueError(f"Cannot parse: {lineno}:{column}: {faulty_line}") + exc = InvalidInput(f"Cannot parse: {lineno}:{column}: {faulty_line}") else: raise exc from None @@ -877,8 +877,8 @@ class BracketTracker: bracket_match: Dict[Tuple[Depth, NodeType], Leaf] = Factory(dict) delimiters: Dict[LeafID, Priority] = Factory(dict) previous: Optional[Leaf] = None - _for_loop_variable: int = 0 - _lambda_arguments: int = 0 + _for_loop_depths: List[int] = Factory(list) + _lambda_argument_depths: List[int] = Factory(list) def mark(self, leaf: Leaf) -> None: """Mark `leaf` with bracket-related metadata. Keep track of delimiters. @@ -951,16 +951,21 @@ class BracketTracker: """ if leaf.type == token.NAME and leaf.value == "for": self.depth += 1 - self._for_loop_variable += 1 + self._for_loop_depths.append(self.depth) return True return False def maybe_decrement_after_for_loop_variable(self, leaf: Leaf) -> bool: """See `maybe_increment_for_loop_variable` above for explanation.""" - if self._for_loop_variable and leaf.type == token.NAME and leaf.value == "in": + if ( + self._for_loop_depths + and self._for_loop_depths[-1] == self.depth + and leaf.type == token.NAME + and leaf.value == "in" + ): self.depth -= 1 - self._for_loop_variable -= 1 + self._for_loop_depths.pop() return True return False @@ -973,16 +978,20 @@ class BracketTracker: """ if leaf.type == token.NAME and leaf.value == "lambda": self.depth += 1 - self._lambda_arguments += 1 + self._lambda_argument_depths.append(self.depth) return True return False def maybe_decrement_after_lambda_arguments(self, leaf: Leaf) -> bool: """See `maybe_increment_lambda_arguments` above for explanation.""" - if self._lambda_arguments and leaf.type == token.COLON: + if ( + self._lambda_argument_depths + and self._lambda_argument_depths[-1] == self.depth + and leaf.type == token.COLON + ): self.depth -= 1 - self._lambda_arguments -= 1 + self._lambda_argument_depths.pop() return True return False @@ -1880,7 +1889,7 @@ def is_split_after_delimiter(leaf: Leaf, previous: Leaf = None) -> int: def is_split_before_delimiter(leaf: Leaf, previous: Leaf = None) -> int: - """Return the priority of the `leaf` delimiter, given a line before after it. + """Return the priority of the `leaf` delimiter, given a line break before it. The delimiter priorities returned here are from those delimiters that would cause a line break before themselves. @@ -1917,15 +1926,20 @@ def is_split_before_delimiter(leaf: Leaf, previous: Leaf = None) -> int: ): return STRING_PRIORITY - if leaf.type != token.NAME: + if leaf.type not in {token.NAME, token.ASYNC}: return 0 if ( leaf.value == "for" and leaf.parent and leaf.parent.type in {syms.comp_for, syms.old_comp_for} + or leaf.type == token.ASYNC ): - return COMPREHENSION_PRIORITY + if ( + not isinstance(leaf.prev_sibling, Leaf) + or leaf.prev_sibling.value != "async" + ): + return COMPREHENSION_PRIORITY if ( leaf.value == "if" @@ -1999,6 +2013,16 @@ def generate_comments(leaf: LN) -> Iterator[Leaf]: @dataclass class ProtoComment: + """Describes a piece of syntax that is a comment. + + It's not a :class:`blib2to3.pytree.Leaf` so that: + + * it can be cached (`Leaf` objects should not be reused more than once as + they store their lineno, column, prefix, and parent information); + * `newlines` and `consumed` fields are kept separate from the `value`. This + simplifies handling of special marker comments like ``# fmt: off/on``. + """ + type: int # token.COMMENT or STANDALONE_COMMENT value: str # content of the comment newlines: int # how many newlines before the comment @@ -2007,6 +2031,7 @@ class ProtoComment: @lru_cache(maxsize=4096) def list_comments(prefix: str, *, is_endmarker: bool) -> List[ProtoComment]: + """Return a list of :class:`ProtoComment` objects parsed from the given `prefix`.""" result: List[ProtoComment] = [] if not prefix or "#" not in prefix: return result @@ -2513,8 +2538,8 @@ def normalize_string_quotes(leaf: Leaf) -> None: def normalize_numeric_literal(leaf: Leaf, allow_underscores: bool) -> None: """Normalizes numeric (float, int, and complex) literals. - All letters used in the representation are normalized to lowercase, long number - literals are split using underscores. + All letters used in the representation are normalized to lowercase (except + in Python 2 long literals), and long number literals are split using underscores. """ text = leaf.value.lower() if text.startswith(("0o", "0x", "0b")): @@ -2534,6 +2559,9 @@ def normalize_numeric_literal(leaf: Leaf, allow_underscores: bool) -> None: elif text.endswith(("j", "l")): number = text[:-1] suffix = text[-1] + # Capitalize in "2L" because "l" looks too similar to "1". + if suffix == "l": + suffix = "L" text = f"{format_float_or_int_string(number, allow_underscores)}{suffix}" else: text = format_float_or_int_string(text, allow_underscores) @@ -2547,14 +2575,22 @@ def format_float_or_int_string(text: str, allow_underscores: bool) -> str: before, after = text.split(".") before = format_int_string(before, allow_underscores) if before else "0" - after = format_int_string(after, allow_underscores) if after else "0" + if after: + after = format_int_string(after, allow_underscores, count_from_end=False) + else: + after = "0" return f"{before}.{after}" -def format_int_string(text: str, allow_underscores: bool) -> str: +def format_int_string( + text: str, allow_underscores: bool, count_from_end: bool = True +) -> str: """Normalizes underscores in a string to e.g. 1_000_000. - Input must be a string of at least six digits and optional underscores. + Input must be a string of digits and optional underscores. + If count_from_end is False, we add underscores after groups of three digits + counting from the beginning instead of the end of the strings. This is used + for the fractional part of float literals. """ if not allow_underscores: return text @@ -2564,7 +2600,12 @@ def format_int_string(text: str, allow_underscores: bool) -> str: # No underscores for numbers <= 6 digits long. return text - return format(int(text), "3_") + if count_from_end: + # Avoid removing leading zeros, which are important if we're formatting + # part of a number like "0.001". + return format(int("1" + text), "3_")[1:].lstrip("_") + else: + return "_".join(text[i : i + 3] for i in range(0, len(text), 3)) def normalize_invisible_parens(node: Node, parens_after: Set[str]) -> None: @@ -2585,7 +2626,11 @@ def normalize_invisible_parens(node: Node, parens_after: Set[str]) -> None: for index, child in enumerate(list(node.children)): if check_lpar: if child.type == syms.atom: - maybe_make_parens_invisible_in_atom(child) + if maybe_make_parens_invisible_in_atom(child): + lpar = Leaf(token.LPAR, "") + rpar = Leaf(token.RPAR, "") + index = child.remove() or 0 + node.insert_child(index, Node(syms.atom, [lpar, child, rpar])) elif is_one_tuple(child): # wrap child in visible parentheses lpar = Leaf(token.LPAR, "(") @@ -2693,7 +2738,11 @@ def generate_ignored_nodes(leaf: Leaf) -> Iterator[LN]: def maybe_make_parens_invisible_in_atom(node: LN) -> bool: - """If it's safe, make the parens in the atom `node` invisible, recursively.""" + """If it's safe, make the parens in the atom `node` invisible, recursively. + + Returns whether the node should itself be wrapped in invisible parentheses. + + """ if ( node.type != syms.atom or is_empty_tuple(node) @@ -2711,9 +2760,9 @@ def maybe_make_parens_invisible_in_atom(node: LN) -> bool: last.value = "" # type: ignore if len(node.children) > 1: maybe_make_parens_invisible_in_atom(node.children[1]) - return True + return False - return False + return True def is_empty_tuple(node: LN) -> bool: @@ -2938,7 +2987,6 @@ def generate_trailers_to_omit(line: Line, line_length: int) -> Iterator[Set[Leaf length = 4 * line.depth opening_bracket = None closing_bracket = None - optional_brackets: Set[LeafID] = set() inner_brackets: Set[LeafID] = set() for index, leaf, leaf_length in enumerate_with_length(line, reversed=True): length += leaf_length @@ -2949,17 +2997,12 @@ def generate_trailers_to_omit(line: Line, line_length: int) -> Iterator[Set[Leaf if leaf.type == STANDALONE_COMMENT or has_inline_comment: break - optional_brackets.discard(id(leaf)) if opening_bracket: if leaf is opening_bracket: opening_bracket = None elif leaf.type in CLOSING_BRACKETS: inner_brackets.add(id(leaf)) elif leaf.type in CLOSING_BRACKETS: - if not leaf.value: - optional_brackets.add(id(opening_bracket)) - continue - if index > 0 and line.leaves[index - 1].type in OPENING_BRACKETS: # Empty brackets would fail a split so treat them as "inner" # brackets (e.g. only add them to the `omit` set if another @@ -2967,13 +3010,15 @@ def generate_trailers_to_omit(line: Line, line_length: int) -> Iterator[Set[Leaf inner_brackets.add(id(leaf)) continue - opening_bracket = leaf.opening_bracket if closing_bracket: omit.add(id(closing_bracket)) omit.update(inner_brackets) inner_brackets.clear() yield omit - closing_bracket = leaf + + if leaf.value: + opening_bracket = leaf.opening_bracket + closing_bracket = leaf def get_future_imports(node: Node) -> Set[str]: