]> git.madduck.net Git - etc/vim.git/blobdiff - black.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:

Treat comments less magically
[etc/vim.git] / black.py
index 6bfef500d6d44ea79d8d1eca2030127c7433339d..81efa59a50669070b4005e4f4c5a73916ab3c43a 100644 (file)
--- a/black.py
+++ b/black.py
@@ -1,4 +1,5 @@
 #!/usr/bin/env python3
+
 import asyncio
 from asyncio.base_events import BaseEventLoop
 from concurrent.futures import Executor, ProcessPoolExecutor
@@ -102,7 +103,9 @@ def main(
         report = Report()
         try:
             if not p.is_file() and str(p) == '-':
-                changed = format_stdin_to_stdout(line_length=line_length, fast=fast)
+                changed = format_stdin_to_stdout(
+                    line_length=line_length, fast=fast, write_back=not check
+                )
             else:
                 changed = format_file_in_place(
                     p, line_length=line_length, fast=fast, write_back=not check
@@ -178,7 +181,9 @@ def format_file_in_place(
     return True
 
 
-def format_stdin_to_stdout(line_length: int, fast: bool) -> bool:
+def format_stdin_to_stdout(
+    line_length: int, fast: bool, write_back: bool = False
+) -> bool:
     """Format file on stdin and pipe output to stdout. Return True if changed."""
     contents = sys.stdin.read()
     try:
@@ -189,7 +194,8 @@ def format_stdin_to_stdout(line_length: int, fast: bool) -> bool:
         return False
 
     finally:
-        sys.stdout.write(contents)
+        if write_back:
+            sys.stdout.write(contents)
 
 
 def format_file_contents(
@@ -213,7 +219,6 @@ def format_str(src_contents: str, line_length: int) -> FileContent:
     """Reformats a string and returns new contents."""
     src_node = lib2to3_parse(src_contents)
     dst_contents = ""
-    comments: List[Line] = []
     lines = LineGenerator()
     elt = EmptyLineTracker()
     py36 = is_python36(src_node)
@@ -225,21 +230,8 @@ def format_str(src_contents: str, line_length: int) -> FileContent:
         before, after = elt.maybe_empty_lines(current_line)
         for _ in range(before):
             dst_contents += str(empty_line)
-        if not current_line.is_comment:
-            for comment in comments:
-                dst_contents += str(comment)
-            comments = []
-            for line in split_line(current_line, line_length=line_length, py36=py36):
-                dst_contents += str(line)
-        else:
-            comments.append(current_line)
-    if comments:
-        if elt.previous_defs:
-            # Separate postscriptum comments from the last module-level def.
-            dst_contents += str(empty_line)
-            dst_contents += str(empty_line)
-        for comment in comments:
-            dst_contents += str(comment)
+        for line in split_line(current_line, line_length=line_length, py36=py36):
+            dst_contents += str(line)
     return dst_contents
 
 
@@ -486,8 +478,7 @@ class Line:
         return (
             (first_leaf.type == token.NAME and first_leaf.value == 'def')
             or (
-                first_leaf.type == token.NAME
-                and first_leaf.value == 'async'
+                first_leaf.type == token.ASYNC
                 and second_leaf is not None
                 and second_leaf.type == token.NAME
                 and second_leaf.value == 'def'
@@ -658,10 +649,6 @@ class EmptyLineTracker:
         (two on module-level), as well as providing an extra empty line after flow
         control keywords to make them more prominent.
         """
-        if current_line.is_comment:
-            # Don't count standalone comments towards previous empty lines.
-            return 0, 0
-
         before, after = self._maybe_empty_lines(current_line)
         before -= self.previous_after
         self.previous_after = after
@@ -669,10 +656,14 @@ class EmptyLineTracker:
         return before, after
 
     def _maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
+        max_allowed = 1
+        if current_line.is_comment and current_line.depth == 0:
+            max_allowed = 2
         if current_line.leaves:
             # Consume the first leaf's extra newlines.
             first_leaf = current_line.leaves[0]
-            before = int('\n' in first_leaf.prefix)
+            before = first_leaf.prefix.count('\n')
+            before = min(before, max(before, max_allowed))
             first_leaf.prefix = ''
         else:
             before = 0
@@ -726,7 +717,6 @@ class LineGenerator(Visitor[Line]):
     in ways that will no longer stringify to valid Python code on the tree.
     """
     current_line: Line = Factory(Line)
-    standalone_comments: List[Leaf] = Factory(list)
 
     def line(self, indent: int = 0) -> Iterator[Line]:
         """Generate a line.
@@ -746,8 +736,9 @@ class LineGenerator(Visitor[Line]):
 
     def visit_default(self, node: LN) -> Iterator[Line]:
         if isinstance(node, Leaf):
+            any_open_brackets = self.current_line.bracket_tracker.any_open_brackets()
             for comment in generate_comments(node):
-                if self.current_line.bracket_tracker.any_open_brackets():
+                if any_open_brackets:
                     # any comment within brackets is subject to splitting
                     self.current_line.append(comment)
                 elif comment.type == token.COMMENT:
@@ -756,33 +747,22 @@ class LineGenerator(Visitor[Line]):
                     yield from self.line()
 
                 else:
-                    # regular standalone comment, to be processed later (see
-                    # docstring in `generate_comments()`
-                    self.standalone_comments.append(comment)
-            normalize_prefix(node)
-            if node.type not in WHITESPACE:
-                for comment in self.standalone_comments:
+                    # regular standalone comment
                     yield from self.line()
 
                     self.current_line.append(comment)
                     yield from self.line()
 
-                self.standalone_comments = []
+            normalize_prefix(node, inside_brackets=any_open_brackets)
+            if node.type not in WHITESPACE:
                 self.current_line.append(node)
         yield from super().visit_default(node)
 
-    def visit_suite(self, node: Node) -> Iterator[Line]:
-        """Body of a statement after a colon."""
-        children = iter(node.children)
-        # Process newline before indenting.  It might contain an inline
-        # comment that should go right after the colon.
-        newline = next(children)
-        yield from self.visit(newline)
+    def visit_INDENT(self, node: Node) -> Iterator[Line]:
         yield from self.line(+1)
+        yield from self.visit_default(node)
 
-        for child in children:
-            yield from self.visit(child)
-
+    def visit_DEDENT(self, node: Node) -> Iterator[Line]:
         yield from self.line(-1)
 
     def visit_stmt(self, node: Node, keywords: Set[str]) -> Iterator[Line]:
@@ -816,7 +796,7 @@ class LineGenerator(Visitor[Line]):
         for child in children:
             yield from self.visit(child)
 
-            if child.type == token.NAME and child.value == 'async':  # type: ignore
+            if child.type == token.ASYNC:
                 break
 
         internal_stmt = next(children)
@@ -872,7 +852,7 @@ def whitespace(leaf: Leaf) -> str:  # noqa C901
         return DOUBLESPACE
 
     assert p is not None, f"INTERNAL ERROR: hand-made leaf without parent: {leaf!r}"
-    if t == token.COLON and p.type != syms.subscript:
+    if t == token.COLON and p.type not in {syms.subscript, syms.subscriptlist}:
         return NO
 
     prev = leaf.prev_sibling
@@ -905,7 +885,7 @@ def whitespace(leaf: Leaf) -> str:  # noqa C901
                 return NO
 
         elif prevp.type == token.COLON:
-            if prevp.parent and prevp.parent.type == syms.subscript:
+            if prevp.parent and prevp.parent.type in {syms.subscript, syms.sliceop}:
                 return NO
 
         elif prevp.parent and prevp.parent.type in {syms.factor, syms.star_expr}:
@@ -1133,30 +1113,40 @@ def generate_comments(leaf: Leaf) -> Iterator[Leaf]:
     Inline comments are emitted as regular token.COMMENT leaves.  Standalone
     are emitted with a fake STANDALONE_COMMENT token identifier.
     """
-    if not leaf.prefix:
+    p = leaf.prefix
+    if not p:
         return
 
-    if '#' not in leaf.prefix:
+    if '#' not in p:
         return
 
-    before_comment, content = leaf.prefix.split('#', 1)
-    content = content.rstrip()
-    if content and (content[0] not in {' ', '!', '#'}):
-        content = ' ' + content
-    is_standalone_comment = (
-        '\n' in before_comment or '\n' in content or leaf.type == token.ENDMARKER
-    )
-    if not is_standalone_comment:
-        # simple trailing comment
-        yield Leaf(token.COMMENT, value='#' + content)
-        return
-
-    for line in ('#' + content).split('\n'):
+    nlines = 0
+    for index, line in enumerate(p.split('\n')):
         line = line.lstrip()
+        if not line:
+            nlines += 1
         if not line.startswith('#'):
             continue
 
-        yield Leaf(STANDALONE_COMMENT, line)
+        if index == 0 and leaf.type != token.ENDMARKER:
+            comment_type = token.COMMENT  # simple trailing comment
+        else:
+            comment_type = STANDALONE_COMMENT
+        yield Leaf(comment_type, make_comment(line), prefix='\n' * nlines)
+
+        nlines = 0
+
+
+def make_comment(content: str) -> str:
+    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 split_line(
@@ -1239,7 +1229,7 @@ def left_hand_split(line: Line, py36: bool = False) -> Iterator[Line]:
                 current_leaves = body_leaves
     # Since body is a new indent level, remove spurious leading whitespace.
     if body_leaves:
-        normalize_prefix(body_leaves[0])
+        normalize_prefix(body_leaves[0], inside_brackets=True)
     # Build the new lines.
     for result, leaves in (
         (head, head_leaves), (body, body_leaves), (tail, tail_leaves)
@@ -1279,7 +1269,7 @@ def right_hand_split(line: Line, py36: bool = False) -> Iterator[Line]:
     head_leaves.reverse()
     # Since body is a new indent level, remove spurious leading whitespace.
     if body_leaves:
-        normalize_prefix(body_leaves[0])
+        normalize_prefix(body_leaves[0], inside_brackets=True)
     # Build the new lines.
     for result, leaves in (
         (head, head_leaves), (body, body_leaves), (tail, tail_leaves)
@@ -1343,7 +1333,7 @@ def delimiter_split(line: Line, py36: bool = False) -> Iterator[Line]:
             trailing_comma_safe = trailing_comma_safe and py36
         leaf_priority = delimiters.get(id(leaf))
         if leaf_priority == delimiter_priority:
-            normalize_prefix(current_line.leaves[0])
+            normalize_prefix(current_line.leaves[0], inside_brackets=True)
             yield current_line
 
             current_line = Line(depth=line.depth, inside_brackets=line.inside_brackets)
@@ -1354,7 +1344,7 @@ def delimiter_split(line: Line, py36: bool = False) -> Iterator[Line]:
             and trailing_comma_safe
         ):
             current_line.append(Leaf(token.COMMA, ','))
-        normalize_prefix(current_line.leaves[0])
+        normalize_prefix(current_line.leaves[0], inside_brackets=True)
         yield current_line
 
 
@@ -1372,13 +1362,20 @@ def is_import(leaf: Leaf) -> bool:
     )
 
 
-def normalize_prefix(leaf: Leaf) -> None:
-    """Leave existing extra newlines for imports.  Remove everything else."""
-    if is_import(leaf):
-        spl = leaf.prefix.split('#', 1)
-        nl_count = spl[0].count('\n')
-        leaf.prefix = '\n' * nl_count
-        return
+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 = ''