]> 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:

Fix numpy-style array indexing for real
[etc/vim.git] / black.py
index 10346a9143c4e490d7fb6a8f314cc00a94e9664a..203fbfa9055be595a358e94bd1c16b9a3911f23f 100644 (file)
--- a/black.py
+++ b/black.py
@@ -74,7 +74,9 @@ class CannotSplit(Exception):
 @click.argument(
     'src',
     nargs=-1,
-    type=click.Path(exists=True, file_okay=True, dir_okay=True, readable=True),
+    type=click.Path(
+        exists=True, file_okay=True, dir_okay=True, readable=True, allow_dash=True
+    ),
 )
 @click.pass_context
 def main(
@@ -89,6 +91,8 @@ def main(
         elif p.is_file():
             # if a file was explicitly given, we don't care about its extension
             sources.append(p)
+        elif s == '-':
+            sources.append(Path('-'))
         else:
             err(f'invalid path: {s}')
     if len(sources) == 0:
@@ -97,9 +101,12 @@ def main(
         p = sources[0]
         report = Report()
         try:
-            changed = format_file_in_place(
-                p, line_length=line_length, fast=fast, write_back=not check
-            )
+            if not p.is_file() and str(p) == '-':
+                changed = format_stdin_to_stdout(line_length=line_length, fast=fast)
+            else:
+                changed = format_file_in_place(
+                    p, line_length=line_length, fast=fast, write_back=not check
+                )
             report.done(p, changed)
         except Exception as exc:
             report.failed(p, str(exc))
@@ -156,34 +163,50 @@ def format_file_in_place(
     src: Path, line_length: int, fast: bool, write_back: bool = False
 ) -> bool:
     """Format the file and rewrite if changed. Return True if changed."""
+    with tokenize.open(src) as src_buffer:
+        src_contents = src_buffer.read()
     try:
-        contents, encoding = format_file(src, line_length=line_length, fast=fast)
+        contents = format_file_contents(
+            src_contents, line_length=line_length, fast=fast
+        )
     except NothingChanged:
         return False
 
     if write_back:
-        with open(src, "w", encoding=encoding) as f:
+        with open(src, "w", encoding=src_buffer.encoding) as f:
             f.write(contents)
     return True
 
 
-def format_file(
-    src: Path, line_length: int, fast: bool
-) -> Tuple[FileContent, Encoding]:
+def format_stdin_to_stdout(line_length: int, fast: bool) -> bool:
+    """Format file on stdin and pipe output to stdout. Return True if changed."""
+    contents = sys.stdin.read()
+    try:
+        contents = format_file_contents(contents, line_length=line_length, fast=fast)
+        return True
+
+    except NothingChanged:
+        return False
+
+    finally:
+        sys.stdout.write(contents)
+
+
+def format_file_contents(
+    src_contents: str, line_length: int, fast: bool
+) -> FileContent:
     """Reformats a file and returns its contents and encoding."""
-    with tokenize.open(src) as src_buffer:
-        src_contents = src_buffer.read()
     if src_contents.strip() == '':
-        raise NothingChanged(src)
+        raise NothingChanged
 
     dst_contents = format_str(src_contents, line_length=line_length)
     if src_contents == dst_contents:
-        raise NothingChanged(src)
+        raise NothingChanged
 
     if not fast:
         assert_equivalent(src_contents, dst_contents)
         assert_stable(src_contents, dst_contents, line_length=line_length)
-    return dst_contents, src_buffer.encoding
+    return dst_contents
 
 
 def format_str(src_contents: str, line_length: int) -> FileContent:
@@ -210,8 +233,13 @@ def format_str(src_contents: str, line_length: int) -> FileContent:
                 dst_contents += str(line)
         else:
             comments.append(current_line)
-    for comment in comments:
-        dst_contents += str(comment)
+    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)
     return dst_contents
 
 
@@ -458,8 +486,7 @@ 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'
@@ -615,7 +642,9 @@ class EmptyLineTracker:
     """Provides a stateful method that returns the number of potential extra
     empty lines needed before and after the currently processed line.
 
-    Note: this tracker works on lines that haven't been split yet.
+    Note: this tracker works on lines that haven't been split yet.  It assumes
+    the prefix of the first leaf consists of optional newlines.  Those newlines
+    are consumed by `maybe_empty_lines()` and included in the computation.
     """
     previous_line: Optional[Line] = None
     previous_after: int = 0
@@ -628,17 +657,28 @@ 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
         self.previous_line = current_line
         return before, after
 
     def _maybe_empty_lines(self, current_line: Line) -> Tuple[int, int]:
-        before = 0
+        if current_line.leaves:
+            # Consume the first leaf's extra newlines.
+            first_leaf = current_line.leaves[0]
+            before = int('\n' in first_leaf.prefix)
+            first_leaf.prefix = ''
+        else:
+            before = 0
         depth = current_line.depth
         while self.previous_defs and self.previous_defs[-1] >= depth:
             self.previous_defs.pop()
-            before = (1 if depth else 2) - self.previous_after
+            before = 1 if depth else 2
         is_decorator = current_line.is_decorator
         if is_decorator or current_line.is_def or current_line.is_class:
             if not is_decorator:
@@ -654,7 +694,6 @@ class EmptyLineTracker:
             newlines = 2
             if current_line.depth:
                 newlines -= 1
-            newlines -= self.previous_after
             return newlines, 0
 
         if current_line.is_flow_control:
@@ -706,8 +745,9 @@ 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:
@@ -719,7 +759,7 @@ class LineGenerator(Visitor[Line]):
                     # regular standalone comment, to be processed later (see
                     # docstring in `generate_comments()`
                     self.standalone_comments.append(comment)
-            normalize_prefix(node)
+            normalize_prefix(node, inside_brackets=any_open_brackets)
             if node.type not in WHITESPACE:
                 for comment in self.standalone_comments:
                     yield from self.line()
@@ -776,7 +816,7 @@ 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)
@@ -814,7 +854,7 @@ BRACKET = {token.LPAR: token.RPAR, token.LSQB: token.RSQB, token.LBRACE: token.R
 OPENING_BRACKETS = set(BRACKET.keys())
 CLOSING_BRACKETS = set(BRACKET.values())
 BRACKETS = OPENING_BRACKETS | CLOSING_BRACKETS
-ALWAYS_NO_SPACE = CLOSING_BRACKETS | {token.COMMA, token.COLON, STANDALONE_COMMENT}
+ALWAYS_NO_SPACE = CLOSING_BRACKETS | {token.COMMA, STANDALONE_COMMENT}
 
 
 def whitespace(leaf: Leaf) -> str:  # noqa C901
@@ -832,12 +872,18 @@ 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 not in {syms.subscript, syms.subscriptlist}:
+        return NO
+
     prev = leaf.prev_sibling
     if not prev:
         prevp = preceding_leaf(p)
         if not prevp or prevp.type in OPENING_BRACKETS:
             return NO
 
+        if t == token.COLON:
+            return SPACE if prevp.type == token.COMMA else NO
+
         if prevp.type == token.EQUAL:
             if prevp.parent and prevp.parent.type in {
                 syms.typedargslist,
@@ -859,7 +905,7 @@ 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}:
@@ -966,7 +1012,7 @@ def whitespace(leaf: Leaf) -> str:  # noqa C901
 
             return NO
 
-        elif prev.type == token.COLON:
+        else:
             return NO
 
     elif p.type == syms.atom:
@@ -1098,7 +1144,7 @@ def generate_comments(leaf: Leaf) -> Iterator[Leaf]:
     if content and (content[0] not in {' ', '!', '#'}):
         content = ' ' + content
     is_standalone_comment = (
-        '\n' in before_comment or '\n' in content or leaf.type == token.DEDENT
+        '\n' in before_comment or '\n' in content or leaf.type == token.ENDMARKER
     )
     if not is_standalone_comment:
         # simple trailing comment
@@ -1193,7 +1239,7 @@ 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)
@@ -1233,7 +1279,7 @@ 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)
@@ -1297,7 +1343,7 @@ 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)
@@ -1308,7 +1354,7 @@ 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
 
 
@@ -1326,16 +1372,18 @@ 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):
+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('#', 1)
-        nl_count = spl[0].count('\n')
-        if len(spl) > 1:
-            # Skip one newline since it was for a standalone comment.
-            nl_count -= 1
-        leaf.prefix = '\n' * nl_count
-        return
+        if '\\' not in spl[0]:
+            nl_count = spl[0].count('\n')
+            leaf.prefix = '\n' * nl_count
+            return
 
     leaf.prefix = ''