- `mode` determines formatting options, such as how many characters per line are
- allowed. Example:
-
- >>> import black
- >>> print(black.format_str("def f(arg:str='')->None:...", mode=Mode()))
- def f(arg: str = "") -> None:
- ...
-
- A more complex example:
- >>> print(
- ... black.format_str(
- ... "def f(arg:str='')->None: hey",
- ... mode=black.Mode(
- ... target_versions={black.TargetVersion.PY36},
- ... line_length=10,
- ... string_normalization=False,
- ... is_pyi=False,
- ... ),
- ... ),
- ... )
- def f(
- arg: str = '',
- ) -> None:
- hey
-
- """
- src_node = lib2to3_parse(src_contents.lstrip(), mode.target_versions)
- dst_contents = []
- future_imports = get_future_imports(src_node)
- if mode.target_versions:
- versions = mode.target_versions
- else:
- versions = detect_target_versions(src_node)
- normalize_fmt_off(src_node)
- lines = LineGenerator(
- remove_u_prefix="unicode_literals" in future_imports
- or supports_feature(versions, Feature.UNICODE_LITERALS),
- is_pyi=mode.is_pyi,
- normalize_strings=mode.string_normalization,
- )
- elt = EmptyLineTracker(is_pyi=mode.is_pyi)
- empty_line = Line()
- after = 0
- split_line_features = {
- feature
- for feature in {Feature.TRAILING_COMMA_IN_CALL, Feature.TRAILING_COMMA_IN_DEF}
- if supports_feature(versions, feature)
- }
- for current_line in lines.visit(src_node):
- dst_contents.append(str(empty_line) * after)
- before, after = elt.maybe_empty_lines(current_line)
- dst_contents.append(str(empty_line) * before)
- for line in transform_line(
- current_line,
- line_length=mode.line_length,
- normalize_strings=mode.string_normalization,
- features=split_line_features,
- ):
- dst_contents.append(str(line))
- return "".join(dst_contents)
-
-
-def decode_bytes(src: bytes) -> Tuple[FileContent, Encoding, NewLine]:
- """Return a tuple of (decoded_contents, encoding, newline).
-
- `newline` is either CRLF or LF but `decoded_contents` is decoded with
- universal newlines (i.e. only contains LF).
- """
- srcbuf = io.BytesIO(src)
- encoding, lines = tokenize.detect_encoding(srcbuf.readline)
- if not lines:
- return "", encoding, "\n"
-
- newline = "\r\n" if b"\r\n" == lines[0][-2:] else "\n"
- srcbuf.seek(0)
- with io.TextIOWrapper(srcbuf, encoding) as tiow:
- return tiow.read(), encoding, newline
-
-
-def get_grammars(target_versions: Set[TargetVersion]) -> List[Grammar]:
- if not target_versions:
- # No target_version specified, so try all grammars.
- return [
- # Python 3.7+
- pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords,
- # Python 3.0-3.6
- pygram.python_grammar_no_print_statement_no_exec_statement,
- # Python 2.7 with future print_function import
- pygram.python_grammar_no_print_statement,
- # Python 2.7
- pygram.python_grammar,
- ]
-
- if all(version.is_python2() for version in target_versions):
- # Python 2-only code, so try Python 2 grammars.
- return [
- # Python 2.7 with future print_function import
- pygram.python_grammar_no_print_statement,
- # Python 2.7
- pygram.python_grammar,
- ]
-
- # Python 3-compatible code, so only try Python 3 grammar.
- grammars = []
- # If we have to parse both, try to parse async as a keyword first
- if not supports_feature(target_versions, Feature.ASYNC_IDENTIFIERS):
- # Python 3.7+
- grammars.append(
- pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords
- )
- if not supports_feature(target_versions, Feature.ASYNC_KEYWORDS):
- # Python 3.0-3.6
- grammars.append(pygram.python_grammar_no_print_statement_no_exec_statement)
- # At least one of the above branches must have been taken, because every Python
- # version has exactly one of the two 'ASYNC_*' flags
- return grammars
-
-
-def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) -> Node:
- """Given a string with source, return the lib2to3 Node."""
- if src_txt[-1:] != "\n":
- src_txt += "\n"
-
- for grammar in get_grammars(set(target_versions)):
- drv = driver.Driver(grammar, pytree.convert)
- try:
- result = drv.parse_string(src_txt, True)
- break
-
- except ParseError as pe:
- lineno, column = pe.context[1]
- lines = src_txt.splitlines()
- try:
- faulty_line = lines[lineno - 1]
- except IndexError:
- faulty_line = "<line number missing in source>"
- exc = InvalidInput(f"Cannot parse: {lineno}:{column}: {faulty_line}")
- else:
- raise exc from None
-
- if isinstance(result, Leaf):
- result = Node(syms.file_input, [result])
- return result
-
-
-def lib2to3_unparse(node: Node) -> str:
- """Given a lib2to3 node, return its string representation."""
- code = str(node)
- return code
-
-
-class Visitor(Generic[T]):
- """Basic lib2to3 visitor that yields things of type `T` on `visit()`."""
-
- def visit(self, node: LN) -> Iterator[T]:
- """Main method to visit `node` and its children.
-
- It tries to find a `visit_*()` method for the given `node.type`, like
- `visit_simple_stmt` for Node objects or `visit_INDENT` for Leaf objects.
- If no dedicated `visit_*()` method is found, chooses `visit_default()`
- instead.
-
- Then yields objects of type `T` from the selected visitor.
- """
- if node.type < 256:
- name = token.tok_name[node.type]
- else:
- name = str(type_repr(node.type))
- # We explicitly branch on whether a visitor exists (instead of
- # using self.visit_default as the default arg to getattr) in order
- # to save needing to create a bound method object and so mypyc can
- # generate a native call to visit_default.
- visitf = getattr(self, f"visit_{name}", None)
- if visitf:
- yield from visitf(node)
- else:
- yield from self.visit_default(node)
-
- def visit_default(self, node: LN) -> Iterator[T]:
- """Default `visit_*()` implementation. Recurses to children of `node`."""
- if isinstance(node, Node):
- for child in node.children:
- yield from self.visit(child)
-
-
-@dataclass
-class DebugVisitor(Visitor[T]):
- tree_depth: int = 0
-
- def visit_default(self, node: LN) -> Iterator[T]:
- indent = " " * (2 * self.tree_depth)
- if isinstance(node, Node):
- _type = type_repr(node.type)
- out(f"{indent}{_type}", fg="yellow")
- self.tree_depth += 1
- for child in node.children:
- yield from self.visit(child)
-
- self.tree_depth -= 1
- out(f"{indent}/{_type}", fg="yellow", bold=False)
- else:
- _type = token.tok_name.get(node.type, str(node.type))
- out(f"{indent}{_type}", fg="blue", nl=False)
- if node.prefix:
- # We don't have to handle prefixes for `Node` objects since
- # that delegates to the first child anyway.
- 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: Union[str, Leaf, Node]) -> None:
- """Pretty-print the lib2to3 AST of a given string of `code`.
-
- Convenience method for debugging.
- """
- v: DebugVisitor[None] = DebugVisitor()
- if isinstance(code, str):
- code = lib2to3_parse(code)
- list(v.visit(code))
-
-
-WHITESPACE: Final = {token.DEDENT, token.INDENT, token.NEWLINE}
-STATEMENT: Final = {
- syms.if_stmt,
- syms.while_stmt,
- syms.for_stmt,
- syms.try_stmt,
- syms.except_clause,
- syms.with_stmt,
- syms.funcdef,
- syms.classdef,
-}
-STANDALONE_COMMENT: Final = 153
-token.tok_name[STANDALONE_COMMENT] = "STANDALONE_COMMENT"
-LOGIC_OPERATORS: Final = {"and", "or"}
-COMPARATORS: Final = {
- token.LESS,
- token.GREATER,
- token.EQEQUAL,
- token.NOTEQUAL,
- token.LESSEQUAL,
- token.GREATEREQUAL,
-}
-MATH_OPERATORS: Final = {
- token.VBAR,
- token.CIRCUMFLEX,
- token.AMPER,
- token.LEFTSHIFT,
- token.RIGHTSHIFT,
- token.PLUS,
- token.MINUS,
- token.STAR,
- token.SLASH,
- token.DOUBLESLASH,
- token.PERCENT,
- token.AT,
- token.TILDE,
- token.DOUBLESTAR,
-}
-STARS: Final = {token.STAR, token.DOUBLESTAR}
-VARARGS_SPECIALS: Final = STARS | {token.SLASH}
-VARARGS_PARENTS: Final = {
- syms.arglist,
- syms.argument, # double star in arglist
- syms.trailer, # single argument to call
- syms.typedargslist,
- syms.varargslist, # lambdas
-}
-UNPACKING_PARENTS: Final = {
- syms.atom, # single element of a list or set literal
- syms.dictsetmaker,
- syms.listmaker,
- syms.testlist_gexp,
- syms.testlist_star_expr,
-}
-TEST_DESCENDANTS: Final = {
- syms.test,
- syms.lambdef,
- syms.or_test,
- syms.and_test,
- syms.not_test,
- syms.comparison,
- syms.star_expr,
- syms.expr,
- syms.xor_expr,
- syms.and_expr,
- syms.shift_expr,
- syms.arith_expr,
- syms.trailer,
- syms.term,
- syms.power,
-}
-ASSIGNMENTS: Final = {
- "=",
- "+=",
- "-=",
- "*=",
- "@=",
- "/=",
- "%=",
- "&=",
- "|=",
- "^=",
- "<<=",
- ">>=",
- "**=",
- "//=",
-}
-COMPREHENSION_PRIORITY: Final = 20
-COMMA_PRIORITY: Final = 18
-TERNARY_PRIORITY: Final = 16
-LOGIC_PRIORITY: Final = 14
-STRING_PRIORITY: Final = 12
-COMPARATOR_PRIORITY: Final = 10
-MATH_PRIORITIES: Final = {
- token.VBAR: 9,
- token.CIRCUMFLEX: 8,
- token.AMPER: 7,
- token.LEFTSHIFT: 6,
- token.RIGHTSHIFT: 6,
- token.PLUS: 5,
- token.MINUS: 5,
- token.STAR: 4,
- token.SLASH: 4,
- token.DOUBLESLASH: 4,
- token.PERCENT: 4,
- token.AT: 4,
- token.TILDE: 3,
- token.DOUBLESTAR: 2,
-}
-DOT_PRIORITY: Final = 1
-
-
-@dataclass
-class BracketTracker:
- """Keeps track of brackets on a line."""
-
- depth: int = 0
- bracket_match: Dict[Tuple[Depth, NodeType], Leaf] = field(default_factory=dict)
- delimiters: Dict[LeafID, Priority] = field(default_factory=dict)
- previous: Optional[Leaf] = None
- _for_loop_depths: List[int] = field(default_factory=list)
- _lambda_argument_depths: List[int] = field(default_factory=list)
-
- def mark(self, leaf: Leaf) -> None:
- """Mark `leaf` with bracket-related metadata. Keep track of delimiters.
-
- All leaves receive an int `bracket_depth` field that stores how deep
- within brackets a given leaf is. 0 means there are no enclosing brackets
- that started on this line.
-
- If a leaf is itself a closing bracket, it receives an `opening_bracket`
- field that it forms a pair with. This is a one-directional link to
- avoid reference cycles.
-
- If a leaf is a delimiter (a token on which Black can split the line if
- needed) and it's on depth 0, its `id()` is stored in the tracker's
- `delimiters` field.
- """
- if leaf.type == token.COMMENT:
- return
-
- self.maybe_decrement_after_for_loop_variable(leaf)
- self.maybe_decrement_after_lambda_arguments(leaf)
- if leaf.type in CLOSING_BRACKETS:
- self.depth -= 1
- opening_bracket = self.bracket_match.pop((self.depth, leaf.type))
- leaf.opening_bracket = opening_bracket
- leaf.bracket_depth = self.depth
- if self.depth == 0:
- delim = is_split_before_delimiter(leaf, self.previous)
- if delim and self.previous is not None:
- self.delimiters[id(self.previous)] = delim
- else:
- delim = is_split_after_delimiter(leaf, self.previous)
- if delim:
- self.delimiters[id(leaf)] = delim
- if leaf.type in OPENING_BRACKETS:
- self.bracket_match[self.depth, BRACKET[leaf.type]] = leaf
- self.depth += 1
- self.previous = leaf
- self.maybe_increment_lambda_arguments(leaf)
- self.maybe_increment_for_loop_variable(leaf)
-
- def any_open_brackets(self) -> bool:
- """Return True if there is an yet unmatched open bracket on the line."""
- return bool(self.bracket_match)
-
- def max_delimiter_priority(self, exclude: Iterable[LeafID] = ()) -> Priority:
- """Return the highest priority of a delimiter found on the line.
-
- Values are consistent with what `is_split_*_delimiter()` return.
- Raises ValueError on no delimiters.
- """
- return max(v for k, v in self.delimiters.items() if k not in exclude)
-
- def delimiter_count_with_priority(self, priority: Priority = 0) -> int:
- """Return the number of delimiters with the given `priority`.
-
- If no `priority` is passed, defaults to max priority on the line.
- """
- if not self.delimiters:
- return 0
-
- priority = priority or self.max_delimiter_priority()
- return sum(1 for p in self.delimiters.values() if p == priority)
-
- def maybe_increment_for_loop_variable(self, leaf: Leaf) -> bool:
- """In a for loop, or comprehension, the variables are often unpacks.
-
- To avoid splitting on the comma in this situation, increase the depth of
- tokens between `for` and `in`.
- """
- if leaf.type == token.NAME and leaf.value == "for":
- self.depth += 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_depths
- and self._for_loop_depths[-1] == self.depth
- and leaf.type == token.NAME
- and leaf.value == "in"
- ):
- self.depth -= 1
- self._for_loop_depths.pop()
- return True
-
- return False
-
- def maybe_increment_lambda_arguments(self, leaf: Leaf) -> bool:
- """In a lambda expression, there might be more than one argument.
-
- To avoid splitting on the comma in this situation, increase the depth of
- tokens between `lambda` and `:`.
- """
- if leaf.type == token.NAME and leaf.value == "lambda":
- self.depth += 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_argument_depths
- and self._lambda_argument_depths[-1] == self.depth
- and leaf.type == token.COLON
- ):
- self.depth -= 1
- self._lambda_argument_depths.pop()
- return True
-
- return False
-
- def get_open_lsqb(self) -> Optional[Leaf]:
- """Return the most recent opening square bracket (if any)."""
- return self.bracket_match.get((self.depth - 1, token.RSQB))
-
-
-@dataclass
-class Line:
- """Holds leaves and comments. Can be printed with `str(line)`."""
-
- depth: int = 0
- leaves: List[Leaf] = field(default_factory=list)
- # keys ordered like `leaves`
- comments: Dict[LeafID, List[Leaf]] = field(default_factory=dict)
- bracket_tracker: BracketTracker = field(default_factory=BracketTracker)
- inside_brackets: bool = False
- should_explode: bool = False
-
- def append(self, leaf: Leaf, preformatted: bool = False) -> None:
- """Add a new `leaf` to the end of the line.
-
- Unless `preformatted` is True, the `leaf` will receive a new consistent
- whitespace prefix and metadata applied by :class:`BracketTracker`.
- Trailing commas are maybe removed, unpacked for loop variables are
- demoted from being delimiters.
-
- Inline comments are put aside.
- """
- has_value = leaf.type in BRACKETS or bool(leaf.value.strip())
- if not has_value:
- return
-
- if token.COLON == leaf.type and self.is_class_paren_empty:
- del self.leaves[-2:]
- if self.leaves and not preformatted:
- # Note: at this point leaf.prefix should be empty except for
- # imports, for which we only preserve newlines.
- leaf.prefix += whitespace(
- leaf, complex_subscript=self.is_complex_subscript(leaf)
- )
- if self.inside_brackets or not preformatted:
- self.bracket_tracker.mark(leaf)
- self.maybe_remove_trailing_comma(leaf)
- if not self.append_comment(leaf):
- self.leaves.append(leaf)
-
- def append_safe(self, leaf: Leaf, preformatted: bool = False) -> None:
- """Like :func:`append()` but disallow invalid standalone comment structure.
-
- Raises ValueError when any `leaf` is appended after a standalone comment
- or when a standalone comment is not the first leaf on the line.
- """
- if self.bracket_tracker.depth == 0:
- if self.is_comment:
- raise ValueError("cannot append to standalone comments")
-
- if self.leaves and leaf.type == STANDALONE_COMMENT:
- raise ValueError(
- "cannot append standalone comments to a populated line"
- )
-
- self.append(leaf, preformatted=preformatted)
-
- @property
- def is_comment(self) -> bool:
- """Is this line a standalone comment?"""
- return len(self.leaves) == 1 and self.leaves[0].type == STANDALONE_COMMENT
-
- @property
- def is_decorator(self) -> bool:
- """Is this line a decorator?"""
- return bool(self) and self.leaves[0].type == token.AT
-
- @property
- def is_import(self) -> bool:
- """Is this an import line?"""
- return bool(self) and is_import(self.leaves[0])
-
- @property
- def is_class(self) -> bool:
- """Is this line a class definition?"""
- return (
- bool(self)
- and self.leaves[0].type == token.NAME
- and self.leaves[0].value == "class"
- )
-
- @property
- def is_stub_class(self) -> bool:
- """Is this line a class definition with a body consisting only of "..."?"""
- return self.is_class and self.leaves[-3:] == [
- Leaf(token.DOT, ".") for _ in range(3)
- ]
-
- @property
- def is_collection_with_optional_trailing_comma(self) -> bool:
- """Is this line a collection literal with a trailing comma that's optional?
-
- Note that the trailing comma in a 1-tuple is not optional.
- """
- if not self.leaves or len(self.leaves) < 4:
- return False
-
- # Look for and address a trailing colon.
- if self.leaves[-1].type == token.COLON:
- closer = self.leaves[-2]
- close_index = -2
- else:
- closer = self.leaves[-1]
- close_index = -1
- if closer.type not in CLOSING_BRACKETS or self.inside_brackets:
- return False
-
- if closer.type == token.RPAR:
- # Tuples require an extra check, because if there's only
- # one element in the tuple removing the comma unmakes the
- # tuple.
- #
- # We also check for parens before looking for the trailing
- # comma because in some cases (eg assigning a dict
- # literal) the literal gets wrapped in temporary parens
- # during parsing. This case is covered by the
- # collections.py test data.
- opener = closer.opening_bracket
- for _open_index, leaf in enumerate(self.leaves):
- if leaf is opener:
- break
-
- else:
- # Couldn't find the matching opening paren, play it safe.
- return False
-
- commas = 0
- comma_depth = self.leaves[close_index - 1].bracket_depth
- for leaf in self.leaves[_open_index + 1 : close_index]:
- if leaf.bracket_depth == comma_depth and leaf.type == token.COMMA:
- commas += 1
- if commas > 1:
- # We haven't looked yet for the trailing comma because
- # we might also have caught noop parens.
- return self.leaves[close_index - 1].type == token.COMMA
-
- elif commas == 1:
- return False # it's either a one-tuple or didn't have a trailing comma
-
- if self.leaves[close_index - 1].type in CLOSING_BRACKETS:
- close_index -= 1
- closer = self.leaves[close_index]
- if closer.type == token.RPAR:
- # TODO: this is a gut feeling. Will we ever see this?
- return False
-
- if self.leaves[close_index - 1].type != token.COMMA:
- return False
-
- return True
-
- @property
- def is_def(self) -> bool:
- """Is this a function definition? (Also returns True for async defs.)"""
- try:
- first_leaf = self.leaves[0]
- except IndexError:
- return False
-
- try:
- second_leaf: Optional[Leaf] = self.leaves[1]
- except IndexError:
- second_leaf = None
- return (first_leaf.type == token.NAME and first_leaf.value == "def") or (
- first_leaf.type == token.ASYNC
- and second_leaf is not None
- and second_leaf.type == token.NAME
- and second_leaf.value == "def"
- )
-
- @property
- def is_class_paren_empty(self) -> bool:
- """Is this a class with no base classes but using parentheses?
-
- Those are unnecessary and should be removed.
- """
- return (
- bool(self)
- and len(self.leaves) == 4
- and self.is_class
- and self.leaves[2].type == token.LPAR
- and self.leaves[2].value == "("
- and self.leaves[3].type == token.RPAR
- and self.leaves[3].value == ")"
- )
-
- @property
- def is_triple_quoted_string(self) -> bool:
- """Is the line a triple quoted string?"""
- return (
- bool(self)
- and self.leaves[0].type == token.STRING
- and self.leaves[0].value.startswith(('"""', "'''"))
- )
-
- def contains_standalone_comments(self, depth_limit: int = sys.maxsize) -> bool:
- """If so, needs to be split before emitting."""
- for leaf in self.leaves:
- if leaf.type == STANDALONE_COMMENT and leaf.bracket_depth <= depth_limit:
- return True
-
- return False
-
- def contains_uncollapsable_type_comments(self) -> bool:
- ignored_ids = set()
- try:
- last_leaf = self.leaves[-1]
- ignored_ids.add(id(last_leaf))
- if last_leaf.type == token.COMMA or (
- last_leaf.type == token.RPAR and not last_leaf.value
- ):
- # When trailing commas or optional parens are inserted by Black for
- # consistency, comments after the previous last element are not moved
- # (they don't have to, rendering will still be correct). So we ignore
- # trailing commas and invisible.
- last_leaf = self.leaves[-2]
- ignored_ids.add(id(last_leaf))
- except IndexError:
- return False
-
- # A type comment is uncollapsable if it is attached to a leaf
- # that isn't at the end of the line (since that could cause it
- # to get associated to a different argument) or if there are
- # comments before it (since that could cause it to get hidden
- # behind a comment.
- comment_seen = False
- for leaf_id, comments in self.comments.items():
- for comment in comments:
- if is_type_comment(comment):
- if comment_seen or (
- not is_type_comment(comment, " ignore")
- and leaf_id not in ignored_ids
- ):
- return True
-
- comment_seen = True
-
- return False
-
- def contains_unsplittable_type_ignore(self) -> bool:
- if not self.leaves:
- return False
-
- # If a 'type: ignore' is attached to the end of a line, we
- # can't split the line, because we can't know which of the
- # subexpressions the ignore was meant to apply to.
- #
- # We only want this to apply to actual physical lines from the
- # original source, though: we don't want the presence of a
- # 'type: ignore' at the end of a multiline expression to
- # justify pushing it all onto one line. Thus we
- # (unfortunately) need to check the actual source lines and
- # only report an unsplittable 'type: ignore' if this line was
- # one line in the original code.
-
- # Grab the first and last line numbers, skipping generated leaves
- first_line = next((l.lineno for l in self.leaves if l.lineno != 0), 0)
- last_line = next((l.lineno for l in reversed(self.leaves) if l.lineno != 0), 0)
-
- if first_line == last_line:
- # We look at the last two leaves since a comma or an
- # invisible paren could have been added at the end of the
- # line.
- for node in self.leaves[-2:]:
- for comment in self.comments.get(id(node), []):
- if is_type_comment(comment, " ignore"):
- return True
-
- return False
-
- def contains_multiline_strings(self) -> bool:
- return any(is_multiline_string(leaf) for leaf in self.leaves)
-
- def maybe_remove_trailing_comma(self, closing: Leaf) -> bool:
- """Remove trailing comma if there is one and it's safe."""
- if not (self.leaves and self.leaves[-1].type == token.COMMA):
- return False
-
- # We remove trailing commas only in the case of importing a
- # single name from a module.
- if not (
- self.leaves
- and self.is_import
- and len(self.leaves) > 4
- and self.leaves[-1].type == token.COMMA
- and closing.type in CLOSING_BRACKETS
- and self.leaves[-4].type == token.NAME
- and (
- # regular `from foo import bar,`
- self.leaves[-4].value == "import"
- # `from foo import (bar as baz,)
- or (
- len(self.leaves) > 6
- and self.leaves[-6].value == "import"
- and self.leaves[-3].value == "as"
- )
- # `from foo import bar as baz,`
- or (
- len(self.leaves) > 5
- and self.leaves[-5].value == "import"
- and self.leaves[-3].value == "as"
- )
- )
- and closing.type == token.RPAR
- ):
- return False
-
- self.remove_trailing_comma()
- return True
-
- def append_comment(self, comment: Leaf) -> bool:
- """Add an inline or standalone comment to the line."""
- if (
- comment.type == STANDALONE_COMMENT
- and self.bracket_tracker.any_open_brackets()
- ):
- comment.prefix = ""
- return False
-
- if comment.type != token.COMMENT:
- return False
-
- if not self.leaves:
- comment.type = STANDALONE_COMMENT
- comment.prefix = ""
- return False
-
- last_leaf = self.leaves[-1]
- if (
- last_leaf.type == token.RPAR
- and not last_leaf.value
- and last_leaf.parent
- and len(list(last_leaf.parent.leaves())) <= 3
- and not is_type_comment(comment)
- ):
- # Comments on an optional parens wrapping a single leaf should belong to
- # the wrapped node except if it's a type comment. Pinning the comment like
- # this avoids unstable formatting caused by comment migration.
- if len(self.leaves) < 2:
- comment.type = STANDALONE_COMMENT
- comment.prefix = ""
- return False
-
- last_leaf = self.leaves[-2]
- self.comments.setdefault(id(last_leaf), []).append(comment)
- return True
-
- def comments_after(self, leaf: Leaf) -> List[Leaf]:
- """Generate comments that should appear directly after `leaf`."""
- return self.comments.get(id(leaf), [])
-
- def remove_trailing_comma(self) -> None:
- """Remove the trailing comma and moves the comments attached to it."""
- trailing_comma = self.leaves.pop()
- trailing_comma_comments = self.comments.pop(id(trailing_comma), [])
- self.comments.setdefault(id(self.leaves[-1]), []).extend(
- trailing_comma_comments
- )
-
- def is_complex_subscript(self, leaf: Leaf) -> bool:
- """Return True iff `leaf` is part of a slice with non-trivial exprs."""
- open_lsqb = self.bracket_tracker.get_open_lsqb()
- if open_lsqb is None:
- return False
-
- subscript_start = open_lsqb.next_sibling
-
- if isinstance(subscript_start, Node):
- if subscript_start.type == syms.listmaker:
- return False
-
- if subscript_start.type == syms.subscriptlist:
- subscript_start = child_towards(subscript_start, leaf)
- return subscript_start is not None and any(
- n.type in TEST_DESCENDANTS for n in subscript_start.pre_order()
- )
-
- def clone(self) -> "Line":
- return Line(
- depth=self.depth,
- inside_brackets=self.inside_brackets,
- should_explode=self.should_explode,
- )
-
- def __str__(self) -> str:
- """Render the line."""
- if not self:
- return "\n"
-
- indent = " " * self.depth
- leaves = iter(self.leaves)
- first = next(leaves)
- res = f"{first.prefix}{indent}{first.value}"
- for leaf in leaves:
- res += str(leaf)
- for comment in itertools.chain.from_iterable(self.comments.values()):
- res += str(comment)
-
- return res + "\n"
-
- def __bool__(self) -> bool:
- """Return True if the line has leaves or comments."""
- return bool(self.leaves or self.comments)
-
-
-@dataclass
-class EmptyLineTracker:
- """Provides a stateful method that returns the number of potential extra
- empty lines needed before and after the currently processed line.
-
- Note: this tracker works on lines that haven't been split yet. It assumes
- the prefix of the first leaf consists of optional newlines. Those newlines
- are consumed by `maybe_empty_lines()` and included in the computation.
- """
-
- is_pyi: bool = False
- previous_line: Optional[Line] = None
- previous_after: int = 0
- previous_defs: List[int] = field(default_factory=list)
-
- def maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
- """Return the number of extra empty lines before and after the `current_line`.
-
- This is for separating `def`, `async def` and `class` with extra empty
- lines (two on module-level).
- """
- before, after = self._maybe_empty_lines(current_line)
- before = (
- # Black should not insert empty lines at the beginning
- # of the file
- 0
- if self.previous_line is None
- else before - self.previous_after
- )
- self.previous_after = after
- self.previous_line = current_line
- return before, after
-
- def _maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
- max_allowed = 1
- if current_line.depth == 0:
- max_allowed = 1 if self.is_pyi else 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_allowed)
- first_leaf.prefix = ""
- else:
- before = 0
- depth = current_line.depth
- while self.previous_defs and self.previous_defs[-1] >= depth:
- self.previous_defs.pop()
- if self.is_pyi:
- before = 0 if depth else 1
- else:
- before = 1 if depth else 2
- if current_line.is_decorator or current_line.is_def or current_line.is_class:
- return self._maybe_empty_lines_for_class_or_def(current_line, before)
-
- if (
- self.previous_line
- and self.previous_line.is_import
- and not current_line.is_import
- and depth == self.previous_line.depth
- ):
- return (before or 1), 0
-
- if (
- self.previous_line
- and self.previous_line.is_class
- and current_line.is_triple_quoted_string
- ):
- return before, 1
-
- return before, 0
-
- def _maybe_empty_lines_for_class_or_def(
- self, current_line: Line, before: int
- ) -> Tuple[int, int]:
- if not current_line.is_decorator:
- self.previous_defs.append(current_line.depth)
- if self.previous_line is None:
- # Don't insert empty lines before the first line in the file.
- return 0, 0
-
- if self.previous_line.is_decorator:
- return 0, 0
-
- if self.previous_line.depth < current_line.depth and (
- self.previous_line.is_class or self.previous_line.is_def
- ):
- return 0, 0
-
- if (
- self.previous_line.is_comment
- and self.previous_line.depth == current_line.depth
- and before == 0
- ):
- return 0, 0
-
- if self.is_pyi:
- if self.previous_line.depth > current_line.depth:
- newlines = 1
- elif current_line.is_class or self.previous_line.is_class:
- if current_line.is_stub_class and self.previous_line.is_stub_class:
- # No blank line between classes with an empty body
- newlines = 0
- else:
- newlines = 1
- elif current_line.is_def and not self.previous_line.is_def:
- # Blank line between a block of functions and a block of non-functions
- newlines = 1
- else:
- newlines = 0
- else:
- newlines = 2
- if current_line.depth and newlines:
- newlines -= 1
- return newlines, 0
-
-
-@dataclass
-class LineGenerator(Visitor[Line]):
- """Generates reformatted Line objects. Empty lines are not emitted.
-
- Note: destroys the tree it's visiting by mutating prefixes of its leaves
- in ways that will no longer stringify to valid Python code on the tree.
- """
-
- is_pyi: bool = False
- normalize_strings: bool = True
- current_line: Line = field(default_factory=Line)
- remove_u_prefix: bool = False
-
- def line(self, indent: int = 0) -> Iterator[Line]:
- """Generate a line.
-
- If the line is empty, only emit if it makes sense.
- If the line is too long, split it first and then generate.
-
- If any lines were generated, set up a new current_line.
- """
- if not self.current_line:
- 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)
- yield complete_line
-
- def visit_default(self, node: LN) -> Iterator[Line]:
- """Default `visit_*()` implementation. Recurses to children of `node`."""
- 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 self.normalize_strings and node.type == token.STRING:
- normalize_string_prefix(node, remove_u_prefix=self.remove_u_prefix)
- normalize_string_quotes(node)
- if node.type == token.NUMBER:
- normalize_numeric_literal(node)
- if node.type not in WHITESPACE:
- self.current_line.append(node)
- yield from super().visit_default(node)
-
- def visit_INDENT(self, node: Leaf) -> Iterator[Line]:
- """Increase indentation level, maybe yield a line."""
- # In blib2to3 INDENT never holds comments.
- yield from self.line(+1)
- yield from self.visit_default(node)
-
- def visit_DEDENT(self, node: Leaf) -> Iterator[Line]:
- """Decrease indentation level, maybe yield a line."""
- # The current line might still wait for trailing comments. At DEDENT time
- # there won't be any (they would be prefixes on the preceding NEWLINE).
- # Emit the line then.
- yield from self.line()
-
- # While DEDENT has no value, its prefix may contain standalone comments
- # that belong to the current indentation level. Get 'em.
- yield from self.visit_default(node)
-
- # Finally, emit the dedent.
- yield from self.line(-1)
-
- def visit_stmt(
- self, node: Node, keywords: Set[str], parens: Set[str]
- ) -> Iterator[Line]:
- """Visit a statement.
-
- This implementation is shared for `if`, `while`, `for`, `try`, `except`,
- `def`, `with`, `class`, `assert` and assignments.
-
- The relevant Python language `keywords` for a given statement will be
- NAME leaves within it. This methods puts those on a separate line.
-
- `parens` holds a set of string leaf values immediately after which
- invisible parens should be put.
- """
- normalize_invisible_parens(node, parens_after=parens)
- for child in node.children:
- if child.type == token.NAME and child.value in keywords: # type: ignore
- yield from self.line()
-
- yield from self.visit(child)
-
- def visit_suite(self, node: Node) -> Iterator[Line]:
- """Visit a suite."""
- if self.is_pyi and is_stub_suite(node):
- yield from self.visit(node.children[2])
- else:
- yield from self.visit_default(node)
-
- def visit_simple_stmt(self, node: Node) -> Iterator[Line]:
- """Visit a statement without nested statements."""
- is_suite_like = node.parent and node.parent.type in STATEMENT
- if is_suite_like:
- if self.is_pyi 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.is_pyi or not node.parent or not is_stub_suite(node.parent):
- yield from self.line()
- yield from self.visit_default(node)
-
- def visit_async_stmt(self, node: Node) -> Iterator[Line]:
- """Visit `async def`, `async for`, `async with`."""
- yield from self.line()
-
- children = iter(node.children)
- for child in children:
- yield from self.visit(child)
-
- if child.type == token.ASYNC:
- break
-
- internal_stmt = next(children)
- for child in internal_stmt.children:
- yield from self.visit(child)
-
- def visit_decorators(self, node: Node) -> Iterator[Line]:
- """Visit decorators."""
- for child in node.children:
- yield from self.line()
- yield from self.visit(child)
-
- def visit_SEMI(self, leaf: Leaf) -> Iterator[Line]:
- """Remove a semicolon and put the other statement on a separate line."""
- yield from self.line()
-
- def visit_ENDMARKER(self, leaf: Leaf) -> Iterator[Line]:
- """End of file. Process outstanding comments and end with a newline."""
- yield from self.visit_default(leaf)
- yield from self.line()
-
- def visit_STANDALONE_COMMENT(self, leaf: Leaf) -> Iterator[Line]:
- if not self.current_line.bracket_tracker.any_open_brackets():
- yield from self.line()
- yield from self.visit_default(leaf)
-
- def visit_factor(self, node: Node) -> Iterator[Line]:
- """Force parentheses between a unary op and a binary power:
-
- -2 ** 8 -> -(2 ** 8)
- """
- _operator, operand = node.children
- if (
- operand.type == syms.power
- and len(operand.children) == 3
- and operand.children[1].type == token.DOUBLESTAR
- ):
- lpar = Leaf(token.LPAR, "(")
- rpar = Leaf(token.RPAR, ")")
- index = operand.remove() or 0
- node.insert_child(index, Node(syms.atom, [lpar, operand, rpar]))
- yield from self.visit_default(node)
-
- def visit_STRING(self, leaf: Leaf) -> Iterator[Line]:
- # Check if it's a docstring
- if prev_siblings_are(
- leaf.parent, [None, token.NEWLINE, token.INDENT, syms.simple_stmt]
- ) and is_multiline_string(leaf):
- prefix = " " * self.current_line.depth
- docstring = fix_docstring(leaf.value[3:-3], prefix)
- leaf.value = leaf.value[0:3] + docstring + leaf.value[-3:]
- normalize_string_quotes(leaf)
-
- yield from self.visit_default(leaf)
-
- def __post_init__(self) -> None:
- """You are in a twisty little maze of passages."""
- v = self.visit_stmt
- Ø: Set[str] = set()
- self.visit_assert_stmt = partial(v, keywords={"assert"}, parens={"assert", ","})
- self.visit_if_stmt = partial(
- v, keywords={"if", "else", "elif"}, parens={"if", "elif"}
- )
- self.visit_while_stmt = partial(v, keywords={"while", "else"}, parens={"while"})
- self.visit_for_stmt = partial(v, keywords={"for", "else"}, parens={"for", "in"})
- self.visit_try_stmt = partial(
- v, keywords={"try", "except", "else", "finally"}, parens=Ø
- )
- self.visit_except_clause = partial(v, keywords={"except"}, parens=Ø)
- self.visit_with_stmt = partial(v, keywords={"with"}, parens=Ø)
- self.visit_funcdef = partial(v, keywords={"def"}, parens=Ø)
- self.visit_classdef = partial(v, keywords={"class"}, parens=Ø)
- 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"})
- self.visit_async_funcdef = self.visit_async_stmt
- self.visit_decorated = self.visit_decorators
-
-
-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}
-
-
-def whitespace(leaf: Leaf, *, complex_subscript: bool) -> 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 = " "
- t = leaf.type
- p = leaf.parent
- v = leaf.value
- if t in ALWAYS_NO_SPACE:
- return NO
-
- if t == token.COMMENT:
- return DOUBLESPACE
-
- assert p is not None, f"INTERNAL ERROR: hand-made leaf without parent: {leaf!r}"
- if t == token.COLON and p.type not in {
- syms.subscript,
- syms.subscriptlist,
- syms.sliceop,
- }:
- return NO
-
- prev = leaf.prev_sibling
- if not prev:
- prevp = preceding_leaf(p)
- if not prevp or prevp.type in OPENING_BRACKETS:
- return NO
-
- if t == token.COLON:
- if prevp.type == token.COLON:
- return NO
-
- elif prevp.type != token.COMMA and not complex_subscript:
- return NO
-
- return SPACE
-
- if prevp.type == token.EQUAL:
- 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 in VARARGS_SPECIALS:
- if is_vararg(prevp, within=VARARGS_PARENTS | UNPACKING_PARENTS):
- return NO
-
- elif prevp.type == token.COLON:
- if prevp.parent and prevp.parent.type in {syms.subscript, syms.sliceop}:
- return SPACE if complex_subscript else NO
-
- elif (
- prevp.parent
- and prevp.parent.type == syms.factor
- and prevp.type in MATH_OPERATORS
- ):
- 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 prev.type in OPENING_BRACKETS:
- return NO
-
- if p.type in {syms.parameters, syms.arglist}:
- # untyped function signatures or calls
- if not prev or prev.type != token.COMMA:
- return NO
-
- elif p.type == syms.varargslist:
- # lambdas
- if prev and prev.type != token.COMMA:
- return NO
-
- elif p.type == syms.typedargslist:
- # typed function signatures
- if not prev:
- return NO
-
- if t == token.EQUAL:
- if prev.type != syms.tname:
- return NO
-
- elif prev.type == token.EQUAL:
- # 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 prev.prefix
-
- elif prev.type != token.COMMA:
- return NO
-
- elif p.type == syms.tname:
- # type names
- if not prev:
- prevp = preceding_leaf(p)
- if not prevp or prevp.type != token.COMMA:
- return NO
-
- elif p.type == syms.trailer:
- # attributes and calls
- if t == token.LPAR or t == token.RPAR:
- 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:
- return NO
-
- elif prev.type != token.COMMA:
- return NO
-
- elif p.type == syms.argument:
- # single argument
- if t == token.EQUAL:
- return NO
-
- if not prev:
- prevp = preceding_leaf(p)
- if not prevp or prevp.type == token.LPAR:
- return NO
-
- elif prev.type in {token.EQUAL} | VARARGS_SPECIALS:
- return NO
-
- elif p.type == syms.decorator:
- # decorators
- return NO
-
- elif p.type == syms.dotted_name:
- if prev:
- return NO
-
- prevp = preceding_leaf(p)
- if not prevp or prevp.type == token.AT or prevp.type == token.DOT:
- return NO
-
- elif p.type == syms.classdef:
- if t == token.LPAR:
- return NO
-
- if prev and prev.type == token.LPAR:
- return NO
-
- elif p.type in {syms.subscript, syms.sliceop}:
- # indexing
- if not prev:
- assert p.parent is not None, "subscripts are always parented"
- if p.parent.type == syms.subscriptlist:
- return SPACE
-
- return NO
-
- elif not complex_subscript:
- return NO
-
- elif p.type == syms.atom:
- if prev and t == token.DOT:
- # dots, but not the first one.
- return NO
-
- elif p.type == syms.dictsetmaker:
- # dict unpacking
- if prev and prev.type == token.DOUBLESTAR:
- return NO
-
- elif p.type in {syms.factor, syms.star_expr}:
- # unary ops
- if not prev:
- prevp = preceding_leaf(p)
- if not prevp or prevp.type in OPENING_BRACKETS:
- return NO
-
- prevp_parent = prevp.parent
- assert prevp_parent is not None
- if prevp.type == token.COLON and prevp_parent.type in {
- syms.subscript,
- syms.sliceop,
- }:
- return NO
-
- elif prevp.type == token.EQUAL and prevp_parent.type == syms.argument:
- return NO
-
- elif t in {token.NAME, token.NUMBER, token.STRING}:
- return NO
-
- elif p.type == syms.import_from:
- if t == token.DOT:
- if prev and prev.type == token.DOT:
- return NO
-
- elif t == token.NAME:
- if v == "import":
- return SPACE
-
- if prev and prev.type == token.DOT:
- return NO
-
- elif p.type == syms.sliceop:
- return NO
-
- return SPACE
-
-
-def preceding_leaf(node: Optional[LN]) -> Optional[Leaf]:
- """Return the first leaf that precedes `node`, if any."""
- while node:
- res = node.prev_sibling
- if res:
- if isinstance(res, Leaf):
- return res
-
- try:
- return list(res.leaves())[-1]
-
- except IndexError:
- return None
-
- node = node.parent
- return None
-
-
-def prev_siblings_are(node: Optional[LN], tokens: List[Optional[NodeType]]) -> bool:
- """Return if the `node` and its previous siblings match types against the provided
- list of tokens; the provided `node`has its type matched against the last element in
- the list. `None` can be used as the first element to declare that the start of the
- list is anchored at the start of its parent's children."""
- if not tokens:
- return True
- if tokens[-1] is None:
- return node is None
- if not node:
- return False
- if node.type != tokens[-1]:
- return False
- return prev_siblings_are(node.prev_sibling, tokens[:-1])
-
-
-def child_towards(ancestor: Node, descendant: LN) -> Optional[LN]:
- """Return the child of `ancestor` that contains `descendant`."""
- node: Optional[LN] = descendant
- while node and node.parent != ancestor:
- node = node.parent
- return node
-
-
-def container_of(leaf: Leaf) -> LN:
- """Return `leaf` or one of its ancestors that is the topmost container of it.
-
- By "container" we mean a node where `leaf` is the very first child.
- """
- same_prefix = leaf.prefix
- container: LN = leaf
- while container:
- parent = container.parent
- if parent is None:
- break
-
- if parent.children[0].prefix != same_prefix:
- break
-
- if parent.type == syms.file_input:
- break
-
- if parent.prev_sibling is not None and parent.prev_sibling.type in BRACKETS:
- break
-
- container = parent
- return container
-
-
-def is_split_after_delimiter(leaf: Leaf, previous: Optional[Leaf] = None) -> Priority:
- """Return the priority of the `leaf` delimiter, given a line break after it.
-
- The delimiter priorities returned here are from those delimiters that would
- cause a line break after themselves.
-
- Higher numbers are higher priority.
- """
- if leaf.type == token.COMMA:
- return COMMA_PRIORITY
-
- return 0
-
-
-def is_split_before_delimiter(leaf: Leaf, previous: Optional[Leaf] = None) -> Priority:
- """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.
-
- Higher numbers are higher priority.
- """
- if is_vararg(leaf, within=VARARGS_PARENTS | UNPACKING_PARENTS):
- # * and ** might also be MATH_OPERATORS but in this case they are not.
- # Don't treat them as a delimiter.
- return 0
-
- if (
- leaf.type == token.DOT
- and leaf.parent
- and leaf.parent.type not in {syms.import_from, syms.dotted_name}
- and (previous is None or previous.type in CLOSING_BRACKETS)
- ):
- return DOT_PRIORITY
-
- if (
- leaf.type in MATH_OPERATORS
- and leaf.parent
- and leaf.parent.type not in {syms.factor, syms.star_expr}
- ):
- return MATH_PRIORITIES[leaf.type]
-
- if leaf.type in COMPARATORS:
- return COMPARATOR_PRIORITY
-
- if (
- leaf.type == token.STRING
- and previous is not None
- and previous.type == token.STRING
- ):
- return STRING_PRIORITY
-
- 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
- ):
- if (
- not isinstance(leaf.prev_sibling, Leaf)
- or leaf.prev_sibling.value != "async"
- ):
- return COMPREHENSION_PRIORITY
-
- if (
- leaf.value == "if"
- and leaf.parent
- and leaf.parent.type in {syms.comp_if, syms.old_comp_if}
- ):
- return COMPREHENSION_PRIORITY
-
- if leaf.value in {"if", "else"} and leaf.parent and leaf.parent.type == syms.test:
- return TERNARY_PRIORITY
-
- if leaf.value == "is":
- return COMPARATOR_PRIORITY
-
- if (
- leaf.value == "in"
- and leaf.parent
- and leaf.parent.type in {syms.comp_op, syms.comparison}
- and not (
- previous is not None
- and previous.type == token.NAME
- and previous.value == "not"
- )
- ):
- return COMPARATOR_PRIORITY
-
- if (
- leaf.value == "not"
- and leaf.parent
- and leaf.parent.type == syms.comp_op
- and not (
- previous is not None
- and previous.type == token.NAME
- and previous.value == "is"
- )
- ):
- return COMPARATOR_PRIORITY
-
- if leaf.value in LOGIC_OPERATORS and leaf.parent:
- return LOGIC_PRIORITY
-
- return 0
-
-
-FMT_OFF = {"# fmt: off", "# fmt:off", "# yapf: disable"}
-FMT_ON = {"# fmt: on", "# fmt:on", "# yapf: enable"}
-
-
-def generate_comments(leaf: LN) -> Iterator[Leaf]:
- """Clean the prefix of the `leaf` and generate comments from it, if any.
-
- Comments in lib2to3 are shoved into the whitespace prefix. This happens
- in `pgen2/driver.py:Driver.parse_tokens()`. This was a brilliant implementation
- move because it does away with modifying the grammar to include all the
- possible places in which comments can be placed.
-
- The sad consequence for us though is that comments don't "belong" anywhere.
- This is why this function generates simple parentless Leaf objects for
- comments. We simply don't know what the correct parent should be.
-
- No matter though, we can live without this. We really only need to
- differentiate between inline and standalone comments. The latter don't
- share the line with any code.
-
- Inline comments are emitted as regular token.COMMENT leaves. Standalone
- are emitted with a fake STANDALONE_COMMENT token identifier.
- """
- for pc in list_comments(leaf.prefix, is_endmarker=leaf.type == token.ENDMARKER):
- yield Leaf(pc.type, pc.value, prefix="\n" * pc.newlines)
-
-
-@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
- consumed: int # how many characters of the original leaf's prefix did we consume
-
-
-@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
-
- consumed = 0
- nlines = 0
- ignored_lines = 0
- for index, line in enumerate(prefix.split("\n")):
- consumed += len(line) + 1 # adding the length of the split '\n'
- line = line.lstrip()
- if not line:
- nlines += 1
- if not line.startswith("#"):
- # Escaped newlines outside of a comment are not really newlines at
- # all. We treat a single-line comment following an escaped newline
- # as a simple trailing comment.
- if line.endswith("\\"):
- ignored_lines += 1
- continue
-
- if index == ignored_lines and not is_endmarker:
- comment_type = token.COMMENT # simple trailing comment
- else:
- comment_type = STANDALONE_COMMENT
- comment = make_comment(line)
- result.append(
- ProtoComment(
- type=comment_type, value=comment, newlines=nlines, consumed=consumed
- )
- )
- nlines = 0
- return result
-
-
-def make_comment(content: str) -> str:
- """Return a consistently formatted comment from the given `content` string.
-
- All comments (except for "##", "#!", "#:", '#'", "#%%") should have a single
- space between the hash sign and the content.
-
- If `content` didn't start with a hash sign, one is provided.
- """
- content = content.rstrip()
- if not content:
- return "#"
-
- if content[0] == "#":
- content = content[1:]
- if content and content[0] not in " !:#'%":
- content = " " + content
- return "#" + content
-
-
-def transform_line(
- line: Line,
- line_length: int,
- normalize_strings: bool,
- features: Collection[Feature] = (),
-) -> Iterator[Line]:
- """Transform a `line`, potentially splitting it into many lines.
-
- They should fit in the allotted `line_length` but might not be able to.
-
- `features` are syntactical features that may be used in the output.
- """
- if line.is_comment:
- yield line
- return
-
- line_str = line_to_string(line)
-
- def init_st(ST: Type[StringTransformer]) -> StringTransformer:
- """Initialize StringTransformer"""
- return ST(line_length, normalize_strings)
-
- string_merge = init_st(StringMerger)
- string_paren_strip = init_st(StringParenStripper)
- string_split = init_st(StringSplitter)
- string_paren_wrap = init_st(StringParenWrapper)
-
- transformers: List[Transformer]
- if (
- not line.contains_uncollapsable_type_comments()
- and not line.should_explode
- and not line.is_collection_with_optional_trailing_comma
- and (
- is_line_short_enough(line, line_length=line_length, line_str=line_str)
- or line.contains_unsplittable_type_ignore()
- )
- and not (line.contains_standalone_comments() and line.inside_brackets)
- ):
- # Only apply basic string preprocessing, since lines shouldn't be split here.
- transformers = [string_merge, string_paren_strip]
- elif line.is_def:
- transformers = [left_hand_split]
- else:
-
- def rhs(line: Line, features: Collection[Feature]) -> Iterator[Line]:
- for omit in generate_trailers_to_omit(line, line_length):
- lines = list(right_hand_split(line, line_length, features, omit=omit))
- if is_line_short_enough(lines[0], line_length=line_length):
- yield from lines
- return
-
- # All splits failed, best effort split with no omits.
- # This mostly happens to multiline strings that are by definition
- # reported as not fitting a single line.
- # line_length=1 here was historically a bug that somehow became a feature.
- # See #762 and #781 for the full story.
- yield from right_hand_split(line, line_length=1, features=features)
-
- if line.inside_brackets:
- transformers = [
- string_merge,
- string_paren_strip,
- delimiter_split,
- standalone_comment_split,
- string_split,
- string_paren_wrap,
- rhs,
- ]
- else:
- transformers = [
- string_merge,
- string_paren_strip,
- string_split,
- string_paren_wrap,
- rhs,
- ]
-
- for transform in transformers:
- # We are accumulating lines in `result` because we might want to abort
- # mission and return the original line in the end, or attempt a different
- # split altogether.
- result: List[Line] = []
- try:
- for l in transform(line, features):
- if str(l).strip("\n") == line_str:
- raise CannotTransform(
- "Line transformer returned an unchanged result"
- )
-
- result.extend(
- transform_line(
- l,
- line_length=line_length,
- normalize_strings=normalize_strings,
- features=features,
- )
- )
- except CannotTransform:
- continue
- else:
- yield from result
- break
-
- else:
- yield line
-
-
-@dataclass # type: ignore
-class StringTransformer(ABC):
- """
- An implementation of the Transformer protocol that relies on its
- subclasses overriding the template methods `do_match(...)` and
- `do_transform(...)`.
-
- This Transformer works exclusively on strings (for example, by merging
- or splitting them).
-
- The following sections can be found among the docstrings of each concrete
- StringTransformer subclass.
-
- Requirements:
- Which requirements must be met of the given Line for this
- StringTransformer to be applied?
-
- Transformations:
- If the given Line meets all of the above requirments, which string
- transformations can you expect to be applied to it by this
- StringTransformer?
-
- Collaborations:
- What contractual agreements does this StringTransformer have with other
- StringTransfomers? Such collaborations should be eliminated/minimized
- as much as possible.
- """
-
- line_length: int
- normalize_strings: bool
-
- @abstractmethod
- def do_match(self, line: Line) -> TMatchResult:
- """
- Returns:
- * Ok(string_idx) such that `line.leaves[string_idx]` is our target
- string, if a match was able to be made.
- OR
- * Err(CannotTransform), if a match was not able to be made.
- """
-
- @abstractmethod
- def do_transform(self, line: Line, string_idx: int) -> Iterator[TResult[Line]]:
- """
- Yields:
- * Ok(new_line) where new_line is the new transformed line.
- OR
- * Err(CannotTransform) if the transformation failed for some reason. The
- `do_match(...)` template method should usually be used to reject
- the form of the given Line, but in some cases it is difficult to
- know whether or not a Line meets the StringTransformer's
- requirements until the transformation is already midway.
-
- Side Effects:
- This method should NOT mutate @line directly, but it MAY mutate the
- Line's underlying Node structure. (WARNING: If the underlying Node
- structure IS altered, then this method should NOT be allowed to
- yield an CannotTransform after that point.)
- """
-
- def __call__(self, line: Line, _features: Collection[Feature]) -> Iterator[Line]:
- """
- StringTransformer instances have a call signature that mirrors that of
- the Transformer type.
-
- Raises:
- CannotTransform(...) if the concrete StringTransformer class is unable
- to transform @line.
- """
- # Optimization to avoid calling `self.do_match(...)` when the line does
- # not contain any string.
- if not any(leaf.type == token.STRING for leaf in line.leaves):
- raise CannotTransform("There are no strings in this line.")
-
- match_result = self.do_match(line)
-
- if isinstance(match_result, Err):
- cant_transform = match_result.err()
- raise CannotTransform(
- f"The string transformer {self.__class__.__name__} does not recognize"
- " this line as one that it can transform."
- ) from cant_transform
-
- string_idx = match_result.ok()
-
- for line_result in self.do_transform(line, string_idx):
- if isinstance(line_result, Err):
- cant_transform = line_result.err()
- raise CannotTransform(
- "StringTransformer failed while attempting to transform string."
- ) from cant_transform
- line = line_result.ok()
- yield line
-
-
-@dataclass
-class CustomSplit:
- """A custom (i.e. manual) string split.
-
- A single CustomSplit instance represents a single substring.
-
- Examples:
- Consider the following string:
- ```
- "Hi there friend."
- " This is a custom"
- f" string {split}."
- ```
-
- This string will correspond to the following three CustomSplit instances:
- ```
- CustomSplit(False, 16)
- CustomSplit(False, 17)
- CustomSplit(True, 16)
- ```
- """
-
- has_prefix: bool
- break_idx: int
-
-
-class CustomSplitMapMixin:
- """
- This mixin class is used to map merged strings to a sequence of
- CustomSplits, which will then be used to re-split the strings iff none of
- the resultant substrings go over the configured max line length.
- """
-
- _Key = Tuple[StringID, str]
- _CUSTOM_SPLIT_MAP: Dict[_Key, Tuple[CustomSplit, ...]] = defaultdict(tuple)
-
- @staticmethod
- def _get_key(string: str) -> "CustomSplitMapMixin._Key":
- """
- Returns:
- A unique identifier that is used internally to map @string to a
- group of custom splits.
- """
- return (id(string), string)
-
- def add_custom_splits(
- self, string: str, custom_splits: Iterable[CustomSplit]
- ) -> None:
- """Custom Split Map Setter Method
-
- Side Effects:
- Adds a mapping from @string to the custom splits @custom_splits.
- """
- key = self._get_key(string)
- self._CUSTOM_SPLIT_MAP[key] = tuple(custom_splits)
-
- def pop_custom_splits(self, string: str) -> List[CustomSplit]:
- """Custom Split Map Getter Method
-
- Returns:
- * A list of the custom splits that are mapped to @string, if any
- exist.
- OR
- * [], otherwise.
-
- Side Effects:
- Deletes the mapping between @string and its associated custom
- splits (which are returned to the caller).
- """
- key = self._get_key(string)
-
- custom_splits = self._CUSTOM_SPLIT_MAP[key]
- del self._CUSTOM_SPLIT_MAP[key]
-
- return list(custom_splits)
-
- def has_custom_splits(self, string: str) -> bool:
- """
- Returns:
- True iff @string is associated with a set of custom splits.
- """
- key = self._get_key(string)
- return key in self._CUSTOM_SPLIT_MAP
-
-
-class StringMerger(CustomSplitMapMixin, StringTransformer):
- """StringTransformer that merges strings together.
-
- Requirements:
- (A) The line contains adjacent strings such that at most one substring
- has inline comments AND none of those inline comments are pragmas AND
- the set of all substring prefixes is either of length 1 or equal to
- {"", "f"} AND none of the substrings are raw strings (i.e. are prefixed
- with 'r').
- OR
- (B) The line contains a string which uses line continuation backslashes.
-
- Transformations:
- Depending on which of the two requirements above where met, either:
-
- (A) The string group associated with the target string is merged.
- OR
- (B) All line-continuation backslashes are removed from the target string.
-
- Collaborations:
- StringMerger provides custom split information to StringSplitter.
- """
-
- def do_match(self, line: Line) -> TMatchResult:
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
-
- for (i, leaf) in enumerate(LL):
- if (
- leaf.type == token.STRING
- and is_valid_index(i + 1)
- and LL[i + 1].type == token.STRING
- ):
- return Ok(i)
-
- if leaf.type == token.STRING and "\\\n" in leaf.value:
- return Ok(i)
-
- return TErr("This line has no strings that need merging.")
-
- def do_transform(self, line: Line, string_idx: int) -> Iterator[TResult[Line]]:
- new_line = line
- rblc_result = self.__remove_backslash_line_continuation_chars(
- new_line, string_idx
- )
- if isinstance(rblc_result, Ok):
- new_line = rblc_result.ok()
-
- msg_result = self.__merge_string_group(new_line, string_idx)
- if isinstance(msg_result, Ok):
- new_line = msg_result.ok()
-
- if isinstance(rblc_result, Err) and isinstance(msg_result, Err):
- msg_cant_transform = msg_result.err()
- rblc_cant_transform = rblc_result.err()
- cant_transform = CannotTransform(
- "StringMerger failed to merge any strings in this line."
- )
-
- # Chain the errors together using `__cause__`.
- msg_cant_transform.__cause__ = rblc_cant_transform
- cant_transform.__cause__ = msg_cant_transform
-
- yield Err(cant_transform)
- else:
- yield Ok(new_line)
-
- @staticmethod
- def __remove_backslash_line_continuation_chars(
- line: Line, string_idx: int
- ) -> TResult[Line]:
- """
- Merge strings that were split across multiple lines using
- line-continuation backslashes.
-
- Returns:
- Ok(new_line), if @line contains backslash line-continuation
- characters.
- OR
- Err(CannotTransform), otherwise.
- """
- LL = line.leaves
-
- string_leaf = LL[string_idx]
- if not (
- string_leaf.type == token.STRING
- and "\\\n" in string_leaf.value
- and not has_triple_quotes(string_leaf.value)
- ):
- return TErr(
- f"String leaf {string_leaf} does not contain any backslash line"
- " continuation characters."
- )
-
- new_line = line.clone()
- new_line.comments = line.comments
- append_leaves(new_line, line, LL)
-
- new_string_leaf = new_line.leaves[string_idx]
- new_string_leaf.value = new_string_leaf.value.replace("\\\n", "")
-
- return Ok(new_line)
-
- def __merge_string_group(self, line: Line, string_idx: int) -> TResult[Line]:
- """
- Merges string group (i.e. set of adjacent strings) where the first
- string in the group is `line.leaves[string_idx]`.
-
- Returns:
- Ok(new_line), if ALL of the validation checks found in
- __validate_msg(...) pass.
- OR
- Err(CannotTransform), otherwise.
- """
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
-
- vresult = self.__validate_msg(line, string_idx)
- if isinstance(vresult, Err):
- return vresult
-
- # If the string group is wrapped inside an Atom node, we must make sure
- # to later replace that Atom with our new (merged) string leaf.
- atom_node = LL[string_idx].parent
-
- # We will place BREAK_MARK in between every two substrings that we
- # merge. We will then later go through our final result and use the
- # various instances of BREAK_MARK we find to add the right values to
- # the custom split map.
- BREAK_MARK = "@@@@@ BLACK BREAKPOINT MARKER @@@@@"
-
- QUOTE = LL[string_idx].value[-1]
-
- def make_naked(string: str, string_prefix: str) -> str:
- """Strip @string (i.e. make it a "naked" string)
-
- Pre-conditions:
- * assert_is_leaf_string(@string)
-
- Returns:
- A string that is identical to @string except that
- @string_prefix has been stripped, the surrounding QUOTE
- characters have been removed, and any remaining QUOTE
- characters have been escaped.
- """
- assert_is_leaf_string(string)
-
- RE_EVEN_BACKSLASHES = r"(?:(?<!\\)(?:\\\\)*)"
- naked_string = string[len(string_prefix) + 1 : -1]
- naked_string = re.sub(
- "(" + RE_EVEN_BACKSLASHES + ")" + QUOTE, r"\1\\" + QUOTE, naked_string
- )
- return naked_string
-
- # Holds the CustomSplit objects that will later be added to the custom
- # split map.
- custom_splits = []
-
- # Temporary storage for the 'has_prefix' part of the CustomSplit objects.
- prefix_tracker = []
-
- # Sets the 'prefix' variable. This is the prefix that the final merged
- # string will have.
- next_str_idx = string_idx
- prefix = ""
- while (
- not prefix
- and is_valid_index(next_str_idx)
- and LL[next_str_idx].type == token.STRING
- ):
- prefix = get_string_prefix(LL[next_str_idx].value)
- next_str_idx += 1
-
- # The next loop merges the string group. The final string will be
- # contained in 'S'.
- #
- # The following convenience variables are used:
- #
- # S: string
- # NS: naked string
- # SS: next string
- # NSS: naked next string
- S = ""
- NS = ""
- num_of_strings = 0
- next_str_idx = string_idx
- while is_valid_index(next_str_idx) and LL[next_str_idx].type == token.STRING:
- num_of_strings += 1
-
- SS = LL[next_str_idx].value
- next_prefix = get_string_prefix(SS)
-
- # If this is an f-string group but this substring is not prefixed
- # with 'f'...
- if "f" in prefix and "f" not in next_prefix:
- # Then we must escape any braces contained in this substring.
- SS = re.subf(r"(\{|\})", "{1}{1}", SS)
-
- NSS = make_naked(SS, next_prefix)
-
- has_prefix = bool(next_prefix)
- prefix_tracker.append(has_prefix)
-
- S = prefix + QUOTE + NS + NSS + BREAK_MARK + QUOTE
- NS = make_naked(S, prefix)
-
- next_str_idx += 1
-
- S_leaf = Leaf(token.STRING, S)
- if self.normalize_strings:
- normalize_string_quotes(S_leaf)
-
- # Fill the 'custom_splits' list with the appropriate CustomSplit objects.
- temp_string = S_leaf.value[len(prefix) + 1 : -1]
- for has_prefix in prefix_tracker:
- mark_idx = temp_string.find(BREAK_MARK)
- assert (
- mark_idx >= 0
- ), "Logic error while filling the custom string breakpoint cache."
-
- temp_string = temp_string[mark_idx + len(BREAK_MARK) :]
- breakpoint_idx = mark_idx + (len(prefix) if has_prefix else 0) + 1
- custom_splits.append(CustomSplit(has_prefix, breakpoint_idx))
-
- string_leaf = Leaf(token.STRING, S_leaf.value.replace(BREAK_MARK, ""))
-
- if atom_node is not None:
- replace_child(atom_node, string_leaf)
-
- # Build the final line ('new_line') that this method will later return.
- new_line = line.clone()
- for (i, leaf) in enumerate(LL):
- if i == string_idx:
- new_line.append(string_leaf)
-
- if string_idx <= i < string_idx + num_of_strings:
- for comment_leaf in line.comments_after(LL[i]):
- new_line.append(comment_leaf, preformatted=True)
- continue
-
- append_leaves(new_line, line, [leaf])
-
- self.add_custom_splits(string_leaf.value, custom_splits)
- return Ok(new_line)
-
- @staticmethod
- def __validate_msg(line: Line, string_idx: int) -> TResult[None]:
- """Validate (M)erge (S)tring (G)roup
-
- Transform-time string validation logic for __merge_string_group(...).
-
- Returns:
- * Ok(None), if ALL validation checks (listed below) pass.
- OR
- * Err(CannotTransform), if any of the following are true:
- - The target string is not in a string group (i.e. it has no
- adjacent strings).
- - The string group has more than one inline comment.
- - The string group has an inline comment that appears to be a pragma.
- - The set of all string prefixes in the string group is of
- length greater than one and is not equal to {"", "f"}.
- - The string group consists of raw strings.
- """
- num_of_inline_string_comments = 0
- set_of_prefixes = set()
- num_of_strings = 0
- for leaf in line.leaves[string_idx:]:
- if leaf.type != token.STRING:
- # If the string group is trailed by a comma, we count the
- # comments trailing the comma to be one of the string group's
- # comments.
- if leaf.type == token.COMMA and id(leaf) in line.comments:
- num_of_inline_string_comments += 1
- break
-
- if has_triple_quotes(leaf.value):
- return TErr("StringMerger does NOT merge multiline strings.")
-
- num_of_strings += 1
- prefix = get_string_prefix(leaf.value)
- if "r" in prefix:
- return TErr("StringMerger does NOT merge raw strings.")
-
- set_of_prefixes.add(prefix)
-
- if id(leaf) in line.comments:
- num_of_inline_string_comments += 1
- if contains_pragma_comment(line.comments[id(leaf)]):
- return TErr("Cannot merge strings which have pragma comments.")
-
- if num_of_strings < 2:
- return TErr(
- f"Not enough strings to merge (num_of_strings={num_of_strings})."
- )
-
- if num_of_inline_string_comments > 1:
- return TErr(
- f"Too many inline string comments ({num_of_inline_string_comments})."
- )
-
- if len(set_of_prefixes) > 1 and set_of_prefixes != {"", "f"}:
- return TErr(f"Too many different prefixes ({set_of_prefixes}).")
-
- return Ok(None)
-
-
-class StringParenStripper(StringTransformer):
- """StringTransformer that strips surrounding parentheses from strings.
-
- Requirements:
- The line contains a string which is surrounded by parentheses and:
- - The target string is NOT the only argument to a function call).
- - The RPAR is NOT followed by an attribute access (i.e. a dot).
-
- Transformations:
- The parentheses mentioned in the 'Requirements' section are stripped.
-
- Collaborations:
- StringParenStripper has its own inherent usefulness, but it is also
- relied on to clean up the parentheses created by StringParenWrapper (in
- the event that they are no longer needed).
- """
-
- def do_match(self, line: Line) -> TMatchResult:
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
-
- for (idx, leaf) in enumerate(LL):
- # Should be a string...
- if leaf.type != token.STRING:
- continue
-
- # Should be preceded by a non-empty LPAR...
- if (
- not is_valid_index(idx - 1)
- or LL[idx - 1].type != token.LPAR
- or is_empty_lpar(LL[idx - 1])
- ):
- continue
-
- # That LPAR should NOT be preceded by a function name or a closing
- # bracket (which could be a function which returns a function or a
- # list/dictionary that contains a function)...
- if is_valid_index(idx - 2) and (
- LL[idx - 2].type == token.NAME or LL[idx - 2].type in CLOSING_BRACKETS
- ):
- continue
-
- string_idx = idx
-
- # Skip the string trailer, if one exists.
- string_parser = StringParser()
- next_idx = string_parser.parse(LL, string_idx)
-
- # Should be followed by a non-empty RPAR...
- if (
- is_valid_index(next_idx)
- and LL[next_idx].type == token.RPAR
- and not is_empty_rpar(LL[next_idx])
- ):
- # That RPAR should NOT be followed by a '.' symbol.
- if is_valid_index(next_idx + 1) and LL[next_idx + 1].type == token.DOT:
- continue
-
- return Ok(string_idx)
-
- return TErr("This line has no strings wrapped in parens.")
-
- def do_transform(self, line: Line, string_idx: int) -> Iterator[TResult[Line]]:
- LL = line.leaves
-
- string_parser = StringParser()
- rpar_idx = string_parser.parse(LL, string_idx)
-
- for leaf in (LL[string_idx - 1], LL[rpar_idx]):
- if line.comments_after(leaf):
- yield TErr(
- "Will not strip parentheses which have comments attached to them."
- )
-
- new_line = line.clone()
- new_line.comments = line.comments.copy()
-
- append_leaves(new_line, line, LL[: string_idx - 1])
-
- string_leaf = Leaf(token.STRING, LL[string_idx].value)
- LL[string_idx - 1].remove()
- replace_child(LL[string_idx], string_leaf)
- new_line.append(string_leaf)
-
- append_leaves(
- new_line, line, LL[string_idx + 1 : rpar_idx] + LL[rpar_idx + 1 :],
- )
-
- LL[rpar_idx].remove()
-
- yield Ok(new_line)
-
-
-class BaseStringSplitter(StringTransformer):
- """
- Abstract class for StringTransformers which transform a Line's strings by splitting
- them or placing them on their own lines where necessary to avoid going over
- the configured line length.
-
- Requirements:
- * The target string value is responsible for the line going over the
- line length limit. It follows that after all of black's other line
- split methods have been exhausted, this line (or one of the resulting
- lines after all line splits are performed) would still be over the
- line_length limit unless we split this string.
- AND
- * The target string is NOT a "pointless" string (i.e. a string that has
- no parent or siblings).
- AND
- * The target string is not followed by an inline comment that appears
- to be a pragma.
- AND
- * The target string is not a multiline (i.e. triple-quote) string.
- """
-
- @abstractmethod
- def do_splitter_match(self, line: Line) -> TMatchResult:
- """
- BaseStringSplitter asks its clients to override this method instead of
- `StringTransformer.do_match(...)`.
-
- Follows the same protocol as `StringTransformer.do_match(...)`.
-
- Refer to `help(StringTransformer.do_match)` for more information.
- """
-
- def do_match(self, line: Line) -> TMatchResult:
- match_result = self.do_splitter_match(line)
- if isinstance(match_result, Err):
- return match_result
-
- string_idx = match_result.ok()
- vresult = self.__validate(line, string_idx)
- if isinstance(vresult, Err):
- return vresult
-
- return match_result
-
- def __validate(self, line: Line, string_idx: int) -> TResult[None]:
- """
- Checks that @line meets all of the requirements listed in this classes'
- docstring. Refer to `help(BaseStringSplitter)` for a detailed
- description of those requirements.
-
- Returns:
- * Ok(None), if ALL of the requirements are met.
- OR
- * Err(CannotTransform), if ANY of the requirements are NOT met.
- """
- LL = line.leaves
-
- string_leaf = LL[string_idx]
-
- max_string_length = self.__get_max_string_length(line, string_idx)
- if len(string_leaf.value) <= max_string_length:
- return TErr(
- "The string itself is not what is causing this line to be too long."
- )
-
- if not string_leaf.parent or [L.type for L in string_leaf.parent.children] == [
- token.STRING,
- token.NEWLINE,
- ]:
- return TErr(
- f"This string ({string_leaf.value}) appears to be pointless (i.e. has"
- " no parent)."
- )
-
- if id(line.leaves[string_idx]) in line.comments and contains_pragma_comment(
- line.comments[id(line.leaves[string_idx])]
- ):
- return TErr(
- "Line appears to end with an inline pragma comment. Splitting the line"
- " could modify the pragma's behavior."
- )
-
- if has_triple_quotes(string_leaf.value):
- return TErr("We cannot split multiline strings.")
-
- return Ok(None)
-
- def __get_max_string_length(self, line: Line, string_idx: int) -> int:
- """
- Calculates the max string length used when attempting to determine
- whether or not the target string is responsible for causing the line to
- go over the line length limit.
-
- WARNING: This method is tightly coupled to both StringSplitter and
- (especially) StringParenWrapper. There is probably a better way to
- accomplish what is being done here.
-
- Returns:
- max_string_length: such that `line.leaves[string_idx].value >
- max_string_length` implies that the target string IS responsible
- for causing this line to exceed the line length limit.
- """
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
-
- # We use the shorthand "WMA4" in comments to abbreviate "We must
- # account for". When giving examples, we use STRING to mean some/any
- # valid string.
- #
- # Finally, we use the following convenience variables:
- #
- # P: The leaf that is before the target string leaf.
- # N: The leaf that is after the target string leaf.
- # NN: The leaf that is after N.
-
- # WMA4 the whitespace at the beginning of the line.
- offset = line.depth * 4
-
- if is_valid_index(string_idx - 1):
- p_idx = string_idx - 1
- if (
- LL[string_idx - 1].type == token.LPAR
- and LL[string_idx - 1].value == ""
- and string_idx >= 2
- ):
- # If the previous leaf is an empty LPAR placeholder, we should skip it.
- p_idx -= 1
-
- P = LL[p_idx]
- if P.type == token.PLUS:
- # WMA4 a space and a '+' character (e.g. `+ STRING`).
- offset += 2
-
- if P.type == token.COMMA:
- # WMA4 a space, a comma, and a closing bracket [e.g. `), STRING`].
- offset += 3
-
- if P.type in [token.COLON, token.EQUAL, token.NAME]:
- # This conditional branch is meant to handle dictionary keys,
- # variable assignments, 'return STRING' statement lines, and
- # 'else STRING' ternary expression lines.
-
- # WMA4 a single space.
- offset += 1
-
- # WMA4 the lengths of any leaves that came before that space.
- for leaf in LL[: p_idx + 1]:
- offset += len(str(leaf))
-
- if is_valid_index(string_idx + 1):
- N = LL[string_idx + 1]
- if N.type == token.RPAR and N.value == "" and len(LL) > string_idx + 2:
- # If the next leaf is an empty RPAR placeholder, we should skip it.
- N = LL[string_idx + 2]
-
- if N.type == token.COMMA:
- # WMA4 a single comma at the end of the string (e.g `STRING,`).
- offset += 1
-
- if is_valid_index(string_idx + 2):
- NN = LL[string_idx + 2]
-
- if N.type == token.DOT and NN.type == token.NAME:
- # This conditional branch is meant to handle method calls invoked
- # off of a string literal up to and including the LPAR character.
-
- # WMA4 the '.' character.
- offset += 1
-
- if (
- is_valid_index(string_idx + 3)
- and LL[string_idx + 3].type == token.LPAR
- ):
- # WMA4 the left parenthesis character.
- offset += 1
-
- # WMA4 the length of the method's name.
- offset += len(NN.value)
-
- has_comments = False
- for comment_leaf in line.comments_after(LL[string_idx]):
- if not has_comments:
- has_comments = True
- # WMA4 two spaces before the '#' character.
- offset += 2
-
- # WMA4 the length of the inline comment.
- offset += len(comment_leaf.value)
-
- max_string_length = self.line_length - offset
- return max_string_length
-
-
-class StringSplitter(CustomSplitMapMixin, BaseStringSplitter):
- """
- StringTransformer that splits "atom" strings (i.e. strings which exist on
- lines by themselves).
-
- Requirements:
- * The line consists ONLY of a single string (with the exception of a
- '+' symbol which MAY exist at the start of the line), MAYBE a string
- trailer, and MAYBE a trailing comma.
- AND
- * All of the requirements listed in BaseStringSplitter's docstring.
-
- Transformations:
- The string mentioned in the 'Requirements' section is split into as
- many substrings as necessary to adhere to the configured line length.
-
- In the final set of substrings, no substring should be smaller than
- MIN_SUBSTR_SIZE characters.
-
- The string will ONLY be split on spaces (i.e. each new substring should
- start with a space).
-
- If the string is an f-string, it will NOT be split in the middle of an
- f-expression (e.g. in f"FooBar: {foo() if x else bar()}", {foo() if x
- else bar()} is an f-expression).
-
- If the string that is being split has an associated set of custom split
- records and those custom splits will NOT result in any line going over
- the configured line length, those custom splits are used. Otherwise the
- string is split as late as possible (from left-to-right) while still
- adhering to the transformation rules listed above.
-
- Collaborations:
- StringSplitter relies on StringMerger to construct the appropriate
- CustomSplit objects and add them to the custom split map.
- """
-
- MIN_SUBSTR_SIZE = 6
- # Matches an "f-expression" (e.g. {var}) that might be found in an f-string.
- RE_FEXPR = r"""
- (?<!\{)\{
- (?:
- [^\{\}]
- | \{\{
- | \}\}
- )+?
- (?<!\})(?:\}\})*\}(?!\})
- """
-
- def do_splitter_match(self, line: Line) -> TMatchResult:
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
-
- idx = 0
-
- # The first leaf MAY be a '+' symbol...
- if is_valid_index(idx) and LL[idx].type == token.PLUS:
- idx += 1
-
- # The next/first leaf MAY be an empty LPAR...
- if is_valid_index(idx) and is_empty_lpar(LL[idx]):
- idx += 1
-
- # The next/first leaf MUST be a string...
- if not is_valid_index(idx) or LL[idx].type != token.STRING:
- return TErr("Line does not start with a string.")
-
- string_idx = idx
-
- # Skip the string trailer, if one exists.
- string_parser = StringParser()
- idx = string_parser.parse(LL, string_idx)
-
- # That string MAY be followed by an empty RPAR...
- if is_valid_index(idx) and is_empty_rpar(LL[idx]):
- idx += 1
-
- # That string / empty RPAR leaf MAY be followed by a comma...
- if is_valid_index(idx) and LL[idx].type == token.COMMA:
- idx += 1
-
- # But no more leaves are allowed...
- if is_valid_index(idx):
- return TErr("This line does not end with a string.")
-
- return Ok(string_idx)
-
- def do_transform(self, line: Line, string_idx: int) -> Iterator[TResult[Line]]:
- LL = line.leaves
-
- QUOTE = LL[string_idx].value[-1]
-
- is_valid_index = is_valid_index_factory(LL)
- insert_str_child = insert_str_child_factory(LL[string_idx])
-
- prefix = get_string_prefix(LL[string_idx].value)
-
- # We MAY choose to drop the 'f' prefix from substrings that don't
- # contain any f-expressions, but ONLY if the original f-string
- # containes at least one f-expression. Otherwise, we will alter the AST
- # of the program.
- drop_pointless_f_prefix = ("f" in prefix) and re.search(
- self.RE_FEXPR, LL[string_idx].value, re.VERBOSE
- )
-
- first_string_line = True
- starts_with_plus = LL[0].type == token.PLUS
-
- def line_needs_plus() -> bool:
- return first_string_line and starts_with_plus
-
- def maybe_append_plus(new_line: Line) -> None:
- """
- Side Effects:
- If @line starts with a plus and this is the first line we are
- constructing, this function appends a PLUS leaf to @new_line
- and replaces the old PLUS leaf in the node structure. Otherwise
- this function does nothing.
- """
- if line_needs_plus():
- plus_leaf = Leaf(token.PLUS, "+")
- replace_child(LL[0], plus_leaf)
- new_line.append(plus_leaf)
-
- ends_with_comma = (
- is_valid_index(string_idx + 1) and LL[string_idx + 1].type == token.COMMA
- )
-
- def max_last_string() -> int:
- """
- Returns:
- The max allowed length of the string value used for the last
- line we will construct.
- """
- result = self.line_length
- result -= line.depth * 4
- result -= 1 if ends_with_comma else 0
- result -= 2 if line_needs_plus() else 0
- return result
-
- # --- Calculate Max Break Index (for string value)
- # We start with the line length limit
- max_break_idx = self.line_length
- # The last index of a string of length N is N-1.
- max_break_idx -= 1
- # Leading whitespace is not present in the string value (e.g. Leaf.value).
- max_break_idx -= line.depth * 4
- if max_break_idx < 0:
- yield TErr(
- f"Unable to split {LL[string_idx].value} at such high of a line depth:"
- f" {line.depth}"
- )
- return
-
- # Check if StringMerger registered any custom splits.
- custom_splits = self.pop_custom_splits(LL[string_idx].value)
- # We use them ONLY if none of them would produce lines that exceed the
- # line limit.
- use_custom_breakpoints = bool(
- custom_splits
- and all(csplit.break_idx <= max_break_idx for csplit in custom_splits)
- )
-
- # Temporary storage for the remaining chunk of the string line that
- # can't fit onto the line currently being constructed.
- rest_value = LL[string_idx].value
-
- def more_splits_should_be_made() -> bool:
- """
- Returns:
- True iff `rest_value` (the remaining string value from the last
- split), should be split again.
- """
- if use_custom_breakpoints:
- return len(custom_splits) > 1
- else:
- return len(rest_value) > max_last_string()
-
- string_line_results: List[Ok[Line]] = []
- while more_splits_should_be_made():
- if use_custom_breakpoints:
- # Custom User Split (manual)
- csplit = custom_splits.pop(0)
- break_idx = csplit.break_idx
- else:
- # Algorithmic Split (automatic)
- max_bidx = max_break_idx - 2 if line_needs_plus() else max_break_idx
- maybe_break_idx = self.__get_break_idx(rest_value, max_bidx)
- if maybe_break_idx is None:
- # If we are unable to algorthmically determine a good split
- # and this string has custom splits registered to it, we
- # fall back to using them--which means we have to start
- # over from the beginning.
- if custom_splits:
- rest_value = LL[string_idx].value
- string_line_results = []
- first_string_line = True
- use_custom_breakpoints = True
- continue
-
- # Otherwise, we stop splitting here.
- break
-
- break_idx = maybe_break_idx
-
- # --- Construct `next_value`
- next_value = rest_value[:break_idx] + QUOTE
- if (
- # Are we allowed to try to drop a pointless 'f' prefix?
- drop_pointless_f_prefix
- # If we are, will we be successful?
- and next_value != self.__normalize_f_string(next_value, prefix)
- ):
- # If the current custom split did NOT originally use a prefix,
- # then `csplit.break_idx` will be off by one after removing
- # the 'f' prefix.
- break_idx = (
- break_idx + 1
- if use_custom_breakpoints and not csplit.has_prefix
- else break_idx
- )
- next_value = rest_value[:break_idx] + QUOTE
- next_value = self.__normalize_f_string(next_value, prefix)
-
- # --- Construct `next_leaf`
- next_leaf = Leaf(token.STRING, next_value)
- insert_str_child(next_leaf)
- self.__maybe_normalize_string_quotes(next_leaf)
-
- # --- Construct `next_line`
- next_line = line.clone()
- maybe_append_plus(next_line)
- next_line.append(next_leaf)
- string_line_results.append(Ok(next_line))
-
- rest_value = prefix + QUOTE + rest_value[break_idx:]
- first_string_line = False
-
- yield from string_line_results
-
- if drop_pointless_f_prefix:
- rest_value = self.__normalize_f_string(rest_value, prefix)
-
- rest_leaf = Leaf(token.STRING, rest_value)
- insert_str_child(rest_leaf)
-
- # NOTE: I could not find a test case that verifies that the following
- # line is actually necessary, but it seems to be. Otherwise we risk
- # not normalizing the last substring, right?
- self.__maybe_normalize_string_quotes(rest_leaf)
-
- last_line = line.clone()
- maybe_append_plus(last_line)
-
- # If there are any leaves to the right of the target string...
- if is_valid_index(string_idx + 1):
- # We use `temp_value` here to determine how long the last line
- # would be if we were to append all the leaves to the right of the
- # target string to the last string line.
- temp_value = rest_value
- for leaf in LL[string_idx + 1 :]:
- temp_value += str(leaf)
- if leaf.type == token.LPAR:
- break
-
- # Try to fit them all on the same line with the last substring...
- if (
- len(temp_value) <= max_last_string()
- or LL[string_idx + 1].type == token.COMMA
- ):
- last_line.append(rest_leaf)
- append_leaves(last_line, line, LL[string_idx + 1 :])
- yield Ok(last_line)
- # Otherwise, place the last substring on one line and everything
- # else on a line below that...
- else:
- last_line.append(rest_leaf)
- yield Ok(last_line)
-
- non_string_line = line.clone()
- append_leaves(non_string_line, line, LL[string_idx + 1 :])
- yield Ok(non_string_line)
- # Else the target string was the last leaf...
- else:
- last_line.append(rest_leaf)
- last_line.comments = line.comments.copy()
- yield Ok(last_line)
-
- def __get_break_idx(self, string: str, max_break_idx: int) -> Optional[int]:
- """
- This method contains the algorithm that StringSplitter uses to
- determine which character to split each string at.
-
- Args:
- @string: The substring that we are attempting to split.
- @max_break_idx: The ideal break index. We will return this value if it
- meets all the necessary conditions. In the likely event that it
- doesn't we will try to find the closest index BELOW @max_break_idx
- that does. If that fails, we will expand our search by also
- considering all valid indices ABOVE @max_break_idx.
-
- Pre-Conditions:
- * assert_is_leaf_string(@string)
- * 0 <= @max_break_idx < len(@string)
-
- Returns:
- break_idx, if an index is able to be found that meets all of the
- conditions listed in the 'Transformations' section of this classes'
- docstring.
- OR
- None, otherwise.
- """
- is_valid_index = is_valid_index_factory(string)
-
- assert is_valid_index(max_break_idx)
- assert_is_leaf_string(string)
-
- _fexpr_slices: Optional[List[Tuple[Index, Index]]] = None
-
- def fexpr_slices() -> Iterator[Tuple[Index, Index]]:
- """
- Yields:
- All ranges of @string which, if @string were to be split there,
- would result in the splitting of an f-expression (which is NOT
- allowed).
- """
- nonlocal _fexpr_slices
-
- if _fexpr_slices is None:
- _fexpr_slices = []
- for match in re.finditer(self.RE_FEXPR, string, re.VERBOSE):
- _fexpr_slices.append(match.span())
-
- yield from _fexpr_slices
-
- is_fstring = "f" in get_string_prefix(string)
-
- def breaks_fstring_expression(i: Index) -> bool:
- """
- Returns:
- True iff returning @i would result in the splitting of an
- f-expression (which is NOT allowed).
- """
- if not is_fstring:
- return False
-
- for (start, end) in fexpr_slices():
- if start <= i < end:
- return True
-
- return False
-
- def passes_all_checks(i: Index) -> bool:
- """
- Returns:
- True iff ALL of the conditions listed in the 'Transformations'
- section of this classes' docstring would be be met by returning @i.
- """
- is_space = string[i] == " "
- is_big_enough = (
- len(string[i:]) >= self.MIN_SUBSTR_SIZE
- and len(string[:i]) >= self.MIN_SUBSTR_SIZE
- )
- return is_space and is_big_enough and not breaks_fstring_expression(i)
-
- # First, we check all indices BELOW @max_break_idx.
- break_idx = max_break_idx
- while is_valid_index(break_idx - 1) and not passes_all_checks(break_idx):
- break_idx -= 1
-
- if not passes_all_checks(break_idx):
- # If that fails, we check all indices ABOVE @max_break_idx.
- #
- # If we are able to find a valid index here, the next line is going
- # to be longer than the specified line length, but it's probably
- # better than doing nothing at all.
- break_idx = max_break_idx + 1
- while is_valid_index(break_idx + 1) and not passes_all_checks(break_idx):
- break_idx += 1
-
- if not is_valid_index(break_idx) or not passes_all_checks(break_idx):
- return None
-
- return break_idx
-
- def __maybe_normalize_string_quotes(self, leaf: Leaf) -> None:
- if self.normalize_strings:
- normalize_string_quotes(leaf)
-
- def __normalize_f_string(self, string: str, prefix: str) -> str:
- """
- Pre-Conditions:
- * assert_is_leaf_string(@string)
-
- Returns:
- * If @string is an f-string that contains no f-expressions, we
- return a string identical to @string except that the 'f' prefix
- has been stripped and all double braces (i.e. '{{' or '}}') have
- been normalized (i.e. turned into '{' or '}').
- OR
- * Otherwise, we return @string.
- """
- assert_is_leaf_string(string)
-
- if "f" in prefix and not re.search(self.RE_FEXPR, string, re.VERBOSE):
- new_prefix = prefix.replace("f", "")
-
- temp = string[len(prefix) :]
- temp = re.sub(r"\{\{", "{", temp)
- temp = re.sub(r"\}\}", "}", temp)
- new_string = temp
-
- return f"{new_prefix}{new_string}"
- else:
- return string
-
-
-class StringParenWrapper(CustomSplitMapMixin, BaseStringSplitter):
- """
- StringTransformer that splits non-"atom" strings (i.e. strings that do not
- exist on lines by themselves).
-
- Requirements:
- All of the requirements listed in BaseStringSplitter's docstring in
- addition to the requirements listed below:
-
- * The line is a return/yield statement, which returns/yields a string.
- OR
- * The line is part of a ternary expression (e.g. `x = y if cond else
- z`) such that the line starts with `else <string>`, where <string> is
- some string.
- OR
- * The line is an assert statement, which ends with a string.
- OR
- * The line is an assignment statement (e.g. `x = <string>` or `x +=
- <string>`) such that the variable is being assigned the value of some
- string.
- OR
- * The line is a dictionary key assignment where some valid key is being
- assigned the value of some string.
-
- Transformations:
- The chosen string is wrapped in parentheses and then split at the LPAR.
-
- We then have one line which ends with an LPAR and another line that
- starts with the chosen string. The latter line is then split again at
- the RPAR. This results in the RPAR (and possibly a trailing comma)
- being placed on its own line.
-
- NOTE: If any leaves exist to the right of the chosen string (except
- for a trailing comma, which would be placed after the RPAR), those
- leaves are placed inside the parentheses. In effect, the chosen
- string is not necessarily being "wrapped" by parentheses. We can,
- however, count on the LPAR being placed directly before the chosen
- string.
-
- In other words, StringParenWrapper creates "atom" strings. These
- can then be split again by StringSplitter, if necessary.
-
- Collaborations:
- In the event that a string line split by StringParenWrapper is
- changed such that it no longer needs to be given its own line,
- StringParenWrapper relies on StringParenStripper to clean up the
- parentheses it created.
- """
-
- def do_splitter_match(self, line: Line) -> TMatchResult:
- LL = line.leaves
-
- string_idx = None
- string_idx = string_idx or self._return_match(LL)
- string_idx = string_idx or self._else_match(LL)
- string_idx = string_idx or self._assert_match(LL)
- string_idx = string_idx or self._assign_match(LL)
- string_idx = string_idx or self._dict_match(LL)
-
- if string_idx is not None:
- string_value = line.leaves[string_idx].value
- # If the string has no spaces...
- if " " not in string_value:
- # And will still violate the line length limit when split...
- max_string_length = self.line_length - ((line.depth + 1) * 4)
- if len(string_value) > max_string_length:
- # And has no associated custom splits...
- if not self.has_custom_splits(string_value):
- # Then we should NOT put this string on its own line.
- return TErr(
- "We do not wrap long strings in parentheses when the"
- " resultant line would still be over the specified line"
- " length and can't be split further by StringSplitter."
- )
- return Ok(string_idx)
-
- return TErr("This line does not contain any non-atomic strings.")
-
- @staticmethod
- def _return_match(LL: List[Leaf]) -> Optional[int]:
- """
- Returns:
- string_idx such that @LL[string_idx] is equal to our target (i.e.
- matched) string, if this line matches the return/yield statement
- requirements listed in the 'Requirements' section of this classes'
- docstring.
- OR
- None, otherwise.
- """
- # If this line is apart of a return/yield statement and the first leaf
- # contains either the "return" or "yield" keywords...
- if parent_type(LL[0]) in [syms.return_stmt, syms.yield_expr] and LL[
- 0
- ].value in ["return", "yield"]:
- is_valid_index = is_valid_index_factory(LL)
-
- idx = 2 if is_valid_index(1) and is_empty_par(LL[1]) else 1
- # The next visible leaf MUST contain a string...
- if is_valid_index(idx) and LL[idx].type == token.STRING:
- return idx
-
- return None
-
- @staticmethod
- def _else_match(LL: List[Leaf]) -> Optional[int]:
- """
- Returns:
- string_idx such that @LL[string_idx] is equal to our target (i.e.
- matched) string, if this line matches the ternary expression
- requirements listed in the 'Requirements' section of this classes'
- docstring.
- OR
- None, otherwise.
- """
- # If this line is apart of a ternary expression and the first leaf
- # contains the "else" keyword...
- if (
- parent_type(LL[0]) == syms.test
- and LL[0].type == token.NAME
- and LL[0].value == "else"
- ):
- is_valid_index = is_valid_index_factory(LL)
-
- idx = 2 if is_valid_index(1) and is_empty_par(LL[1]) else 1
- # The next visible leaf MUST contain a string...
- if is_valid_index(idx) and LL[idx].type == token.STRING:
- return idx
-
- return None
-
- @staticmethod
- def _assert_match(LL: List[Leaf]) -> Optional[int]:
- """
- Returns:
- string_idx such that @LL[string_idx] is equal to our target (i.e.
- matched) string, if this line matches the assert statement
- requirements listed in the 'Requirements' section of this classes'
- docstring.
- OR
- None, otherwise.
- """
- # If this line is apart of an assert statement and the first leaf
- # contains the "assert" keyword...
- if parent_type(LL[0]) == syms.assert_stmt and LL[0].value == "assert":
- is_valid_index = is_valid_index_factory(LL)
-
- for (i, leaf) in enumerate(LL):
- # We MUST find a comma...
- if leaf.type == token.COMMA:
- idx = i + 2 if is_empty_par(LL[i + 1]) else i + 1
-
- # That comma MUST be followed by a string...
- if is_valid_index(idx) and LL[idx].type == token.STRING:
- string_idx = idx
-
- # Skip the string trailer, if one exists.
- string_parser = StringParser()
- idx = string_parser.parse(LL, string_idx)
-
- # But no more leaves are allowed...
- if not is_valid_index(idx):
- return string_idx
-
- return None
-
- @staticmethod
- def _assign_match(LL: List[Leaf]) -> Optional[int]:
- """
- Returns:
- string_idx such that @LL[string_idx] is equal to our target (i.e.
- matched) string, if this line matches the assignment statement
- requirements listed in the 'Requirements' section of this classes'
- docstring.
- OR
- None, otherwise.
- """
- # If this line is apart of an expression statement or is a function
- # argument AND the first leaf contains a variable name...
- if (
- parent_type(LL[0]) in [syms.expr_stmt, syms.argument, syms.power]
- and LL[0].type == token.NAME
- ):
- is_valid_index = is_valid_index_factory(LL)
-
- for (i, leaf) in enumerate(LL):
- # We MUST find either an '=' or '+=' symbol...
- if leaf.type in [token.EQUAL, token.PLUSEQUAL]:
- idx = i + 2 if is_empty_par(LL[i + 1]) else i + 1
-
- # That symbol MUST be followed by a string...
- if is_valid_index(idx) and LL[idx].type == token.STRING:
- string_idx = idx
-
- # Skip the string trailer, if one exists.
- string_parser = StringParser()
- idx = string_parser.parse(LL, string_idx)
-
- # The next leaf MAY be a comma iff this line is apart
- # of a function argument...
- if (
- parent_type(LL[0]) == syms.argument
- and is_valid_index(idx)
- and LL[idx].type == token.COMMA
- ):
- idx += 1
-
- # But no more leaves are allowed...
- if not is_valid_index(idx):
- return string_idx
-
- return None
-
- @staticmethod
- def _dict_match(LL: List[Leaf]) -> Optional[int]:
- """
- Returns:
- string_idx such that @LL[string_idx] is equal to our target (i.e.
- matched) string, if this line matches the dictionary key assignment
- statement requirements listed in the 'Requirements' section of this
- classes' docstring.
- OR
- None, otherwise.
- """
- # If this line is apart of a dictionary key assignment...
- if syms.dictsetmaker in [parent_type(LL[0]), parent_type(LL[0].parent)]:
- is_valid_index = is_valid_index_factory(LL)
-
- for (i, leaf) in enumerate(LL):
- # We MUST find a colon...
- if leaf.type == token.COLON:
- idx = i + 2 if is_empty_par(LL[i + 1]) else i + 1
-
- # That colon MUST be followed by a string...
- if is_valid_index(idx) and LL[idx].type == token.STRING:
- string_idx = idx
-
- # Skip the string trailer, if one exists.
- string_parser = StringParser()
- idx = string_parser.parse(LL, string_idx)
-
- # That string MAY be followed by a comma...
- if is_valid_index(idx) and LL[idx].type == token.COMMA:
- idx += 1
-
- # But no more leaves are allowed...
- if not is_valid_index(idx):
- return string_idx
-
- return None
-
- def do_transform(self, line: Line, string_idx: int) -> Iterator[TResult[Line]]:
- LL = line.leaves
-
- is_valid_index = is_valid_index_factory(LL)
- insert_str_child = insert_str_child_factory(LL[string_idx])
-
- comma_idx = len(LL) - 1
- ends_with_comma = False
- if LL[comma_idx].type == token.COMMA:
- ends_with_comma = True
-
- leaves_to_steal_comments_from = [LL[string_idx]]
- if ends_with_comma:
- leaves_to_steal_comments_from.append(LL[comma_idx])
-
- # --- First Line
- first_line = line.clone()
- left_leaves = LL[:string_idx]
-
- # We have to remember to account for (possibly invisible) LPAR and RPAR
- # leaves that already wrapped the target string. If these leaves do
- # exist, we will replace them with our own LPAR and RPAR leaves.
- old_parens_exist = False
- if left_leaves and left_leaves[-1].type == token.LPAR:
- old_parens_exist = True
- leaves_to_steal_comments_from.append(left_leaves[-1])
- left_leaves.pop()
-
- append_leaves(first_line, line, left_leaves)
-
- lpar_leaf = Leaf(token.LPAR, "(")
- if old_parens_exist:
- replace_child(LL[string_idx - 1], lpar_leaf)
- else:
- insert_str_child(lpar_leaf)
- first_line.append(lpar_leaf)
-
- # We throw inline comments that were originally to the right of the
- # target string to the top line. They will now be shown to the right of
- # the LPAR.
- for leaf in leaves_to_steal_comments_from:
- for comment_leaf in line.comments_after(leaf):
- first_line.append(comment_leaf, preformatted=True)
-
- yield Ok(first_line)
-
- # --- Middle (String) Line
- # We only need to yield one (possibly too long) string line, since the
- # `StringSplitter` will break it down further if necessary.
- string_value = LL[string_idx].value
- string_line = Line(
- depth=line.depth + 1,
- inside_brackets=True,
- should_explode=line.should_explode,
- )
- string_leaf = Leaf(token.STRING, string_value)
- insert_str_child(string_leaf)
- string_line.append(string_leaf)
-
- old_rpar_leaf = None
- if is_valid_index(string_idx + 1):
- right_leaves = LL[string_idx + 1 :]
- if ends_with_comma:
- right_leaves.pop()
-
- if old_parens_exist:
- assert (
- right_leaves and right_leaves[-1].type == token.RPAR
- ), "Apparently, old parentheses do NOT exist?!"
- old_rpar_leaf = right_leaves.pop()
-
- append_leaves(string_line, line, right_leaves)
-
- yield Ok(string_line)
-
- # --- Last Line
- last_line = line.clone()
- last_line.bracket_tracker = first_line.bracket_tracker
-
- new_rpar_leaf = Leaf(token.RPAR, ")")
- if old_rpar_leaf is not None:
- replace_child(old_rpar_leaf, new_rpar_leaf)
- else:
- insert_str_child(new_rpar_leaf)
- last_line.append(new_rpar_leaf)
-
- # If the target string ended with a comma, we place this comma to the
- # right of the RPAR on the last line.
- if ends_with_comma:
- comma_leaf = Leaf(token.COMMA, ",")
- replace_child(LL[comma_idx], comma_leaf)
- last_line.append(comma_leaf)
-
- yield Ok(last_line)
-
-
-class StringParser:
- """
- A state machine that aids in parsing a string's "trailer", which can be
- either non-existant, an old-style formatting sequence (e.g. `% varX` or `%
- (varX, varY)`), or a method-call / attribute access (e.g. `.format(varX,
- varY)`).
-
- NOTE: A new StringParser object MUST be instantiated for each string
- trailer we need to parse.
-
- Examples:
- We shall assume that `line` equals the `Line` object that corresponds
- to the following line of python code:
- ```
- x = "Some {}.".format("String") + some_other_string
- ```
-
- Furthermore, we will assume that `string_idx` is some index such that:
- ```
- assert line.leaves[string_idx].value == "Some {}."
- ```
-
- The following code snippet then holds:
- ```
- string_parser = StringParser()
- idx = string_parser.parse(line.leaves, string_idx)
- assert line.leaves[idx].type == token.PLUS
- ```
- """
-
- DEFAULT_TOKEN = -1
-
- # String Parser States
- START = 1
- DOT = 2
- NAME = 3
- PERCENT = 4
- SINGLE_FMT_ARG = 5
- LPAR = 6
- RPAR = 7
- DONE = 8
-
- # Lookup Table for Next State
- _goto: Dict[Tuple[ParserState, NodeType], ParserState] = {
- # A string trailer may start with '.' OR '%'.
- (START, token.DOT): DOT,
- (START, token.PERCENT): PERCENT,
- (START, DEFAULT_TOKEN): DONE,
- # A '.' MUST be followed by an attribute or method name.
- (DOT, token.NAME): NAME,
- # A method name MUST be followed by an '(', whereas an attribute name
- # is the last symbol in the string trailer.
- (NAME, token.LPAR): LPAR,
- (NAME, DEFAULT_TOKEN): DONE,
- # A '%' symbol can be followed by an '(' or a single argument (e.g. a
- # string or variable name).
- (PERCENT, token.LPAR): LPAR,
- (PERCENT, DEFAULT_TOKEN): SINGLE_FMT_ARG,
- # If a '%' symbol is followed by a single argument, that argument is
- # the last leaf in the string trailer.
- (SINGLE_FMT_ARG, DEFAULT_TOKEN): DONE,
- # If present, a ')' symbol is the last symbol in a string trailer.
- # (NOTE: LPARS and nested RPARS are not included in this lookup table,
- # since they are treated as a special case by the parsing logic in this
- # classes' implementation.)
- (RPAR, DEFAULT_TOKEN): DONE,
- }
-
- def __init__(self) -> None:
- self._state = self.START
- self._unmatched_lpars = 0
-
- def parse(self, leaves: List[Leaf], string_idx: int) -> int:
- """
- Pre-conditions:
- * @leaves[@string_idx].type == token.STRING
-
- Returns:
- The index directly after the last leaf which is apart of the string
- trailer, if a "trailer" exists.
- OR
- @string_idx + 1, if no string "trailer" exists.
- """
- assert leaves[string_idx].type == token.STRING
-
- idx = string_idx + 1
- while idx < len(leaves) and self._next_state(leaves[idx]):
- idx += 1
- return idx
-
- def _next_state(self, leaf: Leaf) -> bool:
- """
- Pre-conditions:
- * On the first call to this function, @leaf MUST be the leaf that
- was directly after the string leaf in question (e.g. if our target
- string is `line.leaves[i]` then the first call to this method must
- be `line.leaves[i + 1]`).
- * On the next call to this function, the leaf paramater passed in
- MUST be the leaf directly following @leaf.
-
- Returns:
- True iff @leaf is apart of the string's trailer.
- """
- # We ignore empty LPAR or RPAR leaves.
- if is_empty_par(leaf):
- return True
-
- next_token = leaf.type
- if next_token == token.LPAR:
- self._unmatched_lpars += 1
-
- current_state = self._state
-
- # The LPAR parser state is a special case. We will return True until we
- # find the matching RPAR token.
- if current_state == self.LPAR:
- if next_token == token.RPAR:
- self._unmatched_lpars -= 1
- if self._unmatched_lpars == 0:
- self._state = self.RPAR
- # Otherwise, we use a lookup table to determine the next state.
- else:
- # If the lookup table matches the current state to the next
- # token, we use the lookup table.
- if (current_state, next_token) in self._goto:
- self._state = self._goto[current_state, next_token]
- else:
- # Otherwise, we check if a the current state was assigned a
- # default.
- if (current_state, self.DEFAULT_TOKEN) in self._goto:
- self._state = self._goto[current_state, self.DEFAULT_TOKEN]
- # If no default has been assigned, then this parser has a logic
- # error.
- else:
- raise RuntimeError(f"{self.__class__.__name__} LOGIC ERROR!")
-
- if self._state == self.DONE:
- return False
-
- return True
-
-
-def TErr(err_msg: str) -> Err[CannotTransform]:
- """(T)ransform Err
-
- Convenience function used when working with the TResult type.
- """
- cant_transform = CannotTransform(err_msg)
- return Err(cant_transform)
-
-
-def contains_pragma_comment(comment_list: List[Leaf]) -> bool:
- """
- Returns:
- True iff one of the comments in @comment_list is a pragma used by one
- of the more common static analysis tools for python (e.g. mypy, flake8,
- pylint).
- """
- for comment in comment_list:
- if comment.value.startswith(("# type:", "# noqa", "# pylint:")):
- return True
-
- return False
-
-
-def insert_str_child_factory(string_leaf: Leaf) -> Callable[[LN], None]:
- """
- Factory for a convenience function that is used to orphan @string_leaf
- and then insert multiple new leaves into the same part of the node
- structure that @string_leaf had originally occupied.
-
- Examples:
- Let `string_leaf = Leaf(token.STRING, '"foo"')` and `N =
- string_leaf.parent`. Assume the node `N` has the following
- original structure:
-
- Node(
- expr_stmt, [
- Leaf(NAME, 'x'),
- Leaf(EQUAL, '='),
- Leaf(STRING, '"foo"'),
- ]
- )
-
- We then run the code snippet shown below.
- ```
- insert_str_child = insert_str_child_factory(string_leaf)
-
- lpar = Leaf(token.LPAR, '(')
- insert_str_child(lpar)
projects / etc / vim.git / blobdiff