from attr import dataclass, evolve, Factory
import click
import toml
+from typed_ast import ast3, ast27
# lib2to3 fork
from blib2to3.pytree import Node, Leaf, type_repr
NUMERIC_UNDERSCORES = 3
TRAILING_COMMA_IN_CALL = 4
TRAILING_COMMA_IN_DEF = 5
+ # The following two feature-flags are mutually exclusive, and exactly one should be
+ # set for every version of python.
+ ASYNC_IDENTIFIERS = 6
+ ASYNC_KEYWORDS = 7
VERSION_TO_FEATURES: Dict[TargetVersion, Set[Feature]] = {
- TargetVersion.PY27: set(),
- TargetVersion.PY33: {Feature.UNICODE_LITERALS},
- TargetVersion.PY34: {Feature.UNICODE_LITERALS},
- TargetVersion.PY35: {Feature.UNICODE_LITERALS, Feature.TRAILING_COMMA_IN_CALL},
+ TargetVersion.PY27: {Feature.ASYNC_IDENTIFIERS},
+ TargetVersion.PY33: {Feature.UNICODE_LITERALS, Feature.ASYNC_IDENTIFIERS},
+ TargetVersion.PY34: {Feature.UNICODE_LITERALS, Feature.ASYNC_IDENTIFIERS},
+ TargetVersion.PY35: {
+ Feature.UNICODE_LITERALS,
+ Feature.TRAILING_COMMA_IN_CALL,
+ Feature.ASYNC_IDENTIFIERS,
+ },
TargetVersion.PY36: {
Feature.UNICODE_LITERALS,
Feature.F_STRINGS,
Feature.NUMERIC_UNDERSCORES,
Feature.TRAILING_COMMA_IN_CALL,
Feature.TRAILING_COMMA_IN_DEF,
+ Feature.ASYNC_IDENTIFIERS,
},
TargetVersion.PY37: {
Feature.UNICODE_LITERALS,
Feature.NUMERIC_UNDERSCORES,
Feature.TRAILING_COMMA_IN_CALL,
Feature.TRAILING_COMMA_IN_DEF,
+ Feature.ASYNC_KEYWORDS,
},
TargetVersion.PY38: {
Feature.UNICODE_LITERALS,
Feature.NUMERIC_UNDERSCORES,
Feature.TRAILING_COMMA_IN_CALL,
Feature.TRAILING_COMMA_IN_DEF,
+ Feature.ASYNC_KEYWORDS,
},
}
if not target_versions:
# No target_version specified, so try all grammars.
return [
+ # Python 3.7+
+ 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,
]
elif all(version.is_python2() for version in target_versions):
# Python 2-only code, so try Python 2 grammars.
- return [pygram.python_grammar_no_print_statement, pygram.python_grammar]
+ return [
+ # Python 2.7 with future print_function import
+ pygram.python_grammar_no_print_statement,
+ # Python 2.7
+ pygram.python_grammar,
+ ]
else:
# Python 3-compatible code, so only try Python 3 grammar.
- return [pygram.python_grammar_no_print_statement_no_exec_statement]
+ grammars = []
+ # 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+
+ grammars.append(
+ pygram.python_grammar_no_print_statement_no_exec_statement_async_keywords # noqa: B950
+ )
+ 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)
+ # 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
def lib2to3_parse(src_txt: str, target_versions: Iterable[TargetVersion] = ()) -> Node:
consumed = 0
nlines = 0
+ ignored_lines = 0
for index, line in enumerate(prefix.split("\n")):
consumed += len(line) + 1 # adding the length of the split '\n'
line = line.lstrip()
if not line:
nlines += 1
if not line.startswith("#"):
+ # Escaped newlines outside of a comment are not really newlines at
+ # all. We treat a single-line comment following an escaped newline
+ # as a simple trailing comment.
+ if line.endswith("\\"):
+ ignored_lines += 1
continue
- if index == 0 and not is_endmarker:
+ if index == ignored_lines and not is_endmarker:
comment_type = token.COMMENT # simple trailing comment
else:
comment_type = STANDALONE_COMMENT
return ", ".join(report) + "."
+def parse_ast(src: str) -> Union[ast3.AST, ast27.AST]:
+ for feature_version in (7, 6):
+ try:
+ return ast3.parse(src, feature_version=feature_version)
+ except SyntaxError:
+ continue
+
+ return ast27.parse(src)
+
+
def assert_equivalent(src: str, dst: str) -> None:
"""Raise AssertionError if `src` and `dst` aren't equivalent."""
- import ast
import traceback
- def _v(node: ast.AST, depth: int = 0) -> Iterator[str]:
+ def _v(node: Union[ast3.AST, ast27.AST], depth: int = 0) -> Iterator[str]:
"""Simple visitor generating strings to compare ASTs by content."""
yield f"{' ' * depth}{node.__class__.__name__}("
for field in sorted(node._fields):
+ # TypeIgnore has only one field 'lineno' which breaks this comparison
+ if isinstance(node, (ast3.TypeIgnore, ast27.TypeIgnore)):
+ break
+
+ # Ignore str kind which is case sensitive / and ignores unicode_literals
+ if isinstance(node, (ast3.Str, ast27.Str, ast3.Bytes)) and field == "kind":
+ continue
+
try:
value = getattr(node, field)
except AttributeError:
# parentheses and they change the AST.
if (
field == "targets"
- and isinstance(node, ast.Delete)
- and isinstance(item, ast.Tuple)
+ and isinstance(node, (ast3.Delete, ast27.Delete))
+ and isinstance(item, (ast3.Tuple, ast27.Tuple))
):
for item in item.elts:
yield from _v(item, depth + 2)
- elif isinstance(item, ast.AST):
+ elif isinstance(item, (ast3.AST, ast27.AST)):
yield from _v(item, depth + 2)
- elif isinstance(value, ast.AST):
+ elif isinstance(value, (ast3.AST, ast27.AST)):
yield from _v(value, depth + 2)
else:
yield f"{' ' * depth}) # /{node.__class__.__name__}"
try:
- src_ast = ast.parse(src)
+ src_ast = parse_ast(src)
except Exception as exc:
- major, minor = sys.version_info[:2]
raise AssertionError(
- f"cannot use --safe with this file; failed to parse source file "
- f"with Python {major}.{minor}'s builtin AST. Re-run with --fast "
- f"or stop using deprecated Python 2 syntax. AST error message: {exc}"
+ f"cannot use --safe with this file; failed to parse source file. "
+ f"AST error message: {exc}"
)
try:
- dst_ast = ast.parse(dst)
+ dst_ast = parse_ast(dst)
except Exception as exc:
log = dump_to_file("".join(traceback.format_tb(exc.__traceback__)), dst)
raise AssertionError(