1
0
forked from VimPlug/jedi
Files
jedi-fork/jedi/fast_parser.py
2013-01-12 00:14:30 +01:00

270 lines
9.3 KiB
Python

import re
import operator
from _compatibility import use_metaclass, reduce, property
import settings
import parsing
import cache
class Module(parsing.Simple, parsing.Module):
def __init__(self, parsers):
self._end_pos = None, None
super(Module, self).__init__(self, (1,0))
self.parsers = parsers
self.reset_caches()
self.line_offset = 0
def reset_caches(self):
""" This module does a whole lot of caching, because it uses different
parsers. """
self.cache = {}
for p in self.parsers:
p.user_scope = None
p.user_stmt = None
def _get(self, name, operation, execute=False, *args, **kwargs):
key = (name, args, frozenset(kwargs.items()))
if key not in self.cache:
if execute:
objs = (getattr(p.module, name)(*args, **kwargs)
for p in self.parsers)
else:
objs = (getattr(p.module, name) for p in self.parsers)
self.cache[key] = reduce(operation, objs)
return self.cache[key]
def __getattr__(self, name):
operators = {'get_imports': operator.add,
'get_code': operator.add,
'get_set_vars': operator.add,
'get_defined_names': operator.add,
'is_empty': operator.and_
}
properties = {'subscopes': operator.add,
'imports': operator.add,
'statements': operator.add,
'imports': operator.add,
'asserts': operator.add,
'global_vars': operator.add
}
if name in operators:
return lambda *args, **kwargs: self._get(name, operators[name],
True, *args, **kwargs)
elif name in properties:
return self._get(name, properties[name])
else:
raise AttributeError("__getattr__ doesn't offer %s" % name)
def get_statement_for_position(self, pos):
key = 'get_statement_for_position', pos
if key not in self.cache:
for p in self.parsers:
s = p.module.get_statement_for_position(pos)
if s:
self.cache[key] = s
break
else:
self.cache[key] = None
return self.cache[key]
@property
def used_names(self):
if not self.parsers:
raise NotImplementedError("Parser doesn't exist.")
key = 'used_names'
if key not in self.cache:
dct = {}
for p in self.parsers:
for k, statement_set in p.module.used_names.items():
if k in dct:
dct[k] |= statement_set
else:
dct[k] = set(statement_set)
self.cache[key] = dct
return self.cache[key]
@property
def docstr(self):
if not self.parsers:
raise NotImplementedError("Parser doesn't exist.")
return self.parsers[0].module.docstr
@property
def name(self):
if not self.parsers:
raise NotImplementedError("Parser doesn't exist.")
return self.parsers[0].module.name
@property
def path(self):
if not self.parsers:
raise NotImplementedError("Parser doesn't exist.")
return self.parsers[0].module.path
@property
def is_builtin(self):
if not self.parsers:
raise NotImplementedError("Parser doesn't exist.")
return self.parsers[0].module.is_builtin
@property
def start_pos(self):
""" overwrite start_pos of Simple """
return 1, 0
@start_pos.setter
def start_pos(self):
""" ignore """
pass
@property
def end_pos(self):
return self._end_pos
@end_pos.setter
def end_pos(self, value):
if None in self._end_pos \
or None not in value and self._end_pos < value:
self._end_pos = value
def __repr__(self):
return "<%s: %s@%s-%s>" % (type(self).__name__, self.name,
self.start_pos[0], self.end_pos[0])
class CachedFastParser(type):
""" This is a metaclass for caching `FastParser`. """
def __call__(self, source, module_path=None, user_position=None):
if not settings.fast_parser:
return parsing.PyFuzzyParser(source, module_path, user_position)
pi = cache.parser_cache.get(module_path, None)
if pi is None or isinstance(pi.parser, parsing.PyFuzzyParser):
p = super(CachedFastParser, self).__call__(source, module_path,
user_position)
else:
p = pi.parser # pi is a `cache.ParserCacheItem`
p.update(source, user_position)
return p
class FastParser(use_metaclass(CachedFastParser)):
def __init__(self, code, module_path=None, user_position=None):
# set values like `parsing.Module`.
self.module_path = module_path
self.user_position = user_position
self.parsers = []
self.module = Module(self.parsers)
self.reset_caches()
self._parse(code)
@property
def user_scope(self):
if self._user_scope is None:
for p in self.parsers:
if p.user_scope:
if self._user_scope is not None and not \
isinstance(self._user_scope, parsing.SubModule):
continue
self._user_scope = p.user_scope
if isinstance(self._user_scope, parsing.SubModule):
self._user_scope = self.module
return self._user_scope
@property
def user_stmt(self):
if self._user_stmt is None:
for p in self.parsers:
if p.user_stmt:
self._user_stmt = p.user_stmt
break
return self._user_stmt
def update(self, code, user_position=None):
self.user_position = user_position
self.reset_caches()
self._parse(code)
def scan_user_scope(self, sub_module):
""" Scan with self.user_position.
:type sub_module: parsing.SubModule
"""
for scope in sub_module.statements + sub_module.subscopes:
if isinstance(scope, parsing.Scope):
if scope.start_pos <= self.user_position <= scope.end_pos:
return self.scan_user_scope(scope) or scope
return None
def _parse(self, code):
""" :type code: str """
r = r'(?:\n(?:def|class|@.*?\n(?:def|class))|^).*?' \
r'(?=\n(?:def|class|@)|$)'
parts = re.findall(r, code, re.DOTALL)
if len(parts) > 1 and not re.match('def|class|@', parts[0]):
# Merge the first two because `common.NoErrorTokenizer` is not able
# to know if there's a class/func or not.
# Therefore every part has it's own class/func. Exactly one.
parts[0] += parts[1]
parts.pop(1)
if settings.fast_parser_always_reparse:
self.parsers[:] = []
# dict comprehensions are not available in py2.5/2.6 :-(
hashes = dict((p.hash, p) for p in self.parsers)
line_offset = 0
start = 0
p = None
parser_order = 0
for code_part in parts:
lines = code_part.count('\n')
# the parser is using additional newlines, therefore substract
if p is None or line_offset >= p.end_pos[0] - 2:
# check if code_part has already been parsed
h = hash(code_part)
if h in hashes and hashes[h].code == code_part:
p = hashes[h]
del hashes[h]
m = p.module
m.line_offset += line_offset + 1 - m.start_pos[0]
if self.user_position is not None and \
m.start_pos <= self.user_position <= m.end_pos:
# It's important to take care of the whole user
# positioning stuff, if no reparsing is being done.
p.user_stmt = m.get_statement_for_position(
self.user_position, include_imports=True)
if p.user_stmt:
p.user_scope = p.user_stmt.parent
else:
p.user_scope = self.scan_user_scope(m) \
or self.module
else:
p = parsing.PyFuzzyParser(code[start:],
self.module_path, self.user_position,
line_offset=line_offset, stop_on_scope=True,
top_module=self.module)
p.hash = h
p.code = code_part
p.module.parent = self.module
self.parsers.insert(parser_order, p)
parser_order += 1
line_offset += lines
start += len(code_part)
self.parsers[parser_order + 1:] = []
def reset_caches(self):
self._user_scope = None
self._user_stmt = None
self.module.reset_caches()