mirror of
https://github.com/davidhalter/parso.git
synced 2025-12-06 12:54:29 +08:00
Refactor the version info to use a tuple, always.
This commit is contained in:
@@ -3,7 +3,7 @@ import os
|
||||
|
||||
from parso._compatibility import FileNotFoundError
|
||||
from parso.pgen2.pgen import generate_grammar
|
||||
from parso.utils import splitlines, source_to_unicode, version_string_to_int
|
||||
from parso.utils import splitlines, source_to_unicode, parse_version_string
|
||||
from parso.python.diff import DiffParser
|
||||
from parso.python.tokenize import tokenize_lines, tokenize
|
||||
from parso.cache import parser_cache, load_module, save_module
|
||||
@@ -127,14 +127,14 @@ class Grammar(object):
|
||||
|
||||
|
||||
class PythonGrammar(Grammar):
|
||||
def __init__(self, version_int, bnf_text):
|
||||
def __init__(self, version_info, bnf_text):
|
||||
super(PythonGrammar, self).__init__(
|
||||
bnf_text,
|
||||
tokenizer=self._tokenize_lines,
|
||||
parser=PythonParser,
|
||||
diff_parser=DiffParser
|
||||
)
|
||||
self._version_int = version_int
|
||||
self._version_int = version_info
|
||||
|
||||
def _tokenize_lines(self, lines):
|
||||
return tokenize_lines(lines, self._version_int)
|
||||
@@ -152,16 +152,16 @@ def load_grammar(**kwargs):
|
||||
`version='3.3'`.
|
||||
"""
|
||||
def load_grammar(version=None):
|
||||
version_int = version_string_to_int(version)
|
||||
version_info = parse_version_string(version)
|
||||
|
||||
# For these versions we use the same grammar files, because nothing
|
||||
# changed.
|
||||
if version_int == 33:
|
||||
version_int = 34
|
||||
elif version_int == 26:
|
||||
version_int = 27
|
||||
if version_info == (3, 3):
|
||||
version_info = parse_version_string('3.4')
|
||||
elif version_info == (2, 6):
|
||||
version_info = parse_version_string('2.7')
|
||||
|
||||
file = 'python/grammar' + str(version_int) + '.txt'
|
||||
file = 'python/grammar%s%s.txt' % (version_info.major, version_info.minor)
|
||||
|
||||
global _loaded_grammars
|
||||
path = os.path.join(os.path.dirname(__file__), file)
|
||||
@@ -172,7 +172,7 @@ def load_grammar(**kwargs):
|
||||
with open(path) as f:
|
||||
bnf_text = f.read()
|
||||
|
||||
grammar = PythonGrammar(version_int, bnf_text)
|
||||
grammar = PythonGrammar(version_info, bnf_text)
|
||||
return _loaded_grammars.setdefault(path, grammar)
|
||||
except FileNotFoundError:
|
||||
message = "Python version %s is currently not supported." % version
|
||||
|
||||
@@ -8,12 +8,16 @@
|
||||
from parso.pgen2 import grammar
|
||||
from parso.python import token
|
||||
from parso.python import tokenize
|
||||
from parso.utils import parse_version_string
|
||||
|
||||
|
||||
class ParserGenerator(object):
|
||||
def __init__(self, bnf_text):
|
||||
self._bnf_text = bnf_text
|
||||
self.generator = tokenize.tokenize(bnf_text, version_int=36)
|
||||
self.generator = tokenize.tokenize(
|
||||
bnf_text,
|
||||
version_info=parse_version_string('3.6')
|
||||
)
|
||||
self._gettoken() # Initialize lookahead
|
||||
self.dfas, self.startsymbol = self._parse()
|
||||
self.first = {} # map from symbol name to set of tokens
|
||||
|
||||
@@ -64,14 +64,14 @@ def maybe(*choices):
|
||||
|
||||
|
||||
# Return the empty string, plus all of the valid string prefixes.
|
||||
def _all_string_prefixes(version_int):
|
||||
def _all_string_prefixes(version_info):
|
||||
# The valid string prefixes. Only contain the lower case versions,
|
||||
# and don't contain any permuations (include 'fr', but not
|
||||
# 'rf'). The various permutations will be generated.
|
||||
_valid_string_prefixes = ['b', 'r', 'u', 'br']
|
||||
if version_int >= 36:
|
||||
if version_info >= (3, 6):
|
||||
_valid_string_prefixes += ['f', 'fr']
|
||||
if version_int <= 27:
|
||||
if version_info <= (2, 7):
|
||||
# TODO this is actually not 100% valid. ur is valid in Python 2.7,
|
||||
# while ru is not.
|
||||
_valid_string_prefixes.append('ur')
|
||||
@@ -91,23 +91,23 @@ def _compile(expr):
|
||||
return re.compile(expr, re.UNICODE)
|
||||
|
||||
|
||||
def _get_token_collection(version_int):
|
||||
def _get_token_collection(version_info):
|
||||
try:
|
||||
return _token_collection_cache[version_int]
|
||||
return _token_collection_cache[tuple(version_info)]
|
||||
except KeyError:
|
||||
_token_collection_cache[version_int] = result = \
|
||||
_create_token_collection(version_int)
|
||||
_token_collection_cache[tuple(version_info)] = result = \
|
||||
_create_token_collection(version_info)
|
||||
return result
|
||||
|
||||
|
||||
def _create_token_collection(version_int):
|
||||
def _create_token_collection(version_info):
|
||||
# Note: we use unicode matching for names ("\w") but ascii matching for
|
||||
# number literals.
|
||||
Whitespace = r'[ \f\t]*'
|
||||
Comment = r'#[^\r\n]*'
|
||||
Name = r'\w+'
|
||||
|
||||
if version_int >= 36:
|
||||
if version_info >= (3, 6):
|
||||
Hexnumber = r'0[xX](?:_?[0-9a-fA-F])+'
|
||||
Binnumber = r'0[bB](?:_?[01])+'
|
||||
Octnumber = r'0[oO](?:_?[0-7])+'
|
||||
@@ -122,7 +122,7 @@ def _create_token_collection(version_int):
|
||||
else:
|
||||
Hexnumber = r'0[xX][0-9a-fA-F]+'
|
||||
Binnumber = r'0[bB][01]+'
|
||||
if version_int >= 30:
|
||||
if version_info >= (3, 0):
|
||||
Octnumber = r'0[oO][0-7]+'
|
||||
else:
|
||||
Octnumber = '0[oO]?[0-7]+'
|
||||
@@ -137,7 +137,7 @@ def _create_token_collection(version_int):
|
||||
|
||||
# Note that since _all_string_prefixes includes the empty string,
|
||||
# StringPrefix can be the empty string (making it optional).
|
||||
possible_prefixes = _all_string_prefixes(version_int)
|
||||
possible_prefixes = _all_string_prefixes(version_info)
|
||||
StringPrefix = group(*possible_prefixes)
|
||||
|
||||
# Tail end of ' string.
|
||||
@@ -161,7 +161,7 @@ def _create_token_collection(version_int):
|
||||
Bracket = '[][(){}]'
|
||||
|
||||
special_args = [r'\r?\n', r'[:;.,@]']
|
||||
if version_int >= 30:
|
||||
if version_info >= (3, 0):
|
||||
special_args.insert(0, r'\.\.\.')
|
||||
Special = group(*special_args)
|
||||
|
||||
@@ -233,13 +233,13 @@ class TokenInfo(namedtuple('Token', ['type', 'string', 'start_pos', 'prefix'])):
|
||||
return self.start_pos[0], self.start_pos[1] + len(self.string)
|
||||
|
||||
|
||||
def tokenize(code, version_int):
|
||||
def tokenize(code, version_info):
|
||||
"""Generate tokens from a the source code (string)."""
|
||||
lines = splitlines(code, keepends=True)
|
||||
return tokenize_lines(lines, version_int)
|
||||
return tokenize_lines(lines, version_info)
|
||||
|
||||
|
||||
def tokenize_lines(lines, version_int):
|
||||
def tokenize_lines(lines, version_info):
|
||||
"""
|
||||
A heavily modified Python standard library tokenizer.
|
||||
|
||||
@@ -248,7 +248,7 @@ def tokenize_lines(lines, version_int):
|
||||
that is irrelevant for the parser like newlines in parentheses or comments.
|
||||
"""
|
||||
pseudo_token, single_quoted, triple_quoted, endpats, always_break_tokens, = \
|
||||
_get_token_collection(version_int)
|
||||
_get_token_collection(version_info)
|
||||
paren_level = 0 # count parentheses
|
||||
indents = [0]
|
||||
max = 0
|
||||
|
||||
@@ -3,7 +3,10 @@ import re
|
||||
import sys
|
||||
from ast import literal_eval
|
||||
|
||||
from parso._compatibility import unicode
|
||||
from parso._compatibility import unicode, total_ordering
|
||||
|
||||
|
||||
Version = namedtuple('Version', 'major, minor, micro')
|
||||
|
||||
|
||||
def splitlines(string, keepends=False):
|
||||
@@ -82,7 +85,6 @@ def version_info():
|
||||
Returns a namedtuple of parso's version, similar to Python's
|
||||
``sys.version_info``.
|
||||
"""
|
||||
Version = namedtuple('Version', 'major, minor, micro')
|
||||
from parso import __version__
|
||||
tupl = re.findall(r'[a-z]+|\d+', __version__)
|
||||
return Version(*[x if i == 3 else int(x) for i, x in enumerate(tupl)])
|
||||
@@ -94,24 +96,48 @@ def _parse_version(version):
|
||||
raise ValueError('The given version is not in the right format. '
|
||||
'Use something like "3.2" or "3".')
|
||||
|
||||
major = match.group(1)
|
||||
major = int(match.group(1))
|
||||
minor = match.group(2)
|
||||
if minor is None:
|
||||
# Use the latest Python in case it's not exactly defined, because the
|
||||
# grammars are typically backwards compatible?
|
||||
if major == "2":
|
||||
if major == 2:
|
||||
minor = "7"
|
||||
elif major == "3":
|
||||
elif major == 3:
|
||||
minor = "6"
|
||||
else:
|
||||
raise NotImplementedError("Sorry, no support yet for those fancy new/old versions.")
|
||||
return int(major + minor)
|
||||
minor = int(minor)
|
||||
return PythonVersionInfo(major, minor)
|
||||
|
||||
|
||||
def version_string_to_int(version=None):
|
||||
@total_ordering
|
||||
class PythonVersionInfo(namedtuple('Version', 'major, minor')):
|
||||
def __gt__(self, other):
|
||||
if isinstance(other, tuple):
|
||||
if len(other) != 2:
|
||||
raise ValueError("Can only compare to tuples of length 2.")
|
||||
return (self.major, self.minor) > other
|
||||
super(PythonVersionInfo, self).__gt__(other)
|
||||
|
||||
return (self.major, self.minor)
|
||||
|
||||
def __eq__(self, other):
|
||||
if isinstance(other, tuple):
|
||||
if len(other) != 2:
|
||||
raise ValueError("Can only compare to tuples of length 2.")
|
||||
return (self.major, self.minor) == other
|
||||
super(PythonVersionInfo, self).__eq__(other)
|
||||
|
||||
def __ne__(self, other):
|
||||
return not self.__eq__(other)
|
||||
|
||||
|
||||
def parse_version_string(version=None):
|
||||
"""
|
||||
Checks for a valid version number (e.g. `3.2` or `2.7.1` or `3`) and
|
||||
returns a corresponding int that is always two characters long in decimal.
|
||||
returns a corresponding version info that is always two characters long in
|
||||
decimal.
|
||||
"""
|
||||
if version is None:
|
||||
version = '%s.%s' % sys.version_info[:2]
|
||||
|
||||
@@ -14,7 +14,7 @@ def test_load_inexisting_grammar():
|
||||
|
||||
|
||||
@pytest.mark.parametrize(('string', 'result'), [
|
||||
('2', 27), ('3', 36), ('1.1', 11), ('1.1.1', 11), ('300.1.31', 3001)
|
||||
('2', (2, 7)), ('3', (3, 6)), ('1.1', (1, 1)), ('1.1.1', (1, 1)), ('300.1.31', (300, 1))
|
||||
])
|
||||
def test_parse_version(string, result):
|
||||
assert utils._parse_version(string) == result
|
||||
|
||||
@@ -10,10 +10,9 @@ from textwrap import dedent
|
||||
|
||||
import pytest
|
||||
|
||||
from parso._compatibility import py_version
|
||||
from parso import load_grammar
|
||||
from parso import ParserSyntaxError
|
||||
from parso.utils import version_string_to_int
|
||||
from parso.utils import parse_version_string
|
||||
|
||||
|
||||
class Checker():
|
||||
@@ -32,8 +31,8 @@ def works_in_py2(each_version):
|
||||
|
||||
@pytest.fixture
|
||||
def works_ge_py3(each_version):
|
||||
version_int = version_string_to_int(each_version)
|
||||
return Checker(each_version, version_int >= 30)
|
||||
version_info = parse_version_string(each_version)
|
||||
return Checker(each_version, version_info >= (3, 0))
|
||||
|
||||
|
||||
@pytest.fixture
|
||||
@@ -41,8 +40,8 @@ def works_ge_py35(each_version):
|
||||
"""
|
||||
Works only greater equal Python 3.3.
|
||||
"""
|
||||
version_int = version_string_to_int(each_version)
|
||||
return Checker(each_version, version_int >= 35)
|
||||
version_info = parse_version_string(each_version)
|
||||
return Checker(each_version, version_info >= (3, 5))
|
||||
|
||||
|
||||
def _parse(code, version=None):
|
||||
|
||||
@@ -5,7 +5,7 @@ from textwrap import dedent
|
||||
import pytest
|
||||
|
||||
from parso._compatibility import py_version
|
||||
from parso.utils import splitlines, version_string_to_int
|
||||
from parso.utils import splitlines, parse_version_string
|
||||
from parso.python.token import (
|
||||
NAME, NEWLINE, STRING, INDENT, DEDENT, ERRORTOKEN, ENDMARKER)
|
||||
from parso.python import tokenize
|
||||
@@ -15,8 +15,8 @@ from parso.python.tokenize import TokenInfo
|
||||
|
||||
def _get_token_list(string):
|
||||
# Load the current version.
|
||||
version_int = version_string_to_int()
|
||||
return list(tokenize.tokenize(string, version_int))
|
||||
version_info = parse_version_string()
|
||||
return list(tokenize.tokenize(string, version_info))
|
||||
|
||||
|
||||
def test_end_pos_one_line():
|
||||
|
||||
Reference in New Issue
Block a user