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

Max line length with bugbear (#3731)
[etc/vim.git] / src / black / parsing.py
index 13fa67ee84dabb07f183f8bee0bd8b8742fac490..70ed99c15494e9fa438309d8b0e1d99626b5080a 100644 (file)
@@ -11,16 +11,14 @@ if sys.version_info < (3, 8):
 else:
     from typing import Final
 
 else:
     from typing import Final
 
-# lib2to3 fork
-from blib2to3.pytree import Node, Leaf
+from black.mode import VERSION_TO_FEATURES, Feature, TargetVersion, supports_feature
+from black.nodes import syms
 from blib2to3 import pygram
 from blib2to3.pgen2 import driver
 from blib2to3.pgen2.grammar import Grammar
 from blib2to3.pgen2.parse import ParseError
 from blib2to3.pgen2.tokenize import TokenError
 from blib2to3 import pygram
 from blib2to3.pgen2 import driver
 from blib2to3.pgen2.grammar import Grammar
 from blib2to3.pgen2.parse import ParseError
 from blib2to3.pgen2.tokenize import TokenError
-
-from black.mode import TargetVersion, Feature, supports_feature
-from black.nodes import syms
+from blib2to3.pytree import Leaf, Node
 
 ast3: Any
 
 
 ast3: Any
 
@@ -29,8 +27,7 @@ _IS_PYPY = platform.python_implementation() == "PyPy"
 try:
     from typed_ast import ast3
 except ImportError:
 try:
     from typed_ast import ast3
 except ImportError:
-    # Either our python version is too low, or we're on pypy
-    if sys.version_info < (3, 7) or (sys.version_info < (3, 8) and not _IS_PYPY):
+    if sys.version_info < (3, 8) and not _IS_PYPY:
         print(
             "The typed_ast package is required but not installed.\n"
             "You can upgrade to Python 3.8+ or install typed_ast with\n"
         print(
             "The typed_ast package is required but not installed.\n"
             "You can upgrade to Python 3.8+ or install typed_ast with\n"
@@ -42,7 +39,6 @@ except ImportError:
         ast3 = ast
 
 
         ast3 = ast
 
 
-PY310_HINT: Final = "Consider using --target-version py310 to parse Python 3.10 code."
 PY2_HINT: Final = "Python 2 support was removed in version 22.0."
 
 
 PY2_HINT: Final = "Python 2 support was removed in version 22.0."
 
 
@@ -54,16 +50,15 @@ def get_grammars(target_versions: Set[TargetVersion]) -> List[Grammar]:
     if not target_versions:
         # No target_version specified, so try all grammars.
         return [
     if not target_versions:
         # No target_version specified, so try all grammars.
         return [
-            # Python 3.7+
+            # Python 3.7-3.9
             pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords,
             # Python 3.0-3.6
             pygram.python_grammar_no_print_statement_no_exec_statement,
             pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords,
             # Python 3.0-3.6
             pygram.python_grammar_no_print_statement_no_exec_statement,
+            # Python 3.10+
+            pygram.python_grammar_soft_keywords,
         ]
 
     grammars = []
         ]
 
     grammars = []
-    if supports_feature(target_versions, Feature.PATTERN_MATCHING):
-        # Python 3.10+
-        grammars.append(pygram.python_grammar_soft_keywords)
     # If we have to parse both, try to parse async as a keyword first
     if not supports_feature(
         target_versions, Feature.ASYNC_IDENTIFIERS
     # If we have to parse both, try to parse async as a keyword first
     if not supports_feature(
         target_versions, Feature.ASYNC_IDENTIFIERS
@@ -75,6 +70,10 @@ def get_grammars(target_versions: Set[TargetVersion]) -> List[Grammar]:
     if not supports_feature(target_versions, Feature.ASYNC_KEYWORDS):
         # Python 3.0-3.6
         grammars.append(pygram.python_grammar_no_print_statement_no_exec_statement)
     if not supports_feature(target_versions, Feature.ASYNC_KEYWORDS):
         # Python 3.0-3.6
         grammars.append(pygram.python_grammar_no_print_statement_no_exec_statement)
+    if any(Feature.PATTERN_MATCHING in VERSION_TO_FEATURES[v] for v in target_versions):
+        # Python 3.10+
+        grammars.append(pygram.python_grammar_soft_keywords)
+
     # At least one of the above branches must have been taken, because every Python
     # version has exactly one of the two 'ASYNC_*' flags
     return grammars
     # At least one of the above branches must have been taken, because every Python
     # version has exactly one of the two 'ASYNC_*' flags
     return grammars
@@ -86,6 +85,7 @@ def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) -
         src_txt += "\n"
 
     grammars = get_grammars(set(target_versions))
         src_txt += "\n"
 
     grammars = get_grammars(set(target_versions))
+    errors = {}
     for grammar in grammars:
         drv = driver.Driver(grammar)
         try:
     for grammar in grammars:
         drv = driver.Driver(grammar)
         try:
@@ -99,20 +99,21 @@ def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) -
                 faulty_line = lines[lineno - 1]
             except IndexError:
                 faulty_line = "<line number missing in source>"
                 faulty_line = lines[lineno - 1]
             except IndexError:
                 faulty_line = "<line number missing in source>"
-            exc = InvalidInput(f"Cannot parse: {lineno}:{column}: {faulty_line}")
+            errors[grammar.version] = InvalidInput(
+                f"Cannot parse: {lineno}:{column}: {faulty_line}"
+            )
 
         except TokenError as te:
             # In edge cases these are raised; and typically don't have a "faulty_line".
             lineno, column = te.args[1]
 
         except TokenError as te:
             # In edge cases these are raised; and typically don't have a "faulty_line".
             lineno, column = te.args[1]
-            exc = InvalidInput(f"Cannot parse: {lineno}:{column}: {te.args[0]}")
+            errors[grammar.version] = InvalidInput(
+                f"Cannot parse: {lineno}:{column}: {te.args[0]}"
+            )
 
     else:
 
     else:
-        if pygram.python_grammar_soft_keywords not in grammars and matches_grammar(
-            src_txt, pygram.python_grammar_soft_keywords
-        ):
-            original_msg = exc.args[0]
-            msg = f"{original_msg}\n{PY310_HINT}"
-            raise InvalidInput(msg) from None
+        # Choose the latest version when raising the actual parsing error.
+        assert len(errors) >= 1
+        exc = errors[max(errors)]
 
         if matches_grammar(src_txt, pygram.python_grammar) or matches_grammar(
             src_txt, pygram.python_grammar_no_print_statement
 
         if matches_grammar(src_txt, pygram.python_grammar) or matches_grammar(
             src_txt, pygram.python_grammar_no_print_statement
@@ -145,18 +146,29 @@ def lib2to3_unparse(node: Node) -> str:
 
 
 def parse_single_version(
 
 
 def parse_single_version(
-    src: str, version: Tuple[int, int]
+    src: str, version: Tuple[int, int], *, type_comments: bool
 ) -> Union[ast.AST, ast3.AST]:
     filename = "<unknown>"
 ) -> Union[ast.AST, ast3.AST]:
     filename = "<unknown>"
-    # typed_ast is needed because of feature version limitations in the builtin ast
+    # typed-ast is needed because of feature version limitations in the builtin ast 3.8>
     if sys.version_info >= (3, 8) and version >= (3,):
     if sys.version_info >= (3, 8) and version >= (3,):
-        return ast.parse(src, filename, feature_version=version)
-    elif version >= (3,):
-        if _IS_PYPY:
-            return ast3.parse(src, filename)
+        return ast.parse(
+            src, filename, feature_version=version, type_comments=type_comments
+        )
+
+    if _IS_PYPY:
+        # PyPy 3.7 doesn't support type comment tracking which is not ideal, but there's
+        # not much we can do as typed-ast won't work either.
+        if sys.version_info >= (3, 8):
+            return ast3.parse(src, filename, type_comments=type_comments)
         else:
         else:
+            return ast3.parse(src, filename)
+    else:
+        if type_comments:
+            # Typed-ast is guaranteed to be used here and automatically tracks type
+            # comments separately.
             return ast3.parse(src, filename, feature_version=version[1])
             return ast3.parse(src, filename, feature_version=version[1])
-    raise AssertionError("INTERNAL ERROR: Tried parsing unsupported Python version!")
+        else:
+            return ast.parse(src, filename)
 
 
 def parse_ast(src: str) -> Union[ast.AST, ast3.AST]:
 
 
 def parse_ast(src: str) -> Union[ast.AST, ast3.AST]:
@@ -166,11 +178,18 @@ def parse_ast(src: str) -> Union[ast.AST, ast3.AST]:
     first_error = ""
     for version in sorted(versions, reverse=True):
         try:
     first_error = ""
     for version in sorted(versions, reverse=True):
         try:
-            return parse_single_version(src, version)
+            return parse_single_version(src, version, type_comments=True)
         except SyntaxError as e:
             if not first_error:
                 first_error = str(e)
 
         except SyntaxError as e:
             if not first_error:
                 first_error = str(e)
 
+    # Try to parse without type comments
+    for version in sorted(versions, reverse=True):
+        try:
+            return parse_single_version(src, version, type_comments=False)
+        except SyntaxError:
+            pass
+
     raise SyntaxError(first_error)
 
 
     raise SyntaxError(first_error)
 
 
@@ -206,7 +225,7 @@ def stringify_ast(node: Union[ast.AST, ast3.AST], depth: int = 0) -> Iterator[st
                 break
 
         try:
                 break
 
         try:
-            value = getattr(node, field)
+            value: object = getattr(node, field)
         except AttributeError:
             continue
 
         except AttributeError:
             continue
 
@@ -221,8 +240,8 @@ def stringify_ast(node: Union[ast.AST, ast3.AST], depth: int = 0) -> Iterator[st
                     and isinstance(node, (ast.Delete, ast3.Delete))
                     and isinstance(item, (ast.Tuple, ast3.Tuple))
                 ):
                     and isinstance(node, (ast.Delete, ast3.Delete))
                     and isinstance(item, (ast.Tuple, ast3.Tuple))
                 ):
-                    for item in item.elts:
-                        yield from stringify_ast(item, depth + 2)
+                    for elt in item.elts:
+                        yield from stringify_ast(elt, depth + 2)
 
                 elif isinstance(item, (ast.AST, ast3.AST)):
                     yield from stringify_ast(item, depth + 2)
 
                 elif isinstance(item, (ast.AST, ast3.AST)):
                     yield from stringify_ast(item, depth + 2)
@@ -237,6 +256,7 @@ def stringify_ast(node: Union[ast.AST, ast3.AST], depth: int = 0) -> Iterator[st
             yield from stringify_ast(value, depth + 2)
 
         else:
             yield from stringify_ast(value, depth + 2)
 
         else:
+            normalized: object
             # Constant strings may be indented across newlines, if they are
             # docstrings; fold spaces after newlines when comparing. Similarly,
             # trailing and leading space may be removed.
             # Constant strings may be indented across newlines, if they are
             # docstrings; fold spaces after newlines when comparing. Similarly,
             # trailing and leading space may be removed.