X-Git-Url: https://git.madduck.net/etc/vim.git/blobdiff_plain/1e0ec543ff3a7de715c8ee3359c8defb2c2c0e0d..2704dc796b49b4c5d348c92ba14ef1cfd908e7f1:/src/black/parsing.py diff --git a/src/black/parsing.py b/src/black/parsing.py index fc540ad..c37c12b 100644 --- a/src/black/parsing.py +++ b/src/black/parsing.py @@ -2,32 +2,46 @@ Parse Python code and perform AST validation. """ import ast +import platform import sys -from typing import Iterable, Iterator, List, Set, Union, Tuple +from typing import Any, Iterable, Iterator, List, Set, Tuple, Type, Union -# lib2to3 fork -from blib2to3.pytree import Node, Leaf -from blib2to3 import pygram, pytree +if sys.version_info < (3, 8): + from typing_extensions import Final +else: + from typing import Final + +from black.mode import 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.pytree import Leaf, Node -from black.mode import TargetVersion, Feature, supports_feature -from black.nodes import syms +ast3: Any + +_IS_PYPY = platform.python_implementation() == "PyPy" try: - from typed_ast import ast3, ast27 + from typed_ast import ast3 except ImportError: - if sys.version_info < (3, 8): + 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" - "`python3 -m pip install typed-ast`.", + ( + "The typed_ast package is required but not installed.\n" + "You can upgrade to Python 3.8+ or install typed_ast with\n" + "`python3 -m pip install typed-ast`." + ), file=sys.stderr, ) sys.exit(1) else: - ast3 = ast27 = ast + ast3 = ast + + +PY2_HINT: Final = "Python 2 support was removed in version 22.0." class InvalidInput(ValueError): @@ -42,35 +56,26 @@ def get_grammars(target_versions: Set[TargetVersion]) -> List[Grammar]: 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 2.7 with future print_function import - pygram.python_grammar_no_print_statement, - # Python 2.7 - pygram.python_grammar, - ] - - if all(version.is_python2() for version in target_versions): - # Python 2-only code, so try Python 2 grammars. - return [ - # Python 2.7 with future print_function import - pygram.python_grammar_no_print_statement, - # Python 2.7 - pygram.python_grammar, + # Python 3.10+ + pygram.python_grammar_soft_keywords, ] - # Python 3-compatible code, so only try Python 3 grammar. 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): - # Python 3.7+ + if not supports_feature( + target_versions, Feature.ASYNC_IDENTIFIERS + ) and not supports_feature(target_versions, Feature.PATTERN_MATCHING): + # Python 3.7-3.9 grammars.append( pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords ) 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 supports_feature(target_versions, Feature.PATTERN_MATCHING): + # 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 @@ -81,8 +86,10 @@ def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) - if not src_txt.endswith("\n"): src_txt += "\n" - for grammar in get_grammars(set(target_versions)): - drv = driver.Driver(grammar, pytree.convert) + grammars = get_grammars(set(target_versions)) + errors = {} + for grammar in grammars: + drv = driver.Driver(grammar) try: result = drv.parse_string(src_txt, True) break @@ -94,8 +101,29 @@ def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) - faulty_line = lines[lineno - 1] except IndexError: faulty_line = "" - 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] + errors[grammar.version] = InvalidInput( + f"Cannot parse: {lineno}:{column}: {te.args[0]}" + ) + else: + # 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 + ): + original_msg = exc.args[0] + msg = f"{original_msg}\n{PY2_HINT}" + raise InvalidInput(msg) from None + raise exc from None if isinstance(result, Leaf): @@ -103,6 +131,16 @@ def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) - return result +def matches_grammar(src_txt: str, grammar: Grammar) -> bool: + drv = driver.Driver(grammar) + try: + drv.parse_string(src_txt, True) + except (ParseError, TokenError, IndentationError): + return False + else: + return True + + def lib2to3_unparse(node: Node) -> str: """Given a lib2to3 node, return its string representation.""" code = str(node) @@ -111,25 +149,29 @@ def lib2to3_unparse(node: Node) -> str: def parse_single_version( src: str, version: Tuple[int, int] -) -> Union[ast.AST, ast3.AST, ast27.AST]: +) -> Union[ast.AST, ast3.AST]: filename = "" - # 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,): - return ast.parse(src, filename, feature_version=version) - elif version >= (3,): + return ast.parse(src, filename, feature_version=version, type_comments=True) + + 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=True) + else: + return ast3.parse(src, filename) + else: + # Typed-ast is guaranteed to be used here and automatically tracks type + # comments separately. return ast3.parse(src, filename, feature_version=version[1]) - elif version == (2, 7): - return ast27.parse(src) - raise AssertionError("INTERNAL ERROR: Tried parsing unsupported Python version!") -def parse_ast(src: str) -> Union[ast.AST, ast3.AST, ast27.AST]: +def parse_ast(src: str) -> Union[ast.AST, ast3.AST]: # TODO: support Python 4+ ;) versions = [(3, minor) for minor in range(3, sys.version_info[1] + 1)] - if ast27.__name__ != "ast": - versions.append((2, 7)) - first_error = "" for version in sorted(versions, reverse=True): try: @@ -141,25 +183,39 @@ def parse_ast(src: str) -> Union[ast.AST, ast3.AST, ast27.AST]: raise SyntaxError(first_error) -def stringify_ast( - node: Union[ast.AST, ast3.AST, ast27.AST], depth: int = 0 -) -> Iterator[str]: +ast3_AST: Final[Type[ast3.AST]] = ast3.AST + + +def _normalize(lineend: str, value: str) -> str: + # To normalize, we strip any leading and trailing space from + # each line... + stripped: List[str] = [i.strip() for i in value.splitlines()] + normalized = lineend.join(stripped) + # ...and remove any blank lines at the beginning and end of + # the whole string + return normalized.strip() + + +def stringify_ast(node: Union[ast.AST, ast3.AST], depth: int = 0) -> Iterator[str]: """Simple visitor generating strings to compare ASTs by content.""" node = fixup_ast_constants(node) yield f"{' ' * depth}{node.__class__.__name__}(" + type_ignore_classes: Tuple[Type[Any], ...] for field in sorted(node._fields): # noqa: F402 - # TypeIgnore has only one field 'lineno' which breaks this comparison - type_ignore_classes = (ast3.TypeIgnore, ast27.TypeIgnore) - if sys.version_info >= (3, 8): - type_ignore_classes += (ast.TypeIgnore,) - if isinstance(node, type_ignore_classes): - break + # TypeIgnore will not be present using pypy < 3.8, so need for this + if not (_IS_PYPY and sys.version_info < (3, 8)): + # TypeIgnore has only one field 'lineno' which breaks this comparison + type_ignore_classes = (ast3.TypeIgnore,) + if sys.version_info >= (3, 8): + type_ignore_classes += (ast.TypeIgnore,) + if isinstance(node, type_ignore_classes): + break try: - value = getattr(node, field) + value: object = getattr(node, field) except AttributeError: continue @@ -171,38 +227,35 @@ def stringify_ast( # parentheses and they change the AST. if ( field == "targets" - and isinstance(node, (ast.Delete, ast3.Delete, ast27.Delete)) - and isinstance(item, (ast.Tuple, ast3.Tuple, ast27.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, ast27.AST)): + elif isinstance(item, (ast.AST, ast3.AST)): yield from stringify_ast(item, depth + 2) - elif isinstance(value, (ast.AST, ast3.AST, ast27.AST)): + # Note that we are referencing the typed-ast ASTs via global variables and not + # direct module attribute accesses because that breaks mypyc. It's probably + # something to do with the ast3 variables being marked as Any leading + # mypy to think this branch is always taken, leaving the rest of the code + # unanalyzed. Tighting up the types for the typed-ast AST types avoids the + # mypyc crash. + elif isinstance(value, (ast.AST, ast3_AST)): 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. - # Note that when formatting Python 2 code, at least with Windows - # line-endings, docstrings can end up here as bytes instead of - # str so make sure that we handle both cases. if ( isinstance(node, ast.Constant) and field == "value" - and isinstance(value, (str, bytes)) + and isinstance(value, str) ): - lineend = "\n" if isinstance(value, str) else b"\n" - # To normalize, we strip any leading and trailing space from - # each line... - stripped = [line.strip() for line in value.splitlines()] - normalized = lineend.join(stripped) # type: ignore[attr-defined] - # ...and remove any blank lines at the beginning and end of - # the whole string - normalized = normalized.strip() + normalized = _normalize("\n", value) else: normalized = value yield f"{' ' * (depth+2)}{normalized!r}, # {value.__class__.__name__}" @@ -210,14 +263,12 @@ def stringify_ast( yield f"{' ' * depth}) # /{node.__class__.__name__}" -def fixup_ast_constants( - node: Union[ast.AST, ast3.AST, ast27.AST] -) -> Union[ast.AST, ast3.AST, ast27.AST]: +def fixup_ast_constants(node: Union[ast.AST, ast3.AST]) -> Union[ast.AST, ast3.AST]: """Map ast nodes deprecated in 3.8 to Constant.""" - if isinstance(node, (ast.Str, ast3.Str, ast27.Str, ast.Bytes, ast3.Bytes)): + if isinstance(node, (ast.Str, ast3.Str, ast.Bytes, ast3.Bytes)): return ast.Constant(value=node.s) - if isinstance(node, (ast.Num, ast3.Num, ast27.Num)): + if isinstance(node, (ast.Num, ast3.Num)): return ast.Constant(value=node.n) if isinstance(node, (ast.NameConstant, ast3.NameConstant)):