]> git.madduck.net Git - etc/vim.git/blobdiff - src/black/lines.py

madduck's git repository

Every one of the projects in this repository is available at the canonical URL git://git.madduck.net/madduck/pub/<projectpath> — see each project's metadata for the exact URL.

All patches and comments are welcome. Please squash your changes to logical commits before using git-format-patch and git-send-email to patches@git.madduck.net. If you'd read over the Git project's submission guidelines and adhered to them, I'd be especially grateful.

SSH access, as well as push access can be individually arranged.

If you use my repositories frequently, consider adding the following snippet to ~/.gitconfig and using the third clone URL listed for each project:

[url "git://git.madduck.net/madduck/"]
  insteadOf = madduck:

Improve multiline string handling (#1879)
[etc/vim.git] / src / black / lines.py
index 30622650d53cd631cdcb800ce8e2a546b0de6840..b65604864a4b73fabb69f893e13b99c39806ea93 100644 (file)
@@ -1,4 +1,5 @@
 import itertools
+import math
 import sys
 from dataclasses import dataclass, field
 from typing import (
@@ -10,6 +11,7 @@ from typing import (
     Sequence,
     Tuple,
     TypeVar,
+    Union,
     cast,
 )
 
@@ -37,6 +39,7 @@ from blib2to3.pytree import Leaf, Node
 T = TypeVar("T")
 Index = int
 LeafID = int
+LN = Union[Leaf, Node]
 
 
 @dataclass
@@ -53,7 +56,9 @@ class Line:
     should_split_rhs: bool = False
     magic_trailing_comma: Optional[Leaf] = None
 
-    def append(self, leaf: Leaf, preformatted: bool = False) -> None:
+    def append(
+        self, leaf: Leaf, preformatted: bool = False, track_bracket: bool = False
+    ) -> None:
         """Add a new `leaf` to the end of the line.
 
         Unless `preformatted` is True, the `leaf` will receive a new consistent
@@ -75,7 +80,7 @@ class Line:
             leaf.prefix += whitespace(
                 leaf, complex_subscript=self.is_complex_subscript(leaf)
             )
-        if self.inside_brackets or not preformatted:
+        if self.inside_brackets or not preformatted or track_bracket:
             self.bracket_tracker.mark(leaf)
             if self.mode.magic_trailing_comma:
                 if self.has_magic_trailing_comma(leaf):
@@ -273,8 +278,7 @@ class Line:
         - it's not a single-element subscript
         Additionally, if ensure_removable:
         - it's not from square bracket indexing
-        (specifically, single-element square bracket indexing with
-        Preview.skip_magic_trailing_comma_in_subscript)
+        (specifically, single-element square bracket indexing)
         """
         if not (
             closing.type in CLOSING_BRACKETS
@@ -288,8 +292,7 @@ class Line:
 
         if closing.type == token.RSQB:
             if (
-                Preview.one_element_subscript in self.mode
-                and closing.parent
+                closing.parent
                 and closing.parent.type == syms.trailer
                 and closing.opening_bracket
                 and is_one_sequence_between(
@@ -307,18 +310,16 @@ class Line:
             comma = self.leaves[-1]
             if comma.parent is None:
                 return False
-            if Preview.skip_magic_trailing_comma_in_subscript in self.mode:
-                return (
-                    comma.parent.type != syms.subscriptlist
-                    or closing.opening_bracket is None
-                    or not is_one_sequence_between(
-                        closing.opening_bracket,
-                        closing,
-                        self.leaves,
-                        brackets=(token.LSQB, token.RSQB),
-                    )
+            return (
+                comma.parent.type != syms.subscriptlist
+                or closing.opening_bracket is None
+                or not is_one_sequence_between(
+                    closing.opening_bracket,
+                    closing,
+                    self.leaves,
+                    brackets=(token.LSQB, token.RSQB),
                 )
-            return comma.parent.type == syms.listmaker
+            )
 
         if self.is_import:
             return True
@@ -448,6 +449,28 @@ class Line:
         return bool(self.leaves or self.comments)
 
 
+@dataclass
+class LinesBlock:
+    """Class that holds information about a block of formatted lines.
+
+    This is introduced so that the EmptyLineTracker can look behind the standalone
+    comments and adjust their empty lines for class or def lines.
+    """
+
+    mode: Mode
+    previous_block: Optional["LinesBlock"]
+    original_line: Line
+    before: int = 0
+    content_lines: List[str] = field(default_factory=list)
+    after: int = 0
+
+    def all_lines(self) -> List[str]:
+        empty_line = str(Line(mode=self.mode))
+        return (
+            [empty_line * self.before] + self.content_lines + [empty_line * self.after]
+        )
+
+
 @dataclass
 class EmptyLineTracker:
     """Provides a stateful method that returns the number of potential extra
@@ -458,33 +481,56 @@ class EmptyLineTracker:
     are consumed by `maybe_empty_lines()` and included in the computation.
     """
 
-    is_pyi: bool = False
+    mode: Mode
     previous_line: Optional[Line] = None
-    previous_after: int = 0
+    previous_block: Optional[LinesBlock] = None
     previous_defs: List[int] = field(default_factory=list)
+    semantic_leading_comment: Optional[LinesBlock] = None
 
-    def maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
+    def maybe_empty_lines(self, current_line: Line) -> LinesBlock:
         """Return the number of extra empty lines before and after the `current_line`.
 
         This is for separating `def`, `async def` and `class` with extra empty
         lines (two on module-level).
         """
         before, after = self._maybe_empty_lines(current_line)
+        previous_after = self.previous_block.after if self.previous_block else 0
         before = (
             # Black should not insert empty lines at the beginning
             # of the file
             0
             if self.previous_line is None
-            else before - self.previous_after
+            else before - previous_after
+        )
+        block = LinesBlock(
+            mode=self.mode,
+            previous_block=self.previous_block,
+            original_line=current_line,
+            before=before,
+            after=after,
         )
-        self.previous_after = after
+
+        # Maintain the semantic_leading_comment state.
+        if current_line.is_comment:
+            if self.previous_line is None or (
+                not self.previous_line.is_decorator
+                # `or before` means this comment already has an empty line before
+                and (not self.previous_line.is_comment or before)
+                and (self.semantic_leading_comment is None or before)
+            ):
+                self.semantic_leading_comment = block
+        # `or before` means this decorator already has an empty line before
+        elif not current_line.is_decorator or before:
+            self.semantic_leading_comment = None
+
         self.previous_line = current_line
-        return before, after
+        self.previous_block = block
+        return block
 
     def _maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
         max_allowed = 1
         if current_line.depth == 0:
-            max_allowed = 1 if self.is_pyi else 2
+            max_allowed = 1 if self.mode.is_pyi else 2
         if current_line.leaves:
             # Consume the first leaf's extra newlines.
             first_leaf = current_line.leaves[0]
@@ -495,7 +541,7 @@ class EmptyLineTracker:
             before = 0
         depth = current_line.depth
         while self.previous_defs and self.previous_defs[-1] >= depth:
-            if self.is_pyi:
+            if self.mode.is_pyi:
                 assert self.previous_line is not None
                 if depth and not current_line.is_def and self.previous_line.is_def:
                     # Empty lines between attributes and methods should be preserved.
@@ -545,11 +591,7 @@ class EmptyLineTracker:
         ):
             return before, 1
 
-        if (
-            Preview.remove_block_trailing_newline in current_line.mode
-            and self.previous_line
-            and self.previous_line.opens_block
-        ):
+        if self.previous_line and self.previous_line.opens_block:
             return 0, 0
         return before, 0
 
@@ -563,7 +605,7 @@ class EmptyLineTracker:
             return 0, 0
 
         if self.previous_line.is_decorator:
-            if self.is_pyi and current_line.is_stub_class:
+            if self.mode.is_pyi and current_line.is_stub_class:
                 # Insert an empty line after a decorated stub class
                 return 0, 1
 
@@ -574,14 +616,25 @@ class EmptyLineTracker:
         ):
             return 0, 0
 
+        comment_to_add_newlines: Optional[LinesBlock] = None
         if (
             self.previous_line.is_comment
             and self.previous_line.depth == current_line.depth
             and before == 0
         ):
-            return 0, 0
+            slc = self.semantic_leading_comment
+            if (
+                slc is not None
+                and slc.previous_block is not None
+                and not slc.previous_block.original_line.is_class
+                and not slc.previous_block.original_line.opens_block
+                and slc.before <= 1
+            ):
+                comment_to_add_newlines = slc
+            else:
+                return 0, 0
 
-        if self.is_pyi:
+        if self.mode.is_pyi:
             if current_line.is_class or self.previous_line.is_class:
                 if self.previous_line.depth < current_line.depth:
                     newlines = 0
@@ -609,6 +662,13 @@ class EmptyLineTracker:
                 newlines = 0
         else:
             newlines = 1 if current_line.depth else 2
+        if comment_to_add_newlines is not None:
+            previous_block = comment_to_add_newlines.previous_block
+            if previous_block is not None:
+                comment_to_add_newlines.before = (
+                    max(comment_to_add_newlines.before, newlines) - previous_block.after
+                )
+                newlines = 0
         return newlines, 0
 
 
@@ -644,18 +704,93 @@ def append_leaves(
             new_line.append(comment_leaf, preformatted=True)
 
 
-def is_line_short_enough(line: Line, *, line_length: int, line_str: str = "") -> bool:
-    """Return True if `line` is no longer than `line_length`.
-
+def is_line_short_enough(  # noqa: C901
+    line: Line, *, mode: Mode, line_str: str = ""
+) -> bool:
+    """For non-multiline strings, return True if `line` is no longer than `line_length`.
+    For multiline strings, looks at the context around `line` to determine
+    if it should be inlined or split up.
     Uses the provided `line_str` rendering, if any, otherwise computes a new one.
     """
     if not line_str:
         line_str = line_to_string(line)
-    return (
-        len(line_str) <= line_length
-        and "\n" not in line_str  # multiline strings
-        and not line.contains_standalone_comments()
-    )
+
+    if Preview.multiline_string_handling not in mode:
+        return (
+            len(line_str) <= mode.line_length
+            and "\n" not in line_str  # multiline strings
+            and not line.contains_standalone_comments()
+        )
+
+    if line.contains_standalone_comments():
+        return False
+    if "\n" not in line_str:
+        # No multiline strings (MLS) present
+        return len(line_str) <= mode.line_length
+
+    first, *_, last = line_str.split("\n")
+    if len(first) > mode.line_length or len(last) > mode.line_length:
+        return False
+
+    # Traverse the AST to examine the context of the multiline string (MLS),
+    # tracking aspects such as depth and comma existence,
+    # to determine whether to split the MLS or keep it together.
+    # Depth (which is based on the existing bracket_depth concept)
+    # is needed to determine nesting level of the MLS.
+    # Includes special case for trailing commas.
+    commas: List[int] = []  # tracks number of commas per depth level
+    multiline_string: Optional[Leaf] = None
+    # store the leaves that contain parts of the MLS
+    multiline_string_contexts: List[LN] = []
+
+    max_level_to_update = math.inf  # track the depth of the MLS
+    for i, leaf in enumerate(line.leaves):
+        if max_level_to_update == math.inf:
+            had_comma: Optional[int] = None
+            if leaf.bracket_depth + 1 > len(commas):
+                commas.append(0)
+            elif leaf.bracket_depth + 1 < len(commas):
+                had_comma = commas.pop()
+            if (
+                had_comma is not None
+                and multiline_string is not None
+                and multiline_string.bracket_depth == leaf.bracket_depth + 1
+            ):
+                # Have left the level with the MLS, stop tracking commas
+                max_level_to_update = leaf.bracket_depth
+                if had_comma > 0:
+                    # MLS was in parens with at least one comma - force split
+                    return False
+
+        if leaf.bracket_depth <= max_level_to_update and leaf.type == token.COMMA:
+            # Ignore non-nested trailing comma
+            # directly after MLS/MLS-containing expression
+            ignore_ctxs: List[Optional[LN]] = [None]
+            ignore_ctxs += multiline_string_contexts
+            if not (leaf.prev_sibling in ignore_ctxs and i == len(line.leaves) - 1):
+                commas[leaf.bracket_depth] += 1
+        if max_level_to_update != math.inf:
+            max_level_to_update = min(max_level_to_update, leaf.bracket_depth)
+
+        if is_multiline_string(leaf):
+            if len(multiline_string_contexts) > 0:
+                # >1 multiline string cannot fit on a single line - force split
+                return False
+            multiline_string = leaf
+            ctx: LN = leaf
+            # fetch the leaf components of the MLS in the AST
+            while str(ctx) in line_str:
+                multiline_string_contexts.append(ctx)
+                if ctx.parent is None:
+                    break
+                ctx = ctx.parent
+
+    # May not have a triple-quoted multiline string at all,
+    # in case of a regular string with embedded newlines and line continuations
+    if len(multiline_string_contexts) == 0:
+        return True
+
+    return all(val == 0 for val in commas)
 
 
 def can_be_split(line: Line) -> bool: