X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/f2ea461e9e9fa5c47bb61fd72d512c748928badc..249c6536c4dff50773f30f222d1f81f0afe41f4c:/src/black/comments.py diff --git a/src/black/comments.py b/src/black/comments.py index 415e391..7069da1 100644 --- a/src/black/comments.py +++ b/src/black/comments.py @@ -1,22 +1,29 @@ +import sys from dataclasses import dataclass from functools import lru_cache -import regex as re +import re from typing import Iterator, List, Optional, Union -from blib2to3.pytree import Node, Leaf +if sys.version_info >= (3, 8): + from typing import Final +else: + from typing_extensions import Final + +from blib2to3.pytree import Node, Leaf, type_repr from blib2to3.pgen2 import token from black.nodes import first_leaf_column, preceding_leaf, container_of -from black.nodes import STANDALONE_COMMENT, WHITESPACE +from black.nodes import CLOSING_BRACKETS, STANDALONE_COMMENT, WHITESPACE # types LN = Union[Leaf, Node] +FMT_OFF: Final = {"# fmt: off", "# fmt:off", "# yapf: disable"} +FMT_SKIP: Final = {"# fmt: skip", "# fmt:skip"} +FMT_PASS: Final = {*FMT_OFF, *FMT_SKIP} +FMT_ON: Final = {"# fmt: on", "# fmt:on", "# yapf: enable"} -FMT_OFF = {"# fmt: off", "# fmt:off", "# yapf: disable"} -FMT_SKIP = {"# fmt: skip", "# fmt:skip"} -FMT_PASS = {*FMT_OFF, *FMT_SKIP} -FMT_ON = {"# fmt: on", "# fmt:on", "# yapf: enable"} +COMMENT_EXCEPTIONS = {True: " !:#'", False: " !:#'%"} @dataclass @@ -37,7 +44,7 @@ class ProtoComment: consumed: int # how many characters of the original leaf's prefix did we consume -def generate_comments(leaf: LN) -> Iterator[Leaf]: +def generate_comments(leaf: LN, *, preview: bool) -> Iterator[Leaf]: """Clean the prefix of the `leaf` and generate comments from it, if any. Comments in lib2to3 are shoved into the whitespace prefix. This happens @@ -56,12 +63,16 @@ def generate_comments(leaf: LN) -> Iterator[Leaf]: Inline comments are emitted as regular token.COMMENT leaves. Standalone are emitted with a fake STANDALONE_COMMENT token identifier. """ - for pc in list_comments(leaf.prefix, is_endmarker=leaf.type == token.ENDMARKER): + for pc in list_comments( + leaf.prefix, is_endmarker=leaf.type == token.ENDMARKER, preview=preview + ): yield Leaf(pc.type, pc.value, prefix="\n" * pc.newlines) @lru_cache(maxsize=4096) -def list_comments(prefix: str, *, is_endmarker: bool) -> List[ProtoComment]: +def list_comments( + prefix: str, *, is_endmarker: bool, preview: 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: @@ -87,7 +98,7 @@ def list_comments(prefix: str, *, is_endmarker: bool) -> List[ProtoComment]: comment_type = token.COMMENT # simple trailing comment else: comment_type = STANDALONE_COMMENT - comment = make_comment(line) + comment = make_comment(line, preview=preview) result.append( ProtoComment( type=comment_type, value=comment, newlines=nlines, consumed=consumed @@ -97,10 +108,10 @@ def list_comments(prefix: str, *, is_endmarker: bool) -> List[ProtoComment]: return result -def make_comment(content: str) -> str: +def make_comment(content: str, *, preview: bool) -> str: """Return a consistently formatted comment from the given `content` string. - All comments (except for "##", "#!", "#:", '#'", "#%%") should have a single + 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. @@ -118,26 +129,26 @@ def make_comment(content: str) -> str: and not content.lstrip().startswith("type:") ): content = " " + content[1:] # Replace NBSP by a simple space - if content and content[0] not in " !:#'%": + if content and content[0] not in COMMENT_EXCEPTIONS[preview]: content = " " + content return "#" + content -def normalize_fmt_off(node: Node) -> None: +def normalize_fmt_off(node: Node, *, preview: bool) -> 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) + try_again = convert_one_fmt_off_pair(node, preview=preview) -def convert_one_fmt_off_pair(node: Node) -> bool: +def convert_one_fmt_off_pair(node: Node, *, preview: bool) -> 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): + for comment in list_comments(leaf.prefix, is_endmarker=False, preview=preview): if comment.value not in FMT_PASS: previous_consumed = comment.consumed continue @@ -152,14 +163,22 @@ def convert_one_fmt_off_pair(node: Node) -> bool: if comment.value in FMT_SKIP and prev.type in WHITESPACE: continue - ignored_nodes = list(generate_ignored_nodes(leaf, comment)) + ignored_nodes = list(generate_ignored_nodes(leaf, comment, preview=preview)) 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 :] + if comment.value in FMT_OFF: + first.prefix = prefix[comment.consumed :] + if comment.value in FMT_SKIP: + first.prefix = "" + standalone_comment_prefix = prefix + else: + standalone_comment_prefix = ( + prefix[:previous_consumed] + "\n" * comment.newlines + ) hidden_value = "".join(str(n) for n in ignored_nodes) if comment.value in FMT_OFF: hidden_value = comment.value + "\n" + hidden_value @@ -181,7 +200,7 @@ def convert_one_fmt_off_pair(node: Node) -> bool: Leaf( STANDALONE_COMMENT, hidden_value, - prefix=prefix[:previous_consumed] + "\n" * comment.newlines, + prefix=standalone_comment_prefix, ), ) return True @@ -189,37 +208,34 @@ def convert_one_fmt_off_pair(node: Node) -> bool: return False -def generate_ignored_nodes(leaf: Leaf, comment: ProtoComment) -> Iterator[LN]: +def generate_ignored_nodes( + leaf: Leaf, comment: ProtoComment, *, preview: bool +) -> 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 + yield from _generate_ignored_nodes_from_fmt_skip(leaf, comment, preview=preview) return + container: Optional[LN] = container_of(leaf) while container is not None and container.type != token.ENDMARKER: - if is_fmt_on(container): + if is_fmt_on(container, preview=preview): return # fix for fmt: on in children - if contains_fmt_on_at_column(container, leaf.column): + if contains_fmt_on_at_column(container, leaf.column, preview=preview): for child in container.children: - if contains_fmt_on_at_column(child, leaf.column): + if isinstance(child, Leaf) and is_fmt_on(child, preview=preview): + if child.type in CLOSING_BRACKETS: + # This means `# fmt: on` is placed at a different bracket level + # than `# fmt: off`. This is an invalid use, but as a courtesy, + # we include this closing bracket in the ignored nodes. + # The alternative is to fail the formatting. + yield child + return + if contains_fmt_on_at_column(child, leaf.column, preview=preview): return yield child else: @@ -227,12 +243,57 @@ def generate_ignored_nodes(leaf: Leaf, comment: ProtoComment) -> Iterator[LN]: container = container.next_sibling -def is_fmt_on(container: LN) -> bool: +def _generate_ignored_nodes_from_fmt_skip( + leaf: Leaf, comment: ProtoComment, *, preview: bool +) -> Iterator[LN]: + """Generate all leaves that should be ignored by the `# fmt: skip` from `leaf`.""" + prev_sibling = leaf.prev_sibling + parent = leaf.parent + # Need to properly format the leaf prefix to compare it to comment.value, + # which is also formatted + comments = list_comments(leaf.prefix, is_endmarker=False, preview=preview) + if not comments or comment.value != comments[0].value: + return + if prev_sibling is not None: + leaf.prefix = "" + 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 ( + parent is not None + and type_repr(parent.type) == "suite" + and leaf.type == token.NEWLINE + ): + # The `# fmt: skip` is on the colon line of the if/while/def/class/... + # statements. The ignored nodes should be previous siblings of the + # parent suite node. + leaf.prefix = "" + ignored_nodes: List[LN] = [] + parent_sibling = parent.prev_sibling + while parent_sibling is not None and type_repr(parent_sibling.type) != "suite": + ignored_nodes.insert(0, parent_sibling) + parent_sibling = parent_sibling.prev_sibling + # Special case for `async_stmt` where the ASYNC token is on the + # grandparent node. + grandparent = parent.parent + if ( + grandparent is not None + and grandparent.prev_sibling is not None + and grandparent.prev_sibling.type == token.ASYNC + ): + ignored_nodes.insert(0, grandparent.prev_sibling) + yield from iter(ignored_nodes) + + +def is_fmt_on(container: LN, preview: bool) -> 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): + for comment in list_comments(container.prefix, is_endmarker=False, preview=preview): if comment.value in FMT_ON: fmt_on = True elif comment.value in FMT_OFF: @@ -240,7 +301,7 @@ def is_fmt_on(container: LN) -> bool: return fmt_on -def contains_fmt_on_at_column(container: LN, column: int) -> bool: +def contains_fmt_on_at_column(container: LN, column: int, *, preview: bool) -> bool: """Determine if children at a given column have formatting switched on.""" for child in container.children: if ( @@ -249,7 +310,7 @@ def contains_fmt_on_at_column(container: LN, column: int) -> bool: or isinstance(child, Leaf) and child.column == column ): - if is_fmt_on(child): + if is_fmt_on(child, preview=preview): return True return False