X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/6316e293ac30a2837ec20eba289fd28a2a18cf89..a9f50cd0b58259a11a1c851bde7b4f11321e5b3b:/black.py diff --git a/black.py b/black.py index 7935cdc..da7af03 100644 --- a/black.py +++ b/black.py @@ -10,7 +10,7 @@ from pathlib import Path import tokenize import sys from typing import ( - Dict, Generic, Iterable, Iterator, List, Optional, Set, Tuple, TypeVar, Union + Dict, Generic, Iterable, Iterator, List, Optional, Set, Tuple, Type, TypeVar, Union ) from attr import dataclass, Factory @@ -22,7 +22,7 @@ from blib2to3 import pygram, pytree from blib2to3.pgen2 import driver, token from blib2to3.pgen2.parse import ParseError -__version__ = "18.3a3" +__version__ = "18.3a4" DEFAULT_LINE_LENGTH = 88 # types syms = pygram.python_symbols @@ -48,6 +48,34 @@ class CannotSplit(Exception): """ +class FormatError(Exception): + """Base fmt: on/off error. + + It holds the number of bytes of the prefix consumed before the format + control comment appeared. + """ + + def __init__(self, consumed: int) -> None: + super().__init__(consumed) + self.consumed = consumed + + def trim_prefix(self, leaf: Leaf) -> None: + leaf.prefix = leaf.prefix[self.consumed:] + + def leaf_from_consumed(self, leaf: Leaf) -> Leaf: + """Returns a new Leaf from the consumed part of the prefix.""" + unformatted_prefix = leaf.prefix[:self.consumed] + return Leaf(token.NEWLINE, unformatted_prefix) + + +class FormatOn(FormatError): + """Found a comment like `# fmt: on` in the file.""" + + +class FormatOff(FormatError): + """Found a comment like `# fmt: off` in the file.""" + + @click.command() @click.option( '-l', @@ -320,6 +348,15 @@ class DebugVisitor(Visitor[T]): out(f' {node.prefix!r}', fg='green', bold=False, nl=False) out(f' {node.value!r}', fg='blue', bold=False) + @classmethod + def show(cls, code: str) -> None: + """Pretty-prints a given string of `code`. + + Convenience method for debugging. + """ + v: DebugVisitor[None] = DebugVisitor() + list(v.visit(lib2to3_parse(code))) + KEYWORDS = set(keyword.kwlist) WHITESPACE = {token.DEDENT, token.INDENT, token.NEWLINE} @@ -419,7 +456,7 @@ class BracketTracker: """Returns True if there is an yet unmatched open bracket on the line.""" return bool(self.bracket_match) - def max_priority(self, exclude: Iterable[LeafID] =()) -> int: + def max_priority(self, exclude: Iterable[LeafID] = ()) -> int: """Returns the highest priority of a delimiter found on the line. Values are consistent with what `is_delimiter()` returns. @@ -649,6 +686,43 @@ class Line: return bool(self.leaves or self.comments) +class UnformattedLines(Line): + + def append(self, leaf: Leaf, preformatted: bool = False) -> None: + try: + list(generate_comments(leaf)) + except FormatOn as f_on: + self.leaves.append(f_on.leaf_from_consumed(leaf)) + raise + + self.leaves.append(leaf) + if leaf.type == token.INDENT: + self.depth += 1 + elif leaf.type == token.DEDENT: + self.depth -= 1 + + def append_comment(self, comment: Leaf) -> bool: + raise NotImplementedError("Unformatted lines don't store comments separately.") + + def maybe_remove_trailing_comma(self, closing: Leaf) -> bool: + return False + + def maybe_increment_for_loop_variable(self, leaf: Leaf) -> bool: + return False + + def maybe_adapt_standalone_comment(self, comment: Leaf) -> bool: + return False + + def __str__(self) -> str: + if not self: + return '\n' + + res = '' + for leaf in self.leaves: + res += str(leaf) + return res + + @dataclass class EmptyLineTracker: """Provides a stateful method that returns the number of potential extra @@ -669,6 +743,9 @@ class EmptyLineTracker: (two on module-level), as well as providing an extra empty line after flow control keywords to make them more prominent. """ + if isinstance(current_line, UnformattedLines): + return 0, 0 + before, after = self._maybe_empty_lines(current_line) before -= self.previous_after self.previous_after = after @@ -677,13 +754,13 @@ class EmptyLineTracker: def _maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]: max_allowed = 1 - if current_line.is_comment and current_line.depth == 0: + if current_line.depth == 0: max_allowed = 2 if current_line.leaves: # Consume the first leaf's extra newlines. first_leaf = current_line.leaves[0] before = first_leaf.prefix.count('\n') - before = min(before, max(before, max_allowed)) + before = min(before, max_allowed) first_leaf.prefix = '' else: before = 0 @@ -738,7 +815,7 @@ class LineGenerator(Visitor[Line]): """ current_line: Line = Factory(Line) - def line(self, indent: int = 0) -> Iterator[Line]: + def line(self, indent: int = 0, type: Type[Line] = Line) -> Iterator[Line]: """Generate a line. If the line is empty, only emit if it makes sense. @@ -747,35 +824,60 @@ class LineGenerator(Visitor[Line]): If any lines were generated, set up a new current_line. """ if not self.current_line: - self.current_line.depth += indent + if self.current_line.__class__ == type: + self.current_line.depth += indent + else: + self.current_line = type(depth=self.current_line.depth + indent) return # Line is empty, don't emit. Creating a new one unnecessary. complete_line = self.current_line - self.current_line = Line(depth=complete_line.depth + indent) + self.current_line = type(depth=complete_line.depth + indent) yield complete_line + def visit(self, node: LN) -> Iterator[Line]: + """High-level entry point to the visitor.""" + if isinstance(self.current_line, UnformattedLines): + # File contained `# fmt: off` + yield from self.visit_unformatted(node) + + else: + yield from super().visit(node) + def visit_default(self, node: LN) -> Iterator[Line]: if isinstance(node, Leaf): any_open_brackets = self.current_line.bracket_tracker.any_open_brackets() - for comment in generate_comments(node): - if any_open_brackets: - # any comment within brackets is subject to splitting - self.current_line.append(comment) - elif comment.type == token.COMMENT: - # regular trailing comment - self.current_line.append(comment) - yield from self.line() - - else: - # regular standalone comment - yield from self.line() - - self.current_line.append(comment) - yield from self.line() - - normalize_prefix(node, inside_brackets=any_open_brackets) - if node.type not in WHITESPACE: - self.current_line.append(node) + try: + for comment in generate_comments(node): + if any_open_brackets: + # any comment within brackets is subject to splitting + self.current_line.append(comment) + elif comment.type == token.COMMENT: + # regular trailing comment + self.current_line.append(comment) + yield from self.line() + + else: + # regular standalone comment + yield from self.line() + + self.current_line.append(comment) + yield from self.line() + + except FormatOff as f_off: + f_off.trim_prefix(node) + yield from self.line(type=UnformattedLines) + yield from self.visit(node) + + except FormatOn as f_on: + # This only happens here if somebody says "fmt: on" multiple + # times in a row. + f_on.trim_prefix(node) + yield from self.visit_default(node) + + else: + normalize_prefix(node, inside_brackets=any_open_brackets) + if node.type not in WHITESPACE: + self.current_line.append(node) yield from super().visit_default(node) def visit_INDENT(self, node: Node) -> Iterator[Line]: @@ -783,6 +885,7 @@ class LineGenerator(Visitor[Line]): yield from self.visit_default(node) def visit_DEDENT(self, node: Node) -> Iterator[Line]: + # DEDENT has no value. Additionally, in blib2to3 it never holds comments. yield from self.line(-1) def visit_stmt(self, node: Node, keywords: Set[str]) -> Iterator[Line]: @@ -835,6 +938,19 @@ class LineGenerator(Visitor[Line]): yield from self.visit_default(leaf) yield from self.line() + def visit_unformatted(self, node: LN) -> Iterator[Line]: + if isinstance(node, Node): + for child in node.children: + yield from self.visit(child) + + else: + try: + self.current_line.append(node) + except FormatOn as f_on: + f_on.trim_prefix(node) + yield from self.line() + yield from self.visit(node) + def __attrs_post_init__(self) -> None: """You are in a twisty little maze of passages.""" v = self.visit_stmt @@ -885,14 +1001,17 @@ def whitespace(leaf: Leaf) -> str: # noqa C901 return SPACE if prevp.type == token.COMMA else NO if prevp.type == token.EQUAL: - if prevp.parent and prevp.parent.type in { - syms.arglist, - syms.argument, - syms.parameters, - syms.typedargslist, - syms.varargslist, - }: - return NO + if prevp.parent: + if prevp.parent.type in { + syms.arglist, syms.argument, syms.parameters, syms.varargslist + }: + return NO + + elif prevp.parent.type == syms.typedargslist: + # A bit hacky: if the equal sign has whitespace, it means we + # previously found it's a typed argument. So, we're using + # that, too. + return prevp.prefix elif prevp.type == token.DOUBLESTAR: if prevp.parent and prevp.parent.type in { @@ -922,7 +1041,7 @@ def whitespace(leaf: Leaf) -> str: # noqa C901 and prevp.parent.type == syms.shift_expr and prevp.prev_sibling and prevp.prev_sibling.type == token.NAME - and prevp.prev_sibling.value == 'print' + and prevp.prev_sibling.value == 'print' # type: ignore ): # Python 2 print chevron return NO @@ -938,7 +1057,7 @@ def whitespace(leaf: Leaf) -> str: # noqa C901 if not prev or prev.type != token.COMMA: return NO - if p.type == syms.varargslist: + elif p.type == syms.varargslist: # lambdas if t == token.RPAR: return NO @@ -1156,8 +1275,10 @@ def generate_comments(leaf: Leaf) -> Iterator[Leaf]: if '#' not in p: return + consumed = 0 nlines = 0 for index, line in enumerate(p.split('\n')): + consumed += len(line) + 1 # adding the length of the split '\n' line = line.lstrip() if not line: nlines += 1 @@ -1168,7 +1289,14 @@ def generate_comments(leaf: Leaf) -> Iterator[Leaf]: comment_type = token.COMMENT # simple trailing comment else: comment_type = STANDALONE_COMMENT - yield Leaf(comment_type, make_comment(line), prefix='\n' * nlines) + comment = make_comment(line) + yield Leaf(comment_type, comment, prefix='\n' * nlines) + + if comment in {'# fmt: on', '# yapf: enable'}: + raise FormatOn(consumed) + + if comment in {'# fmt: off', '# yapf: disable'}: + raise FormatOff(consumed) nlines = 0 @@ -1180,7 +1308,7 @@ def make_comment(content: str) -> str: if content[0] == '#': content = content[1:] - if content and content[0] not in {' ', '!', '#'}: + if content and content[0] not in ' !:#': content = ' ' + content return '#' + content @@ -1198,6 +1326,10 @@ def split_line( If `py36` is True, splitting may generate syntax that is only compatible with Python 3.6 and later. """ + if isinstance(line, UnformattedLines): + yield line + return + line_str = str(line).strip('\n') if len(line_str) <= line_length and '\n' not in line_str: yield line