Files
jedi/evaluate.py

837 lines
28 KiB
Python

"""
follow_statement -> follow_call -> follow_paths -> follow_path
'follow_import'
`get_names_for_scope` and `get_scopes_for_name` are search functions
TODO include super classes
TODO nonlocal statement
TODO doc
TODO list comprehensions, priority?
TODO care for *args **kwargs
TODO annotations
"""
from _compatibility import next
import itertools
import copy
import parsing
import modules
import debug
import builtin
memoize_caches = []
class MultiLevelStopIteration(Exception):
pass
def clear_caches():
for m in memoize_caches:
m.clear()
def memoize_default(default=None):
"""
This is a typical memoization decorator, BUT there is one difference:
To prevent recursion it sets defaults.
Preventing recursion is in this case the much bigger use than speed. I
don't think, that there is a big speed difference, but there are many cases
where recursion could happen (think about a = b; b = a).
"""
def func(function):
memo = {}
memoize_caches.append(memo)
def wrapper(*args, **kwargs):
key = (args, frozenset(kwargs.items()))
if key in memo:
return memo[key]
else:
memo[key] = default
rv = function(*args, **kwargs)
memo[key] = rv
return rv
return wrapper
return func
class Executable(object):
""" An instance is also an executable - because __init__ is called """
def __init__(self, base, var_args=[]):
self.base = base
# the param input array
self.var_args = var_args
self.func = None
def get_parent_until(self, *args):
return self.base.get_parent_until(*args)
@property
def scope(self):
""" Just try through the whole param array to find the own scope """
for param in self.var_args:
for call in param:
try:
return call.parent_stmt.parent
except AttributeError: # if operators are there
pass
raise IndexError('No params available')
@memoize_default(default=[])
def get_params(self):
"""
This returns the params for an Execution/Instance and is injected as a
'hack' into the parsing.Function class.
This needs to be here, because Instance can have __init__ functions,
which act the same way as normal functions
"""
def gen_param_name_copy(param, keys=[], values=[], array_type=None):
calls = parsing.Array(parsing.Array.NOARRAY,
self.var_args.parent_stmt)
calls.values = values
calls.keys = keys
calls.type = array_type
new_param = copy.copy(param)
new_param._assignment_calls_calculated = True
new_param._assignment_calls = calls
name = copy.copy(param.get_name())
name.parent = new_param
#print 'insert', i, name, calls.values, value, self.func.params
return name
result = []
start_offset = 0
#print '\n\nfunc_params', self.func, self.func.parent, self.func
if isinstance(self.func, InstanceElement):
# care for self -> just exclude it and add the instance
start_offset = 1
self_name = copy.copy(self.func.params[0].get_name())
self_name.parent = self.func.instance
result.append(self_name)
param_dict = {}
for param in self.func.params:
param_dict[str(param.get_name())] = param
# There may be calls, which don't fit all the params, this just ignores
# it.
var_arg_iterator = self.get_var_args_iterator()
non_matching_keys = []
for param in self.func.params[start_offset:]:
# The value and key can both be null. There, the defaults apply.
# args / kwargs will just be empty arrays / dicts, respectively.
key, value = next(var_arg_iterator, (None, None))
while key:
try:
key_param = param_dict[str(key)]
except KeyError:
non_matching_keys.append((key, value))
else:
result.append(gen_param_name_copy(key_param,
values=[value]))
key, value = next(var_arg_iterator, (None, None))
#debug.warning('Too many arguments given.', value)
assignment = param.get_assignment_calls().values[0]
keys = []
values = []
array_type = None
if assignment[0] == '*':
# *args param
array_type = parsing.Array.TUPLE
if value:
values.append(value)
for key, value in var_arg_iterator:
# iterate until a key argument is found
if key:
var_arg_iterator.push_back(key, value)
break
values.append(value)
elif assignment[0] == '**':
# **kwargs param
array_type = parsing.Array.DICT
if non_matching_keys:
keys, values = zip(*non_matching_keys)
else:
# normal param
if value:
values = [value]
result.append(gen_param_name_copy(param, keys=keys, values=values,
array_type=array_type))
return result
def get_var_args_iterator(self):
"""
Yields a key/value pair, the key is None, if its not a named arg.
"""
def iterate():
# var_args is typically an Array, and not a list
for var_arg in self.var_args:
# *args
if var_arg[0] == '*':
arrays = follow_call_list(self.scope, [var_arg[1:]])
for array in arrays:
for field in array.get_contents():
yield None, field
# **kwargs
elif var_arg[0] == '**':
arrays = follow_call_list(self.scope, [var_arg[1:]])
for array in arrays:
for key, field in array.get_contents():
yield key[0].name, field
yield var_arg
# normal arguments (including key arguments)
else:
if len(var_arg) > 1 and var_arg[1] == '=':
# this is a named parameter
yield var_arg[0].name, var_arg[2:]
else:
yield None, var_arg
class PushBackIterator(object):
def __init__(self, iterator):
self.pushes = []
self.iterator = iterator
def push_back(self, key, value):
self.pushes.append((key,value))
def __iter__(self):
return self
def next(self):
""" Python 2 Compatibility """
return self.__next__()
def __next__(self):
try:
return self.pushes.pop()
except IndexError:
return next(self.iterator)
return iter(PushBackIterator(iterate()))
def set_param_cb(self, func):
self.func = func
func.param_cb = self.get_params
class Instance(Executable):
""" This class is used to evaluate instances. """
def __init__(self, base, var_args=[]):
super(Instance, self).__init__(base, var_args)
if var_args:
self.set_init_params()
def set_init_params(self):
for sub in self.base.subscopes:
if isinstance(sub, parsing.Function) \
and sub.name.get_code() == '__init__':
self.set_param_cb(InstanceElement(self, sub))
def get_func_self_name(self, func):
"""
Returns the name of the first param in a class method (which is
normally self
"""
try:
return func.params[0].used_vars[0].names[0]
except:
return None
def get_defined_names(self):
"""
Get the instance vars of a class. This includes the vars of all
classes
"""
def add_self_name(name):
n = copy.copy(name)
n.names = n.names[1:]
names.append(InstanceElement(self, n))
names = []
# this loop adds the names of the self object, copies them and removes
# the self.
for s in self.base.subscopes:
# get the self name, if there's one
self_name = self.get_func_self_name(s)
if self_name:
for n in s.get_set_vars():
# Only names with the selfname are being added.
# It is also important, that they have a len() of 2,
# because otherwise, they are just something else
if n.names[0] == self_name and len(n.names) == 2:
add_self_name(n)
for var in self.base.get_defined_names(as_instance=True):
# functions are also instance elements
if isinstance(var.parent, (parsing.Function)):
var = InstanceElement(self, var)
names.append(var)
return names
def parent(self):
return self.base.parent
def __repr__(self):
return "<p%s of %s (var_args: %s)>" % \
(self.__class__.__name__, self.base, len(self.var_args or []))
class InstanceElement(object):
def __init__(self, instance, var):
super(InstanceElement, self).__init__()
self.instance = instance
self.var = var
@property
def parent(self):
return InstanceElement(self.instance, self.var.parent)
@property
def param_cb(self):
return self.var.param_cb
@param_cb.setter
def param_cb(self, value):
self.var.param_cb = value
def __getattr__(self, name):
return getattr(self.var, name)
def __repr__(self):
return "<%s of %s>" % (self.__class__.__name__, self.var)
class Class(object):
def __init__(self, base):
self.base = base
def get_defined_names(self, as_instance=False):
def in_iterable(name, iterable):
for i in iterable:
# only the last name is important, because these names have a
# maximal length of 2, with the first one being `self`.
if i.names[-1] == name.names[-1]:
return True
return False
names = self.base.get_defined_names()
# check super classes:
for s in self.base.supers:
for cls in follow_statement(s):
# get the inherited names
if as_instance:
cls = Instance(cls)
for i in cls.get_defined_names():
if not in_iterable(i, names):
names.append(i)
return names
def __getattr__(self, name):
return getattr(self.base, name)
def __repr__(self):
return "<p%s of %s>" % (self.__class__.__name__, self.base)
class Execution(Executable):
"""
This class is used to evaluate functions and their returns.
"""
cache = {}
@memoize_default(default=[])
def get_return_types(self):
"""
Get the return vars of a function.
"""
stmts = []
#print '\n\n', self.var_args, self.var_args.values, self.var_args.parent_stmt
if isinstance(self.base, Class):
# there maybe executions of executions
stmts = [Instance(self.base, self.var_args)]
else:
# set the callback function to get the var_args
self.set_param_cb(self.base)
# don't do this with exceptions, as usual, because some deeper
# exceptions could be catched - and I wouldn't know what happened.
if hasattr(self.base, 'returns'):
ret = self.base.returns
for s in ret:
#temp, s.parent = s.parent, self
stmts += follow_statement(s)
#s.parent = temp
# reset the callback function on exit
# TODO how can we deactivate this again?
#self.base.param_cb = None
else:
debug.warning("no execution possible", self.base)
debug.dbg('exec stmts=', stmts, self.base, repr(self))
return stmts
def __repr__(self):
return "<%s of %s>" % \
(self.__class__.__name__, self.base)
class Array(object):
"""
Used as a mirror to parsing.Array, if needed. It defines some getter
methods which are important in this module.
"""
def __init__(self, array):
self._array = array
def get_index_types(self, index=None):
values = self._array.values
if index is not None:
# This is indexing only one element, with a fixed index number,
# otherwise it just ignores the index (e.g. [1+1])
try:
index_nr = int(index.get_only_subelement().name)
values = [self._array[index_nr]]
except:
pass
scope = self._array.parent_stmt.parent
return follow_call_list(scope, values)
def get_exact_index_types(self, index):
values = [self._array[index]]
scope = self._array.parent_stmt.parent
return follow_call_list(scope, values)
def get_defined_names(self):
""" This method generates all ArrayElements for one parsing.Array. """
# array.type is a string with the type, e.g. 'list'
scope = get_scopes_for_name(builtin.Builtin.scope, self._array.type)[0]
names = scope.get_defined_names()
return [ArrayElement(n) for n in names]
def get_contents(self):
return self._array
def __repr__(self):
return "<p%s of %s>" % (self.__class__.__name__, self._array)
class ArrayElement(object):
def __init__(self, name):
super(ArrayElement, self).__init__()
self.name = name
@property
def parent(self):
raise NotImplementedError("This shouldn't happen")
@property
def returns(self):
return self.name.parent.returns
@property
def names(self):
return self.name.names
def __repr__(self):
return "<%s of %s>" % (self.__class__.__name__, self.name)
def get_defined_names_for_position(obj, position):
names = obj.get_defined_names()
if not position:
return names
names_new = []
for n in names:
if (n.line_nr, n.indent) < position:
names_new.append(n)
return names_new
def get_names_for_scope(scope, position=None, star_search=True):
"""
Get all completions possible for the current scope.
The star search option is only here to provide an optimization. Otherwise
the whole thing would probably start a little recursive madness.
"""
start_scope = scope
while scope:
# class variables/functions are only available
if (not isinstance(scope, Class) or scope == start_scope) \
and not isinstance(scope, parsing.Flow):
try:
yield scope, get_defined_names_for_position(scope, position)
except StopIteration:
raise MultiLevelStopIteration('StopIteration raised somewhere')
scope = scope.parent
# add star imports
if star_search:
for s in remove_star_imports(start_scope.get_parent_until()):
for g in get_names_for_scope(s, star_search=False):
yield g
# add builtins to the global scope
builtin_scope = builtin.Builtin.scope
yield builtin_scope, builtin_scope.get_defined_names()
def get_scopes_for_name(scope, name_str, position=None, search_global=False):
"""
:param position: Position of the last statement ->tuple of line, indent
:return: List of Names. Their parents are the scopes, they are defined in.
:rtype: list
"""
def remove_statements(result):
"""
This is the part where statements are being stripped.
Due to lazy evaluation, statements like a = func; b = a; b() have to be
evaluated.
"""
res_new = []
for r in result:
if isinstance(r, parsing.Statement) \
or isinstance(r, InstanceElement) \
and isinstance(r.var, parsing.Statement):
# global variables handling
if r.is_global():
for token_name in r.token_list[1:]:
if isinstance(token_name, parsing.Name):
res_new += get_scopes_for_name(r.parent,
str(token_name))
else:
scopes = follow_statement(r, seek_name=name_str)
res_new += remove_statements(scopes)
else:
if isinstance(r, parsing.Class):
r = Class(r)
res_new.append(r)
debug.dbg('sfn remove, new: %s, old: %s' % (res_new, result))
return res_new
def filter_name(scope_generator):
def handle_non_arrays(name):
result = []
par = name.parent
if isinstance(par, parsing.Flow):
if par.command == 'for':
# take the first statement (for has always only
# one, remember `in`). And follow it. After that,
# get the types which are in the array
arrays = follow_statement(par.inits[0])
for array in arrays:
in_vars = array.get_index_types()
if len(par.set_vars) > 1:
var_arr = par.set_stmt.get_assignment_calls()
result += assign_tuples(var_arr, in_vars, name_str)
else:
result += in_vars
else:
debug.warning('Flow: Why are you here? %s' % par.command)
elif isinstance(par, parsing.Param) \
and isinstance(par.parent.parent, parsing.Class) \
and par.position == 0:
# this is where self gets added - this happens at another
# place, if the var_args are clear. But some times the class is
# not known. Therefore set self.
result.append(Instance(Class(par.parent.parent)))
result.append(par)
else:
result.append(par)
return result
result = []
# compare func uses the tuple of line/indent = row/column
comparison_func = lambda name: (name.line_nr, name.indent)
for scope, name_list in scope_generator:
# here is the position stuff happening (sorting of variables)
for name in sorted(name_list, key=comparison_func, reverse=True):
if name_str == name.get_code():
result += handle_non_arrays(name)
#print name, name.parent.parent, scope
# this means that a definition was found and is not e.g.
# in if/else.
if name.parent.parent == scope:
break
# if there are results, ignore the other scopes
if result:
break
debug.dbg('sfn filter', name_str, result)
return result
if search_global:
scope_generator = get_names_for_scope(scope, position=position)
else:
if position:
names = get_defined_names_for_position(scope, position)
else:
names = scope.get_defined_names()
scope_generator = iter([(scope, names)])
#print ' ln', position
return remove_statements(filter_name(scope_generator))
def strip_imports(scopes):
"""
Here we strip the imports - they don't get resolved necessarily.
Really used anymore?
"""
result = []
for s in scopes:
if isinstance(s, parsing.Import):
#print 'dini mueter, steile griech!'
try:
result += follow_import(s)
except modules.ModuleNotFound:
debug.warning('Module not found: ' + str(s))
else:
result.append(s)
return result
def assign_tuples(tup, results, seek_name):
"""
This is a normal assignment checker. In python functions and other things
can return tuples:
>>> a, b = 1, ""
>>> a, (b, c) = 1, ("", 1.0)
Here, if seek_name is "a", the number type will be returned.
The first part (before `=`) is the param tuples, the second one result.
:type tup: parsing.Array
"""
def eval_results(index):
types = []
for r in results:
if hasattr(r, "get_exact_index_types"):
types += r.get_exact_index_types(index)
else:
debug.warning("assign tuples: invalid tuple lookup")
return types
result = []
if tup.type == parsing.Array.NOARRAY:
# here we have unnessecary braces, which we just remove
arr = tup.get_only_subelement()
result = assign_tuples(arr, results, seek_name)
else:
for i, t in enumerate(tup):
# used in assignments. there is just one call and no other things,
# therefor we can just assume, that the first part is important.
if len(t) != 1:
raise AttributeError('Array length should be 1')
t = t[0]
# check the left part, if it's still tuples in it or a Call
if isinstance(t, parsing.Array):
# these are "sub" tuples
result += assign_tuples(t, eval_results(i), seek_name)
else:
if t.name.names[-1] == seek_name:
result += eval_results(i)
return result
@memoize_default(default=[])
def follow_statement(stmt, scope=None, seek_name=None):
"""
:param stmt: contains a statement
:param scope: contains a scope. If not given, takes the parent of stmt.
"""
if scope is None:
scope = stmt.get_parent_until(parsing.Function, Execution,
parsing.Class, Instance,
InstanceElement)
debug.dbg('follow_stmt', stmt, 'in', scope, seek_name)
call_list = stmt.get_assignment_calls()
debug.dbg('calls', call_list, call_list.values)
result = follow_call_list(scope, call_list)
# assignment checking is only important if the statement defines multiple
# variables
if len(stmt.get_set_vars()) > 1 and seek_name and stmt.assignment_details:
# TODO this should have its own call_list, because call_list can also
# return 3 results for 2 variables.
new_result = []
for op, set_vars in stmt.assignment_details:
new_result += assign_tuples(set_vars, result, seek_name)
result = new_result
return result
def follow_call_list(scope, call_list):
"""
The call_list has a special structure.
This can be either `parsing.Array` or `list of list`.
It is used to evaluate a two dimensional object, that has calls, arrays and
operators in it.
"""
if parsing.Array.is_type(call_list, parsing.Array.TUPLE,
parsing.Array.DICT):
# Tuples can stand just alone without any braces. These would be
# recognized as separate calls, but actually are a tuple.
result = follow_call(scope, call_list)
else:
result = []
for calls in call_list:
for call in calls:
if parsing.Array.is_type(call, parsing.Array.NOARRAY):
result += follow_call_list(scope, call)
else:
if not isinstance(call, str):
# The string tokens are just operations (+, -, etc.)
result += follow_call(scope, call)
return set(result)
def follow_call(scope, call):
""" Follow a call is following a function, variable, string, etc. """
path = call.generate_call_list()
position = (call.parent_stmt.line_nr, call.parent_stmt.indent)
current = next(path)
if isinstance(current, parsing.Array):
result = [Array(current)]
else:
# TODO add better care for int/unicode, now str/float are just used
# instead
if not isinstance(current, parsing.NamePart):
if current.type == parsing.Call.STRING:
scopes = get_scopes_for_name(builtin.Builtin.scope, 'str')
elif current.type == parsing.Call.NUMBER:
scopes = get_scopes_for_name(builtin.Builtin.scope, 'float')
else:
debug.warning('unknown type:', current.type, current)
# make instances of those number/string objects
scopes = [Instance(s) for s in scopes]
else:
# this is the first global lookup
scopes = get_scopes_for_name(scope, current, position=position,
search_global=True)
result = strip_imports(scopes)
if result != scopes:
# reset the position, when imports where stripped
position = None
debug.dbg('call before result %s, current %s, scope %s'
% (result, current, scope))
result = follow_paths(path, result, position=position)
return result
def follow_paths(path, results, position=None):
results_new = []
if results:
if len(results) > 1:
iter_paths = itertools.tee(path, len(results))
else:
iter_paths = [path]
for i, r in enumerate(results):
fp = follow_path(iter_paths[i], r, position=position)
if fp is not None:
results_new += fp
else:
# this means stop iteration
return results
return results_new
def follow_path(path, scope, position=None):
"""
Takes a generator and tries to complete the path.
"""
# current is either an Array or a Scope
try:
current = next(path)
except StopIteration:
return None
debug.dbg('follow', current, scope)
result = []
if isinstance(current, parsing.Array):
# this must be an execution, either () or []
if current.type == parsing.Array.LIST:
result = scope.get_index_types(current)
elif current.type not in [parsing.Array.DICT]:
# scope must be a class or func - make an instance or execution
debug.dbg('befexec', scope)
exe = Execution(scope, current)
result = strip_imports(exe.get_return_types())
debug.dbg('exec', result)
else:
# curly braces are not allowed, because they make no sense
debug.warning('strange function call with {}', current, scope)
else:
if isinstance(scope, parsing.Function):
# TODO check default function methods and return them
result = []
else:
# TODO check magic class methods and return them also
# this is the typical lookup while chaining things
result = strip_imports(get_scopes_for_name(scope, current,
position=position))
return follow_paths(path, result, position=position)
def follow_import(_import):
"""
follows a module name and returns the parser.
:param _import: The import statement.
:type _import: parsing.Import
"""
# set path together
ns_list = []
if _import.from_ns:
ns_list += _import.from_ns.names
if _import.namespace:
ns_list += _import.namespace.names
loaded_in = _import.get_parent_until()
scope, rest = modules.find_module(loaded_in, ns_list)
if rest:
scopes = follow_path(iter(rest), scope)
else:
scopes = [scope]
new = []
for scope in scopes:
new += remove_star_imports(scope)
scopes += new
debug.dbg('after import', scopes, rest)
return scopes
def remove_star_imports(scope):
"""
TODO doc
"""
modules = strip_imports(i for i in scope.get_imports() if i.star)
new = []
for m in modules:
new += remove_star_imports(m)
modules += new
# filter duplicate modules
return set(modules)