-@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(re.split("\r?\n", prefix)):
- 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, mode: Mode, 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(mode.line_length, mode.string_normalization)
-
- 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_split_rhs
- and not line.magic_trailing_comma
- and (
- is_line_short_enough(line, line_length=mode.line_length, line_str=line_str)
- or line.contains_unsplittable_type_ignore()
- )
- and not (line.inside_brackets and line.contains_standalone_comments())
- ):
- # Only apply basic string preprocessing, since lines shouldn't be split here.
- if mode.experimental_string_processing:
- transformers = [string_merge, string_paren_strip]
- else:
- transformers = []
- elif line.is_def:
- transformers = [left_hand_split]
- else:
-
- def rhs(line: Line, features: Collection[Feature]) -> Iterator[Line]:
- """Wraps calls to `right_hand_split`.
-
- The calls increasingly `omit` right-hand trailers (bracket pairs with
- content), meaning the trailers get glued together to split on another
- bracket pair instead.
- """
- for omit in generate_trailers_to_omit(line, mode.line_length):
- lines = list(
- right_hand_split(line, mode.line_length, features, omit=omit)
- )
- # Note: this check is only able to figure out if the first line of the
- # *current* transformation fits in the line length. This is true only
- # for simple cases. All others require running more transforms via
- # `transform_line()`. This check doesn't know if those would succeed.
- if is_line_short_enough(lines[0], line_length=mode.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, as well as lines that contain
- # trailing commas (those have to be exploded).
- yield from right_hand_split(
- line, line_length=mode.line_length, features=features
- )
-
- if mode.experimental_string_processing:
- if line.inside_brackets:
- transformers = [
- string_merge,
- string_paren_strip,
- string_split,
- delimiter_split,
- standalone_comment_split,
- string_paren_wrap,
- rhs,
- ]
- else:
- transformers = [
- string_merge,
- string_paren_strip,
- string_split,
- string_paren_wrap,
- rhs,
- ]
- else:
- if line.inside_brackets:
- transformers = [delimiter_split, standalone_comment_split, rhs]
- else:
- transformers = [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.
- try:
- result = run_transformer(line, transform, mode, features, line_str=line_str)
- 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 requirements, 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
- __name__ = "StringTransformer"
-
- @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 ALL of the validation checks
- listed in StringMerger.__validate_msg(...)'s docstring pass.
- 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.copy()
- 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 group does not contain ANY stand-alone comments.
- - 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.
- """
- # We first check for "inner" stand-alone comments (i.e. stand-alone
- # comments that have a string leaf before them AND after them).
- for inc in [1, -1]:
- i = string_idx
- found_sa_comment = False
- is_valid_index = is_valid_index_factory(line.leaves)
- while is_valid_index(i) and line.leaves[i].type in [
- token.STRING,
- STANDALONE_COMMENT,
- ]:
- if line.leaves[i].type == STANDALONE_COMMENT:
- found_sa_comment = True
- elif found_sa_comment:
- return TErr(
- "StringMerger does NOT merge string groups which contain "
- "stand-alone comments."
- )
-
- i += inc
-
- 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 target string is NOT a "pointless" string.
- - If the target string contains a PERCENT, the brackets are not
- preceeded or followed by an operator with higher precedence than
- PERCENT.
-
- 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
-
- # If this is a "pointless" string...
- if (
- leaf.parent
- and leaf.parent.parent
- and leaf.parent.parent.type == syms.simple_stmt
- ):
- 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)
-
- # if the leaves in the parsed string include a PERCENT, we need to
- # make sure the initial LPAR is NOT preceded by an operator with
- # higher or equal precedence to PERCENT
- if is_valid_index(idx - 2):
- # mypy can't quite follow unless we name this
- before_lpar = LL[idx - 2]
- if token.PERCENT in {leaf.type for leaf in LL[idx - 1 : next_idx]} and (
- (
- before_lpar.type
- in {
- token.STAR,
- token.AT,
- token.SLASH,
- token.DOUBLESLASH,
- token.PERCENT,
- token.TILDE,
- token.DOUBLESTAR,
- token.AWAIT,
- token.LSQB,
- token.LPAR,
- }
- )
- or (
- # only unary PLUS/MINUS
- before_lpar.parent
- and before_lpar.parent.type == syms.factor
- and (before_lpar.type in {token.PLUS, token.MINUS})
- )
- ):
- continue
-
- # 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 anything with higher
- # precedence than PERCENT
- if is_valid_index(next_idx + 1) and LL[next_idx + 1].type in {
- token.DOUBLESTAR,
- token.LSQB,
- token.LPAR,
- 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."
- )
- return
-
- new_line = line.clone()
- new_line.comments = line.comments.copy()
- try:
- append_leaves(new_line, line, LL[: string_idx - 1])
- except BracketMatchError:
- # HACK: I believe there is currently a bug somewhere in
- # right_hand_split() that is causing brackets to not be tracked
- # properly by a shared BracketTracker.
- append_leaves(new_line, line, LL[: string_idx - 1], preformatted=True)
-
- 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,
- # but after any closing bracket before that space.
- for leaf in reversed(LL[: p_idx + 1]):
- offset += len(str(leaf))
- if leaf.type in CLOSING_BRACKETS:
- break
-
- 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). Note that the string will NOT be split on a space
- which is escaped with a backslash.
-
- 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"""
- (?<!\{) (?:\{\{)* \{ (?!\{)
- (?:
- [^\{\}]
- | \{\{
- | \}\}
- | (?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
- # contains 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 algorithmically 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_not_escaped = True
- j = i - 1
- while is_valid_index(j) and string[j] == "\\":
- is_not_escaped = not is_not_escaped
- j -= 1
-
- is_big_enough = (
- len(string[i:]) >= self.MIN_SUBSTR_SIZE
- and len(string[:i]) >= self.MIN_SUBSTR_SIZE
- )
- return (
- is_space
- and is_not_escaped
- 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 = (
- self._return_match(LL)
- or self._else_match(LL)
- or self._assert_match(LL)
- or self._assign_match(LL)
- 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 = -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(
- mode=line.mode,
- depth=line.depth + 1,
- inside_brackets=True,
- should_split_rhs=line.should_split_rhs,
- magic_trailing_comma=line.magic_trailing_comma,
- )
- 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-existent, 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 parameter 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)
-
- bar = Leaf(token.STRING, '"bar"')
- insert_str_child(bar)
-
- rpar = Leaf(token.RPAR, ')')
- insert_str_child(rpar)
- ```
-
- After which point, it follows that `string_leaf.parent is None` and
- the node `N` now has the following structure:
-
- Node(
- expr_stmt, [
- Leaf(NAME, 'x'),
- Leaf(EQUAL, '='),
- Leaf(LPAR, '('),
- Leaf(STRING, '"bar"'),
- Leaf(RPAR, ')'),
- ]
- )
- """
- string_parent = string_leaf.parent
- string_child_idx = string_leaf.remove()
-
- def insert_str_child(child: LN) -> None:
- nonlocal string_child_idx
-
- assert string_parent is not None
- assert string_child_idx is not None
-
- string_parent.insert_child(string_child_idx, child)
- string_child_idx += 1
-
- return insert_str_child
-
-
-def has_triple_quotes(string: str) -> bool:
- """
- Returns:
- True iff @string starts with three quotation characters.
- """
- raw_string = string.lstrip(STRING_PREFIX_CHARS)
- return raw_string[:3] in {'"""', "'''"}
-
-
-def parent_type(node: Optional[LN]) -> Optional[NodeType]:
- """
- Returns:
- @node.parent.type, if @node is not None and has a parent.
- OR
- None, otherwise.
- """
- if node is None or node.parent is None:
- return None
-
- return node.parent.type
-
-
-def is_empty_par(leaf: Leaf) -> bool:
- return is_empty_lpar(leaf) or is_empty_rpar(leaf)
-
-
-def is_empty_lpar(leaf: Leaf) -> bool:
- return leaf.type == token.LPAR and leaf.value == ""
-
-
-def is_empty_rpar(leaf: Leaf) -> bool:
- return leaf.type == token.RPAR and leaf.value == ""
-
-
-def is_valid_index_factory(seq: Sequence[Any]) -> Callable[[int], bool]:
- """
- Examples:
- ```
- my_list = [1, 2, 3]
-
- is_valid_index = is_valid_index_factory(my_list)
-
- assert is_valid_index(0)
- assert is_valid_index(2)
-
- assert not is_valid_index(3)
- assert not is_valid_index(-1)
- ```
- """
-
- def is_valid_index(idx: int) -> bool:
- """
- Returns:
- True iff @idx is positive AND seq[@idx] does NOT raise an
- IndexError.
- """
- return 0 <= idx < len(seq)
-
- return is_valid_index
-
-
-def line_to_string(line: Line) -> str:
- """Returns the string representation of @line.
-
- WARNING: This is known to be computationally expensive.
- """
- return str(line).strip("\n")
-
-
-def append_leaves(
- new_line: Line, old_line: Line, leaves: List[Leaf], preformatted: bool = False
-) -> None:
- """
- Append leaves (taken from @old_line) to @new_line, making sure to fix the
- underlying Node structure where appropriate.
-
- All of the leaves in @leaves are duplicated. The duplicates are then
- appended to @new_line and used to replace their originals in the underlying
- Node structure. Any comments attached to the old leaves are reattached to
- the new leaves.
-
- Pre-conditions:
- set(@leaves) is a subset of set(@old_line.leaves).
- """
- for old_leaf in leaves:
- new_leaf = Leaf(old_leaf.type, old_leaf.value)
- replace_child(old_leaf, new_leaf)
- new_line.append(new_leaf, preformatted=preformatted)
-
- for comment_leaf in old_line.comments_after(old_leaf):
- new_line.append(comment_leaf, preformatted=True)
-
-
-def replace_child(old_child: LN, new_child: LN) -> None:
- """
- Side Effects:
- * If @old_child.parent is set, replace @old_child with @new_child in
- @old_child's underlying Node structure.
- OR
- * Otherwise, this function does nothing.
- """
- parent = old_child.parent
- if not parent:
- return
-
- child_idx = old_child.remove()
- if child_idx is not None:
- parent.insert_child(child_idx, new_child)
-
-
-def get_string_prefix(string: str) -> str:
- """
- Pre-conditions:
- * assert_is_leaf_string(@string)
-
- Returns:
- @string's prefix (e.g. '', 'r', 'f', or 'rf').
- """
- assert_is_leaf_string(string)
-
- prefix = ""
- prefix_idx = 0
- while string[prefix_idx] in STRING_PREFIX_CHARS:
- prefix += string[prefix_idx].lower()
- prefix_idx += 1
-
- return prefix
-
-
-def assert_is_leaf_string(string: str) -> None:
- """
- Checks the pre-condition that @string has the format that you would expect
- of `leaf.value` where `leaf` is some Leaf such that `leaf.type ==
- token.STRING`. A more precise description of the pre-conditions that are
- checked are listed below.
-
- Pre-conditions:
- * @string starts with either ', ", <prefix>', or <prefix>" where
- `set(<prefix>)` is some subset of `set(STRING_PREFIX_CHARS)`.
- * @string ends with a quote character (' or ").
-
- Raises:
- AssertionError(...) if the pre-conditions listed above are not
- satisfied.
- """
- dquote_idx = string.find('"')
- squote_idx = string.find("'")
- if -1 in [dquote_idx, squote_idx]:
- quote_idx = max(dquote_idx, squote_idx)
- else:
- quote_idx = min(squote_idx, dquote_idx)
-
- assert (
- 0 <= quote_idx < len(string) - 1
- ), f"{string!r} is missing a starting quote character (' or \")."
- assert string[-1] in (
- "'",
- '"',
- ), f"{string!r} is missing an ending quote character (' or \")."
- assert set(string[:quote_idx]).issubset(
- set(STRING_PREFIX_CHARS)
- ), f"{set(string[:quote_idx])} is NOT a subset of {set(STRING_PREFIX_CHARS)}."
-
-
-def left_hand_split(line: Line, _features: Collection[Feature] = ()) -> Iterator[Line]:
- """Split line into many lines, starting with the first matching bracket pair.
-
- Note: this usually looks weird, only use this for function definitions.
- Prefer RHS otherwise. This is why this function is not symmetrical with
- :func:`right_hand_split` which also handles optional parentheses.
- """
- tail_leaves: List[Leaf] = []
- body_leaves: List[Leaf] = []
- head_leaves: List[Leaf] = []
- current_leaves = head_leaves
- matching_bracket: Optional[Leaf] = None
- for leaf in line.leaves:
- if (
- current_leaves is body_leaves
- and leaf.type in CLOSING_BRACKETS
- and leaf.opening_bracket is matching_bracket
- ):
- current_leaves = tail_leaves if body_leaves else head_leaves
- current_leaves.append(leaf)
- if current_leaves is head_leaves:
- if leaf.type in OPENING_BRACKETS:
- matching_bracket = leaf
- current_leaves = body_leaves
- if not matching_bracket:
- raise CannotSplit("No brackets found")
-
- head = bracket_split_build_line(head_leaves, line, matching_bracket)
- body = bracket_split_build_line(body_leaves, line, matching_bracket, is_body=True)
- tail = bracket_split_build_line(tail_leaves, line, matching_bracket)
- bracket_split_succeeded_or_raise(head, body, tail)
- for result in (head, body, tail):
- if result:
- yield result
-
-
-def right_hand_split(
- line: Line,
- line_length: int,
- features: Collection[Feature] = (),
- omit: Collection[LeafID] = (),
-) -> Iterator[Line]:
- """Split line into many lines, starting with the last matching bracket pair.
-
- If the split was by optional parentheses, attempt splitting without them, too.
- `omit` is a collection of closing bracket IDs that shouldn't be considered for
- this split.
-
- Note: running this function modifies `bracket_depth` on the leaves of `line`.
- """
- tail_leaves: List[Leaf] = []
- body_leaves: List[Leaf] = []
- head_leaves: List[Leaf] = []
- current_leaves = tail_leaves
- opening_bracket: Optional[Leaf] = None
- closing_bracket: Optional[Leaf] = None
- for leaf in reversed(line.leaves):
- if current_leaves is body_leaves:
- if leaf is opening_bracket:
- current_leaves = head_leaves if body_leaves else tail_leaves
- current_leaves.append(leaf)
- if current_leaves is tail_leaves:
- if leaf.type in CLOSING_BRACKETS and id(leaf) not in omit:
- opening_bracket = leaf.opening_bracket
- closing_bracket = leaf
- current_leaves = body_leaves
- if not (opening_bracket and closing_bracket and head_leaves):
- # If there is no opening or closing_bracket that means the split failed and
- # all content is in the tail. Otherwise, if `head_leaves` are empty, it means
- # the matching `opening_bracket` wasn't available on `line` anymore.
- raise CannotSplit("No brackets found")
-
- tail_leaves.reverse()
- body_leaves.reverse()
- head_leaves.reverse()
- head = bracket_split_build_line(head_leaves, line, opening_bracket)
- body = bracket_split_build_line(body_leaves, line, opening_bracket, is_body=True)
- tail = bracket_split_build_line(tail_leaves, line, opening_bracket)
- bracket_split_succeeded_or_raise(head, body, tail)
- if (
- Feature.FORCE_OPTIONAL_PARENTHESES not in features
- # the opening bracket is an optional paren
- and opening_bracket.type == token.LPAR
- and not opening_bracket.value
- # the closing bracket is an optional paren
- and closing_bracket.type == token.RPAR
- and not closing_bracket.value
- # it's not an import (optional parens are the only thing we can split on
- # in this case; attempting a split without them is a waste of time)
- and not line.is_import
- # there are no standalone comments in the body
- and not body.contains_standalone_comments(0)
- # and we can actually remove the parens
- and can_omit_invisible_parens(body, line_length, omit_on_explode=omit)
- ):
- omit = {id(closing_bracket), *omit}
- try:
- yield from right_hand_split(line, line_length, features=features, omit=omit)
- return
-
- except CannotSplit:
- if not (
- can_be_split(body)
- or is_line_short_enough(body, line_length=line_length)
- ):
- raise CannotSplit(
- "Splitting failed, body is still too long and can't be split."
- )
-
- elif head.contains_multiline_strings() or tail.contains_multiline_strings():
- raise CannotSplit(
- "The current optional pair of parentheses is bound to fail to"
- " satisfy the splitting algorithm because the head or the tail"
- " contains multiline strings which by definition never fit one"
- " line."
- )
-
- ensure_visible(opening_bracket)
- ensure_visible(closing_bracket)
- for result in (head, body, tail):
- if result:
- yield result
-
-
-def bracket_split_succeeded_or_raise(head: Line, body: Line, tail: Line) -> None:
- """Raise :exc:`CannotSplit` if the last left- or right-hand split failed.
-
- Do nothing otherwise.
-
- A left- or right-hand split is based on a pair of brackets. Content before
- (and including) the opening bracket is left on one line, content inside the
- brackets is put on a separate line, and finally content starting with and
- following the closing bracket is put on a separate line.
-
- Those are called `head`, `body`, and `tail`, respectively. If the split
- produced the same line (all content in `head`) or ended up with an empty `body`
- and the `tail` is just the closing bracket, then it's considered failed.
- """
- tail_len = len(str(tail).strip())
- if not body:
- if tail_len == 0:
- raise CannotSplit("Splitting brackets produced the same line")
-
- elif tail_len < 3:
- raise CannotSplit(
- f"Splitting brackets on an empty body to save {tail_len} characters is"
- " not worth it"
- )
-
-
-def bracket_split_build_line(
- leaves: List[Leaf], original: Line, opening_bracket: Leaf, *, is_body: bool = False
-) -> Line:
- """Return a new line with given `leaves` and respective comments from `original`.
-
- If `is_body` is True, the result line is one-indented inside brackets and as such
- has its first leaf's prefix normalized and a trailing comma added when expected.
- """
- result = Line(mode=original.mode, depth=original.depth)
- if is_body:
- result.inside_brackets = True
- result.depth += 1
- if leaves:
- # Since body is a new indent level, remove spurious leading whitespace.
- normalize_prefix(leaves[0], inside_brackets=True)
- # Ensure a trailing comma for imports and standalone function arguments, but
- # be careful not to add one after any comments or within type annotations.
- no_commas = (
- original.is_def
- and opening_bracket.value == "("
- and not any(leaf.type == token.COMMA for leaf in leaves)
- )
-
- if original.is_import or no_commas:
- for i in range(len(leaves) - 1, -1, -1):
- if leaves[i].type == STANDALONE_COMMENT:
- continue
-
- if leaves[i].type != token.COMMA:
- new_comma = Leaf(token.COMMA, ",")
- leaves.insert(i + 1, new_comma)
- break
-
- # Populate the line
- for leaf in leaves:
- result.append(leaf, preformatted=True)
- for comment_after in original.comments_after(leaf):
- result.append(comment_after, preformatted=True)
- if is_body and should_split_line(result, opening_bracket):
- result.should_split_rhs = True
- return result
-
-
-def dont_increase_indentation(split_func: Transformer) -> Transformer:
- """Normalize prefix of the first leaf in every line returned by `split_func`.
-
- This is a decorator over relevant split functions.
- """
-
- @wraps(split_func)
- def split_wrapper(line: Line, features: Collection[Feature] = ()) -> Iterator[Line]:
- for line in split_func(line, features):
- normalize_prefix(line.leaves[0], inside_brackets=True)
- yield line
-
- return split_wrapper
-
-
-@dont_increase_indentation
-def delimiter_split(line: Line, features: Collection[Feature] = ()) -> Iterator[Line]:
- """Split according to delimiters of the highest priority.
-
- If the appropriate Features are given, the split will add trailing commas
- also in function signatures and calls that contain `*` and `**`.
- """
- try:
- last_leaf = line.leaves[-1]
- except IndexError:
- raise CannotSplit("Line empty")
-
- bt = line.bracket_tracker
- try:
- delimiter_priority = bt.max_delimiter_priority(exclude={id(last_leaf)})
- except ValueError:
- raise CannotSplit("No delimiters found")
-
- if delimiter_priority == DOT_PRIORITY:
- if bt.delimiter_count_with_priority(delimiter_priority) == 1:
- raise CannotSplit("Splitting a single attribute from its owner looks wrong")
-
- current_line = Line(
- mode=line.mode, depth=line.depth, inside_brackets=line.inside_brackets
- )
- lowest_depth = sys.maxsize
- trailing_comma_safe = True
-
- def append_to_line(leaf: Leaf) -> Iterator[Line]:
- """Append `leaf` to current line or to new line if appending impossible."""
- nonlocal current_line
- try:
- current_line.append_safe(leaf, preformatted=True)
- except ValueError:
- yield current_line
-
- current_line = Line(
- mode=line.mode, depth=line.depth, inside_brackets=line.inside_brackets
- )
- current_line.append(leaf)
-
- for leaf in line.leaves:
- yield from append_to_line(leaf)
-
- for comment_after in line.comments_after(leaf):
- yield from append_to_line(comment_after)
-
- lowest_depth = min(lowest_depth, leaf.bracket_depth)
- if leaf.bracket_depth == lowest_depth:
- if is_vararg(leaf, within={syms.typedargslist}):
- trailing_comma_safe = (
- trailing_comma_safe and Feature.TRAILING_COMMA_IN_DEF in features
- )
- elif is_vararg(leaf, within={syms.arglist, syms.argument}):
- trailing_comma_safe = (
- trailing_comma_safe and Feature.TRAILING_COMMA_IN_CALL in features
- )
-
- leaf_priority = bt.delimiters.get(id(leaf))
- if leaf_priority == delimiter_priority:
- yield current_line
-
- current_line = Line(
- mode=line.mode, depth=line.depth, inside_brackets=line.inside_brackets
- )
- if current_line:
- if (
- trailing_comma_safe
- and delimiter_priority == COMMA_PRIORITY
- and current_line.leaves[-1].type != token.COMMA
- and current_line.leaves[-1].type != STANDALONE_COMMENT
- ):
- new_comma = Leaf(token.COMMA, ",")
- current_line.append(new_comma)
- yield current_line
-
-
-@dont_increase_indentation
-def standalone_comment_split(
- line: Line, features: Collection[Feature] = ()
-) -> Iterator[Line]:
- """Split standalone comments from the rest of the line."""
- if not line.contains_standalone_comments(0):
- raise CannotSplit("Line does not have any standalone comments")
-
- current_line = Line(
- mode=line.mode, depth=line.depth, inside_brackets=line.inside_brackets
- )
-
- def append_to_line(leaf: Leaf) -> Iterator[Line]:
- """Append `leaf` to current line or to new line if appending impossible."""
- nonlocal current_line
- try:
- current_line.append_safe(leaf, preformatted=True)
- except ValueError:
- yield current_line
-
- current_line = Line(
- line.mode, depth=line.depth, inside_brackets=line.inside_brackets
- )
- current_line.append(leaf)
-
- for leaf in line.leaves:
- yield from append_to_line(leaf)
-
- for comment_after in line.comments_after(leaf):
- yield from append_to_line(comment_after)
-
- if current_line:
- yield current_line
-
-
-def is_import(leaf: Leaf) -> bool:
- """Return True if the given leaf starts an import statement."""
- p = leaf.parent
- t = leaf.type
- v = leaf.value
- return bool(
- t == token.NAME
- and (
- (v == "import" and p and p.type == syms.import_name)
- or (v == "from" and p and p.type == syms.import_from)
- )
- )
-
-
-def is_type_comment(leaf: Leaf, suffix: str = "") -> bool:
- """Return True if the given leaf is a special comment.
- Only returns true for type comments for now."""
- t = leaf.type
- v = leaf.value
- return t in {token.COMMENT, STANDALONE_COMMENT} and v.startswith("# type:" + suffix)
-
-
-def normalize_prefix(leaf: Leaf, *, inside_brackets: bool) -> None:
- """Leave existing extra newlines if not `inside_brackets`. Remove everything
- else.
-
- Note: don't use backslashes for formatting or you'll lose your voting rights.
- """
- if not inside_brackets:
- spl = leaf.prefix.split("#")
- if "\\" not in spl[0]:
- nl_count = spl[-1].count("\n")
- if len(spl) > 1:
- nl_count -= 1
- leaf.prefix = "\n" * nl_count
- return
-
- leaf.prefix = ""
-
-
-def normalize_string_prefix(leaf: Leaf, remove_u_prefix: bool = False) -> None:
- """Make all string prefixes lowercase.
-
- If remove_u_prefix is given, also removes any u prefix from the string.
-
- Note: Mutates its argument.
- """
- match = re.match(r"^([" + STRING_PREFIX_CHARS + r"]*)(.*)$", leaf.value, re.DOTALL)
- assert match is not None, f"failed to match string {leaf.value!r}"
- orig_prefix = match.group(1)
- new_prefix = orig_prefix.replace("F", "f").replace("B", "b").replace("U", "u")
- if remove_u_prefix:
- new_prefix = new_prefix.replace("u", "")
- leaf.value = f"{new_prefix}{match.group(2)}"
-
-
-def normalize_string_quotes(leaf: Leaf) -> None:
- """Prefer double quotes but only if it doesn't cause more escaping.
-
- Adds or removes backslashes as appropriate. Doesn't parse and fix
- strings nested in f-strings (yet).
-
- Note: Mutates its argument.
- """
- value = leaf.value.lstrip(STRING_PREFIX_CHARS)
- if value[:3] == '"""':
- return
-
- elif value[:3] == "'''":
- orig_quote = "'''"
- new_quote = '"""'
- elif value[0] == '"':
- orig_quote = '"'
- new_quote = "'"
- else:
- orig_quote = "'"
- new_quote = '"'
- first_quote_pos = leaf.value.find(orig_quote)
- if first_quote_pos == -1:
- return # There's an internal error
-
- prefix = leaf.value[:first_quote_pos]
- unescaped_new_quote = re.compile(rf"(([^\\]|^)(\\\\)*){new_quote}")
- escaped_new_quote = re.compile(rf"([^\\]|^)\\((?:\\\\)*){new_quote}")
- escaped_orig_quote = re.compile(rf"([^\\]|^)\\((?:\\\\)*){orig_quote}")
- body = leaf.value[first_quote_pos + len(orig_quote) : -len(orig_quote)]
- if "r" in prefix.casefold():
- if unescaped_new_quote.search(body):
- # There's at least one unescaped new_quote in this raw string
- # so converting is impossible
- return
-
- # Do not introduce or remove backslashes in raw strings
- new_body = body
- else:
- # remove unnecessary escapes
- new_body = sub_twice(escaped_new_quote, rf"\1\2{new_quote}", body)
- if body != new_body:
- # Consider the string without unnecessary escapes as the original
- body = new_body
- leaf.value = f"{prefix}{orig_quote}{body}{orig_quote}"
- new_body = sub_twice(escaped_orig_quote, rf"\1\2{orig_quote}", new_body)
- new_body = sub_twice(unescaped_new_quote, rf"\1\\{new_quote}", new_body)
- if "f" in prefix.casefold():
- matches = re.findall(
- r"""
- (?:[^{]|^)\{ # start of the string or a non-{ followed by a single {
- ([^{].*?) # contents of the brackets except if begins with {{
- \}(?:[^}]|$) # A } followed by end of the string or a non-}
- """,
- new_body,
- re.VERBOSE,
- )
- for m in matches:
- if "\\" in str(m):
- # Do not introduce backslashes in interpolated expressions
- return
-
- if new_quote == '"""' and new_body[-1:] == '"':
- # edge case:
- new_body = new_body[:-1] + '\\"'
- orig_escape_count = body.count("\\")
- new_escape_count = new_body.count("\\")
- if new_escape_count > orig_escape_count:
- return # Do not introduce more escaping
-
- if new_escape_count == orig_escape_count and orig_quote == '"':
- return # Prefer double quotes
-
- leaf.value = f"{prefix}{new_quote}{new_body}{new_quote}"
-
-
-def normalize_numeric_literal(leaf: Leaf) -> None:
- """Normalizes numeric (float, int, and complex) literals.
-
- All letters used in the representation are normalized to lowercase (except
- in Python 2 long literals).
- """
- text = leaf.value.lower()
- if text.startswith(("0o", "0b")):
- # Leave octal and binary literals alone.
- pass
- elif text.startswith("0x"):
- text = format_hex(text)
- elif "e" in text:
- text = format_scientific_notation(text)
- elif text.endswith(("j", "l")):
- text = format_long_or_complex_number(text)
- else:
- text = format_float_or_int_string(text)
- leaf.value = text
-
-
-def format_hex(text: str) -> str:
- """
- Formats a hexadecimal string like "0x12b3"
-
- Uses lowercase because of similarity between "B" and "8", which
- can cause security issues.
- see: https://github.com/psf/black/issues/1692
- """
-
- before, after = text[:2], text[2:]
- return f"{before}{after.lower()}"
-
-
-def format_scientific_notation(text: str) -> str:
- """Formats a numeric string utilizing scentific notation"""
- before, after = text.split("e")
- sign = ""
- if after.startswith("-"):
- after = after[1:]
- sign = "-"
- elif after.startswith("+"):
- after = after[1:]
- before = format_float_or_int_string(before)
- return f"{before}e{sign}{after}"
-
-
-def format_long_or_complex_number(text: str) -> str:
- """Formats a long or complex string like `10L` or `10j`"""
- number = text[:-1]
- suffix = text[-1]
- # Capitalize in "2L" because "l" looks too similar to "1".
- if suffix == "l":
- suffix = "L"
- return f"{format_float_or_int_string(number)}{suffix}"
-
-
-def format_float_or_int_string(text: str) -> str:
- """Formats a float string like "1.0"."""
- if "." not in text:
- return text
-
- before, after = text.split(".")
- return f"{before or 0}.{after or 0}"
-
-
-def normalize_invisible_parens(node: Node, parens_after: Set[str]) -> None:
- """Make existing optional parentheses invisible or create new ones.
-
- `parens_after` is a set of string leaf values immediately after which parens
- should be put.
-
- Standardizes on visible parentheses for single-element tuples, and keeps
- existing visible parentheses for other tuples and generator expressions.
- """
- for pc in list_comments(node.prefix, is_endmarker=False):
- if pc.value in FMT_OFF:
- # This `node` has a prefix with `# fmt: off`, don't mess with parens.
- return
- check_lpar = False
- for index, child in enumerate(list(node.children)):
- # Fixes a bug where invisible parens are not properly stripped from
- # assignment statements that contain type annotations.
- if isinstance(child, Node) and child.type == syms.annassign:
- normalize_invisible_parens(child, parens_after=parens_after)
-
- # Add parentheses around long tuple unpacking in assignments.
- if (
- index == 0
- and isinstance(child, Node)
- and child.type == syms.testlist_star_expr
- ):
- check_lpar = True
-
- if check_lpar:
- if child.type == syms.atom:
- if maybe_make_parens_invisible_in_atom(child, parent=node):
- wrap_in_parentheses(node, child, visible=False)
- elif is_one_tuple(child):
- wrap_in_parentheses(node, child, visible=True)
- elif node.type == syms.import_from:
- # "import from" nodes store parentheses directly as part of
- # the statement
- if child.type == token.LPAR:
- # make parentheses invisible
- child.value = "" # type: ignore
- node.children[-1].value = "" # type: ignore
- elif child.type != token.STAR:
- # insert invisible parentheses
- node.insert_child(index, Leaf(token.LPAR, ""))
- node.append_child(Leaf(token.RPAR, ""))
- break
-
- elif not (isinstance(child, Leaf) and is_multiline_string(child)):
- wrap_in_parentheses(node, child, visible=False)
-
- check_lpar = isinstance(child, Leaf) and child.value in parens_after
-
-
-def normalize_fmt_off(node: Node) -> None:
- """Convert content between `# fmt: off`/`# fmt: on` into standalone comments."""
- try_again = True
- while try_again:
- try_again = convert_one_fmt_off_pair(node)
-
-
-def convert_one_fmt_off_pair(node: Node) -> bool:
- """Convert content of a single `# fmt: off`/`# fmt: on` into a standalone comment.
-
- Returns True if a pair was converted.
- """
- for leaf in node.leaves():
- previous_consumed = 0
- for comment in list_comments(leaf.prefix, is_endmarker=False):
- if comment.value not in FMT_PASS:
- previous_consumed = comment.consumed
- continue
- # We only want standalone comments. If there's no previous leaf or
- # the previous leaf is indentation, it's a standalone comment in
- # disguise.
- if comment.value in FMT_PASS and comment.type != STANDALONE_COMMENT:
- prev = preceding_leaf(leaf)
- if prev:
- if comment.value in FMT_OFF and prev.type not in WHITESPACE:
- continue
- if comment.value in FMT_SKIP and prev.type in WHITESPACE:
- continue
-
- ignored_nodes = list(generate_ignored_nodes(leaf, comment))
- if not ignored_nodes:
- continue
-
- first = ignored_nodes[0] # Can be a container node with the `leaf`.
- parent = first.parent
- prefix = first.prefix
- first.prefix = prefix[comment.consumed :]
- hidden_value = "".join(str(n) for n in ignored_nodes)
- if comment.value in FMT_OFF:
- hidden_value = comment.value + "\n" + hidden_value
- if comment.value in FMT_SKIP:
- hidden_value += " " + comment.value
- if hidden_value.endswith("\n"):
- # That happens when one of the `ignored_nodes` ended with a NEWLINE
- # leaf (possibly followed by a DEDENT).
- hidden_value = hidden_value[:-1]
- first_idx: Optional[int] = None
- for ignored in ignored_nodes:
- index = ignored.remove()
- if first_idx is None:
- first_idx = index
- assert parent is not None, "INTERNAL ERROR: fmt: on/off handling (1)"
- assert first_idx is not None, "INTERNAL ERROR: fmt: on/off handling (2)"
- parent.insert_child(
- first_idx,
- Leaf(
- STANDALONE_COMMENT,
- hidden_value,
- prefix=prefix[:previous_consumed] + "\n" * comment.newlines,
- ),
- )
- return True
-
- return False
-
-
-def generate_ignored_nodes(leaf: Leaf, comment: ProtoComment) -> Iterator[LN]:
- """Starting from the container of `leaf`, generate all leaves until `# fmt: on`.
-
- If comment is skip, returns leaf only.
- Stops at the end of the block.
- """
- container: Optional[LN] = container_of(leaf)
- if comment.value in FMT_SKIP:
- prev_sibling = leaf.prev_sibling
- if comment.value in leaf.prefix and prev_sibling is not None:
- leaf.prefix = leaf.prefix.replace(comment.value, "")
- siblings = [prev_sibling]
- while (
- "\n" not in prev_sibling.prefix
- and prev_sibling.prev_sibling is not None
- ):
- prev_sibling = prev_sibling.prev_sibling
- siblings.insert(0, prev_sibling)
- for sibling in siblings:
- yield sibling
- elif leaf.parent is not None:
- yield leaf.parent
- return
- while container is not None and container.type != token.ENDMARKER:
- if is_fmt_on(container):
- return
-
- # fix for fmt: on in children
- if contains_fmt_on_at_column(container, leaf.column):
- for child in container.children:
- if contains_fmt_on_at_column(child, leaf.column):
- return
- yield child
- else:
- yield container
- container = container.next_sibling
-
-
-def is_fmt_on(container: LN) -> bool:
- """Determine whether formatting is switched on within a container.
- Determined by whether the last `# fmt:` comment is `on` or `off`.
- """
- fmt_on = False
- for comment in list_comments(container.prefix, is_endmarker=False):
- if comment.value in FMT_ON:
- fmt_on = True
- elif comment.value in FMT_OFF:
- fmt_on = False
- return fmt_on
-
-
-def contains_fmt_on_at_column(container: LN, column: int) -> bool:
- """Determine if children at a given column have formatting switched on."""
- for child in container.children:
- if (
- isinstance(child, Node)
- and first_leaf_column(child) == column
- or isinstance(child, Leaf)
- and child.column == column
- ):
- if is_fmt_on(child):
- return True
-
- return False
-
-
-def first_leaf_column(node: Node) -> Optional[int]:
- """Returns the column of the first leaf child of a node."""
- for child in node.children:
- if isinstance(child, Leaf):
- return child.column
- return None
-
-
-def maybe_make_parens_invisible_in_atom(node: LN, parent: LN) -> bool:
- """If it's safe, make the parens in the atom `node` invisible, recursively.
- Additionally, remove repeated, adjacent invisible parens from the atom `node`
- as they are redundant.
-
- Returns whether the node should itself be wrapped in invisible parentheses.
-
- """
-
- if (
- node.type != syms.atom
- or is_empty_tuple(node)
- or is_one_tuple(node)
- or (is_yield(node) and parent.type != syms.expr_stmt)
- or max_delimiter_priority_in_atom(node) >= COMMA_PRIORITY
- ):
- return False
-
- if is_walrus_assignment(node):
- if parent.type in [syms.annassign, syms.expr_stmt]:
- return False
-
- first = node.children[0]
- last = node.children[-1]
- if first.type == token.LPAR and last.type == token.RPAR:
- middle = node.children[1]
- # make parentheses invisible
- first.value = "" # type: ignore
- last.value = "" # type: ignore
- maybe_make_parens_invisible_in_atom(middle, parent=parent)
-
- if is_atom_with_invisible_parens(middle):
- # Strip the invisible parens from `middle` by replacing
- # it with the child in-between the invisible parens
- middle.replace(middle.children[1])
-
- return False
-
- return True
-
-
-def is_atom_with_invisible_parens(node: LN) -> bool:
- """Given a `LN`, determines whether it's an atom `node` with invisible
- parens. Useful in dedupe-ing and normalizing parens.
- """
- if isinstance(node, Leaf) or node.type != syms.atom:
- return False
-
- first, last = node.children[0], node.children[-1]
- return (
- isinstance(first, Leaf)
- and first.type == token.LPAR
- and first.value == ""
- and isinstance(last, Leaf)
- and last.type == token.RPAR
- and last.value == ""
- )
-
-
-def is_empty_tuple(node: LN) -> bool:
- """Return True if `node` holds an empty tuple."""
- return (
- node.type == syms.atom
- and len(node.children) == 2
- and node.children[0].type == token.LPAR
- and node.children[1].type == token.RPAR
- )
-
-
-def unwrap_singleton_parenthesis(node: LN) -> Optional[LN]:
- """Returns `wrapped` if `node` is of the shape ( wrapped ).
-
- Parenthesis can be optional. Returns None otherwise"""
- if len(node.children) != 3:
- return None
-
- lpar, wrapped, rpar = node.children
- if not (lpar.type == token.LPAR and rpar.type == token.RPAR):
- return None
-
- return wrapped
-
-
-def first_child_is_arith(node: Node) -> bool:
- """Whether first child is an arithmetic or a binary arithmetic expression"""
- expr_types = {
- syms.arith_expr,
- syms.shift_expr,
- syms.xor_expr,
- syms.and_expr,
- }
- return bool(node.children and node.children[0].type in expr_types)
-
-
-def wrap_in_parentheses(parent: Node, child: LN, *, visible: bool = True) -> None:
- """Wrap `child` in parentheses.
-
- This replaces `child` with an atom holding the parentheses and the old
- child. That requires moving the prefix.
-
- If `visible` is False, the leaves will be valueless (and thus invisible).
- """
- lpar = Leaf(token.LPAR, "(" if visible else "")
- rpar = Leaf(token.RPAR, ")" if visible else "")
- prefix = child.prefix
- child.prefix = ""
- index = child.remove() or 0
- new_child = Node(syms.atom, [lpar, child, rpar])
- new_child.prefix = prefix
- parent.insert_child(index, new_child)
-
-
-def is_one_tuple(node: LN) -> bool:
- """Return True if `node` holds a tuple with one element, with or without parens."""
- if node.type == syms.atom:
- gexp = unwrap_singleton_parenthesis(node)
- if gexp is None or gexp.type != syms.testlist_gexp:
- return False
-
- return len(gexp.children) == 2 and gexp.children[1].type == token.COMMA
-
- return (
- node.type in IMPLICIT_TUPLE
- and len(node.children) == 2
- and node.children[1].type == token.COMMA
- )
-
-
-def is_walrus_assignment(node: LN) -> bool:
- """Return True iff `node` is of the shape ( test := test )"""
- inner = unwrap_singleton_parenthesis(node)
- return inner is not None and inner.type == syms.namedexpr_test
-
-
-def is_simple_decorator_trailer(node: LN, last: bool = False) -> bool:
- """Return True iff `node` is a trailer valid in a simple decorator"""
- return node.type == syms.trailer and (
- (
- len(node.children) == 2
- and node.children[0].type == token.DOT
- and node.children[1].type == token.NAME
- )
- # last trailer can be arguments
- or (
- last
- and len(node.children) == 3
- and node.children[0].type == token.LPAR
- # and node.children[1].type == syms.argument
- and node.children[2].type == token.RPAR
- )
- )
-
-
-def is_simple_decorator_expression(node: LN) -> bool:
- """Return True iff `node` could be a 'dotted name' decorator
-
- This function takes the node of the 'namedexpr_test' of the new decorator
- grammar and test if it would be valid under the old decorator grammar.
-
- The old grammar was: decorator: @ dotted_name [arguments] NEWLINE
- The new grammar is : decorator: @ namedexpr_test NEWLINE
- """
- if node.type == token.NAME:
- return True
- if node.type == syms.power:
- if node.children:
- return (
- node.children[0].type == token.NAME
- and all(map(is_simple_decorator_trailer, node.children[1:-1]))
- and (
- len(node.children) < 2
- or is_simple_decorator_trailer(node.children[-1], last=True)
- )
- )
- return False
-
-
-def is_yield(node: LN) -> bool:
- """Return True if `node` holds a `yield` or `yield from` expression."""
- if node.type == syms.yield_expr:
- return True
-
- if node.type == token.NAME and node.value == "yield": # type: ignore
- return True
-
- if node.type != syms.atom:
- return False
-
- if len(node.children) != 3:
- return False
-
- lpar, expr, rpar = node.children
- if lpar.type == token.LPAR and rpar.type == token.RPAR:
- return is_yield(expr)
-
- return False
-
-
-def is_vararg(leaf: Leaf, within: Set[NodeType]) -> bool:
- """Return True if `leaf` is a star or double star in a vararg or kwarg.
-
- If `within` includes VARARGS_PARENTS, this applies to function signatures.
- If `within` includes UNPACKING_PARENTS, it applies to right hand-side
- extended iterable unpacking (PEP 3132) and additional unpacking
- generalizations (PEP 448).
- """
- if leaf.type not in VARARGS_SPECIALS or not leaf.parent:
- return False
-
- p = leaf.parent
- if p.type == syms.star_expr:
- # Star expressions are also used as assignment targets in extended
- # iterable unpacking (PEP 3132). See what its parent is instead.
- if not p.parent:
- return False
-
- p = p.parent
-
- return p.type in within
-
-
-def is_multiline_string(leaf: Leaf) -> bool:
- """Return True if `leaf` is a multiline string that actually spans many lines."""
- return has_triple_quotes(leaf.value) and "\n" in leaf.value
-
-
-def is_stub_suite(node: Node) -> bool:
- """Return True if `node` is a suite with a stub body."""
- if (
- len(node.children) != 4
- or node.children[0].type != token.NEWLINE
- or node.children[1].type != token.INDENT
- or node.children[3].type != token.DEDENT
- ):
- return False
-
- return is_stub_body(node.children[2])
-
-
-def is_stub_body(node: LN) -> bool:
- """Return True if `node` is a simple statement containing an ellipsis."""
- if not isinstance(node, Node) or node.type != syms.simple_stmt:
- return False
-
- if len(node.children) != 2:
- return False
-
- child = node.children[0]
- return (
- child.type == syms.atom
- and len(child.children) == 3
- and all(leaf == Leaf(token.DOT, ".") for leaf in child.children)
- )
-
-
-def max_delimiter_priority_in_atom(node: LN) -> Priority:
- """Return maximum delimiter priority inside `node`.
-
- This is specific to atoms with contents contained in a pair of parentheses.
- If `node` isn't an atom or there are no enclosing parentheses, returns 0.
- """
- if node.type != syms.atom:
- return 0
-
- first = node.children[0]
- last = node.children[-1]
- if not (first.type == token.LPAR and last.type == token.RPAR):
- return 0
-
- bt = BracketTracker()
- for c in node.children[1:-1]:
- if isinstance(c, Leaf):
- bt.mark(c)
- else:
- for leaf in c.leaves():
- bt.mark(leaf)
- try:
- return bt.max_delimiter_priority()
-
- except ValueError:
- return 0
-
-
-def ensure_visible(leaf: Leaf) -> None:
- """Make sure parentheses are visible.
-
- They could be invisible as part of some statements (see
- :func:`normalize_invisible_parens` and :func:`visit_import_from`).
- """
- if leaf.type == token.LPAR:
- leaf.value = "("
- elif leaf.type == token.RPAR:
- leaf.value = ")"
-
-
-def should_split_line(line: Line, opening_bracket: Leaf) -> bool:
- """Should `line` be immediately split with `delimiter_split()` after RHS?"""
-
- if not (opening_bracket.parent and opening_bracket.value in "[{("):
- return False
-
- # We're essentially checking if the body is delimited by commas and there's more
- # than one of them (we're excluding the trailing comma and if the delimiter priority
- # is still commas, that means there's more).
- exclude = set()
- trailing_comma = False
- try:
- last_leaf = line.leaves[-1]
- if last_leaf.type == token.COMMA:
- trailing_comma = True
- exclude.add(id(last_leaf))
- max_priority = line.bracket_tracker.max_delimiter_priority(exclude=exclude)
- except (IndexError, ValueError):
- return False
-
- return max_priority == COMMA_PRIORITY and (
- (line.mode.magic_trailing_comma and trailing_comma)
- # always explode imports
- or opening_bracket.parent.type in {syms.atom, syms.import_from}
- )
-
-
-def is_one_tuple_between(opening: Leaf, closing: Leaf, leaves: List[Leaf]) -> bool:
- """Return True if content between `opening` and `closing` looks like a one-tuple."""
- if opening.type != token.LPAR and closing.type != token.RPAR:
- return False
-
- depth = closing.bracket_depth + 1
- for _opening_index, leaf in enumerate(leaves):
- if leaf is opening:
- break
-
- else:
- raise LookupError("Opening paren not found in `leaves`")
-
- commas = 0
- _opening_index += 1
- for leaf in leaves[_opening_index:]:
- if leaf is closing:
- break
-
- bracket_depth = leaf.bracket_depth
- if bracket_depth == depth and leaf.type == token.COMMA:
- commas += 1
- if leaf.parent and leaf.parent.type in {
- syms.arglist,
- syms.typedargslist,
- }:
- commas += 1
- break
-
- return commas < 2
-
-
-def get_features_used(node: Node) -> Set[Feature]:
- """Return a set of (relatively) new Python features used in this file.
-
- Currently looking for:
- - f-strings;
- - underscores in numeric literals;
- - trailing commas after * or ** in function signatures and calls;
- - positional only arguments in function signatures and lambdas;
- - assignment expression;
- - relaxed decorator syntax;
- """
- features: Set[Feature] = set()
- for n in node.pre_order():
- if n.type == token.STRING:
- value_head = n.value[:2] # type: ignore
- if value_head in {'f"', 'F"', "f'", "F'", "rf", "fr", "RF", "FR"}:
- features.add(Feature.F_STRINGS)
-
- elif n.type == token.NUMBER:
- if "_" in n.value: # type: ignore
- features.add(Feature.NUMERIC_UNDERSCORES)
-
- elif n.type == token.SLASH:
- if n.parent and n.parent.type in {syms.typedargslist, syms.arglist}:
- features.add(Feature.POS_ONLY_ARGUMENTS)
-
- elif n.type == token.COLONEQUAL:
- features.add(Feature.ASSIGNMENT_EXPRESSIONS)
-
- elif n.type == syms.decorator:
- if len(n.children) > 1 and not is_simple_decorator_expression(
- n.children[1]
- ):
- features.add(Feature.RELAXED_DECORATORS)
projects / etc / vim.git / blobdiff