def extract(script, new_name):
""" The `args` / `kwargs` params are the same as in `api.Script`.
:param operation: The refactoring operation to execute.
:type operation: str
:type source: str
:return: list of changed lines/changed files
"""
new_lines = split_lines(python_bytes_to_unicode(script.source))
old_lines = new_lines[:]
user_stmt = script._parser.user_stmt()
# TODO care for multiline extracts
dct = {}
if user_stmt:
pos = script._pos
line_index = pos[0] - 1
arr, index = helpers.array_for_pos(user_stmt, pos)
if arr is not None:
start_pos = arr[index].start_pos
end_pos = arr[index].end_pos
# take full line if the start line is different from end line
e = end_pos[1] if end_pos[0] == start_pos[0] else None
start_line = new_lines[start_pos[0] - 1]
text = start_line[start_pos[1]:e]
for l in range(start_pos[0], end_pos[0] - 1):
text += '\n' + l
if e is None:
end_line = new_lines[end_pos[0] - 1]
text += '\n' + end_line[:end_pos[1]]
# remove code from new lines
t = text.lstrip()
del_start = start_pos[1] + len(text) - len(t)
text = t.rstrip()
del_end = len(t) - len(text)
if e is None:
new_lines[end_pos[0] - 1] = end_line[end_pos[1] - del_end:]
e = len(start_line)
else:
e = e - del_end
start_line = start_line[:del_start] + new_name + start_line[e:]
new_lines[start_pos[0] - 1] = start_line
new_lines[start_pos[0]:end_pos[0] - 1] = []
# add parentheses in multiline case
open_brackets = ['(', '[', '{']
close_brackets = [')', ']', '}']
if '\n' in text and not (text[0] in open_brackets and text[-1] ==
close_brackets[open_brackets.index(text[0])]):
text = '(%s)' % text
# add new line before statement
indent = user_stmt.start_pos[1]
new = "%s%s = %s" % (' ' * indent, new_name, text)
new_lines.insert(line_index, new)
dct[script.path] = script.path, old_lines, new_lines
return Refactoring(dct)
def extract(script, new_name):
""" The `args` / `kwargs` params are the same as in `api.Script`.
:param operation: The refactoring operation to execute.
:type operation: str
:type source: str
:return: list of changed lines/changed files
"""
new_lines = split_lines(python_bytes_to_unicode(script.source))
old_lines = new_lines[:]
user_stmt = script._parser.user_stmt()
# TODO care for multiline extracts
dct = {}
if user_stmt:
pos = script._pos
line_index = pos[0] - 1
arr, index = helpers.array_for_pos(user_stmt, pos)
if arr is not None:
start_pos = arr[index].start_pos
end_pos = arr[index].end_pos
# take full line if the start line is different from end line
e = end_pos[1] if end_pos[0] == start_pos[0] else None
start_line = new_lines[start_pos[0] - 1]
text = start_line[start_pos[1]:e]
for l in range(start_pos[0], end_pos[0] - 1):
text += '\n' + l
if e is None:
end_line = new_lines[end_pos[0] - 1]
text += '\n' + end_line[:end_pos[1]]
# remove code from new lines
t = text.lstrip()
del_start = start_pos[1] + len(text) - len(t)
text = t.rstrip()
del_end = len(t) - len(text)
if e is None:
new_lines[end_pos[0] - 1] = end_line[end_pos[1] - del_end:]
e = len(start_line)
else:
e = e - del_end
start_line = start_line[:del_start] + new_name + start_line[e:]
new_lines[start_pos[0] - 1] = start_line
new_lines[start_pos[0]:end_pos[0] - 1] = []
# add parentheses in multiline case
open_brackets = ['(', '[', '{']
close_brackets = [')', ']', '}']
if '\n' in text and not (text[0] in open_brackets and text[-1] ==
close_brackets[open_brackets.index(text[0])]):
text = '(%s)' % text
# add new line before statement
indent = user_stmt.start_pos[1]
new = "%s%s = %s" % (' ' * indent, new_name, text)
new_lines.insert(line_index, new)
dct[script.path] = script.path, old_lines, new_lines
return Refactoring(dct)
def _py__path__(self):
search_path = self.evaluator.get_sys_path()
init_path = self.py__file__()
if os.path.basename(init_path) == '__init__.py':
with open(init_path, 'rb') as f:
content = python_bytes_to_unicode(f.read(), errors='replace')
# these are strings that need to be used for namespace packages,
# the first one is ``pkgutil``, the second ``pkg_resources``.
options = ('declare_namespace(__name__)',
'extend_path(__path__')
if options[0] in content or options[1] in content:
# It is a namespace, now try to find the rest of the
# modules on sys_path or whatever the search_path is.
paths = set()
for s in search_path:
other = os.path.join(s, self.name.string_name)
if os.path.isdir(other):
paths.add(other)
if paths:
return list(paths)
# TODO I'm not sure if this is how nested namespace
# packages work. The tests are not really good enough to
# show that.
# Default to this.
return [self._get_init_directory()]
def _rename(names, replace_str):
""" For both rename and inline. """
order = sorted(names, key=lambda x: (x.module_path, x.line, x.column),
reverse=True)
def process(path, old_lines, new_lines):
if new_lines is not None: # goto next file, save last
dct[path] = path, old_lines, new_lines
dct = {}
current_path = object()
new_lines = old_lines = None
for name in order:
if name.in_builtin_module():
continue
if current_path != name.module_path:
current_path = name.module_path
process(current_path, old_lines, new_lines)
if current_path is not None:
# None means take the source that is a normal param.
with open(current_path) as f:
source = f.read()
new_lines = split_lines(python_bytes_to_unicode(source))
old_lines = new_lines[:]
nr, indent = name.line, name.column
line = new_lines[nr - 1]
new_lines[nr - 1] = line[:indent] + replace_str + \
line[indent + len(name.name):]
process(current_path, old_lines, new_lines)
return dct
def _rename(names, replace_str):
""" For both rename and inline. """
order = sorted(names,
key=lambda x: (x.module_path, x.line, x.column),
reverse=True)
def process(path, old_lines, new_lines):
if new_lines is not None: # goto next file, save last
dct[path] = path, old_lines, new_lines
dct = {}
current_path = object()
new_lines = old_lines = None
for name in order:
if name.in_builtin_module():
continue
if current_path != name.module_path:
current_path = name.module_path
process(current_path, old_lines, new_lines)
if current_path is not None:
# None means take the source that is a normal param.
with open(current_path) as f:
source = f.read()
new_lines = split_lines(python_bytes_to_unicode(source))
old_lines = new_lines[:]
nr, indent = name.line, name.column
line = new_lines[nr - 1]
new_lines[nr - 1] = line[:indent] + replace_str + \
line[indent + len(name.name):]
process(current_path, old_lines, new_lines)
return dct
def __init__(self, source=None, line=None, column=None, path=None,
encoding='utf-8', sys_path=None):
self._orig_path = path
# An empty path (also empty string) should always result in no path.
self.path = os.path.abspath(path) if path else None
if source is None:
# TODO add a better warning than the traceback!
with open(path, 'rb') as f:
source = f.read()
# TODO do we really want that?
self._source = python_bytes_to_unicode(source, encoding, errors='replace')
self._code_lines = split_lines(self._source)
line = max(len(self._code_lines), 1) if line is None else line
if not (0 < line <= len(self._code_lines)):
raise ValueError('`line` parameter is not in a valid range.')
line_len = len(self._code_lines[line - 1])
column = line_len if column is None else column
if not (0 <= column <= line_len):
raise ValueError('`column` parameter is not in a valid range.')
self._pos = line, column
self._path = path
cache.clear_time_caches()
debug.reset_time()
# Load the Python grammar of the current interpreter.
self._grammar = parso.load_grammar()
project = Project(sys_path=sys_path)
self._evaluator = Evaluator(self._grammar, project)
project.add_script_path(self.path)
debug.speed('init')
def parse_and_get_code(self, code=None, path=None, **kwargs):
if self.allow_different_encoding:
if code is None:
with open(path, 'rb') as f:
code = f.read()
code = python_bytes_to_unicode(code, errors='replace')
return self.grammar.parse(code=code, path=path, **kwargs), code
def check_fs(path):
with open(path, 'rb') as f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name in code:
module = _load_module(evaluator, path, code)
module_name = sys_path.dotted_path_in_sys_path(evaluator.sys_path, path)
if module_name is not None:
add_module(evaluator, module_name, module)
return module
def check_fs(path, base_names):
try:
f = open(path, 'rb')
except FileNotFoundError:
return None
with f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name not in code:
return None
return _load_module_from_path(evaluator, path, base_names, code)
def check_fs(path):
with open(path, 'rb') as f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name in code:
module = _load_module(evaluator, path, code)
module_name = sys_path.dotted_path_in_sys_path(
evaluator.project.sys_path, path)
if module_name is not None:
add_module(evaluator, module_name, module)
return module
def check_fs(path):
with open(path, 'rb') as f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name in code:
e_sys_path = evaluator.get_sys_path()
module_name = sys_path.dotted_path_in_sys_path(e_sys_path, path)
module = _load_module(
evaluator, path, code,
sys_path=e_sys_path, module_name=module_name
)
return module
def check_fs(path):
with open(path, 'rb') as f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name in code:
e_sys_path = evaluator.get_sys_path()
module_name = sys_path.dotted_path_in_sys_path(e_sys_path, path)
module = _load_module(
evaluator, path, code,
sys_path=e_sys_path, module_name=module_name
)
return module
def parse_and_get_code(self, code=None, path=None, encoding='utf-8',
use_latest_grammar=False, file_io=None, **kwargs):
if self.allow_different_encoding:
if code is None:
if file_io is None:
file_io = FileIO(path)
code = file_io.read()
code = python_bytes_to_unicode(code, encoding=encoding, errors='replace')
grammar = self.latest_grammar if use_latest_grammar else self.grammar
return grammar.parse(code=code, path=path, file_io=file_io, **kwargs), code
def check_fs(file_io, base_names):
try:
code = file_io.read()
except FileNotFoundError:
return None
code = python_bytes_to_unicode(code, errors='replace')
if name not in code:
return None
new_file_io = KnownContentFileIO(file_io.path, code)
m = load_module_from_path(inference_state, new_file_io, base_names)
if isinstance(m, compiled.CompiledObject):
return None
return m.as_context()
def _check_fs(inference_state, file_io, regex):
try:
code = file_io.read()
except FileNotFoundError:
return None
code = python_bytes_to_unicode(code, errors='replace')
if not regex.search(code):
return None
new_file_io = KnownContentFileIO(file_io.path, code)
m = load_module_from_path(inference_state, new_file_io)
if m.is_compiled():
return None
return m.as_context()
def parse_and_get_code(self, code=None, path=None,
use_latest_grammar=False, file_io=None, **kwargs):
if path is not None:
path = str(path)
if code is None:
if file_io is None:
file_io = FileIO(path)
code = file_io.read()
# We cannot just use parso, because it doesn't use errors='replace'.
code = parso.python_bytes_to_unicode(code, encoding='utf-8', errors='replace')
if len(code) > settings._cropped_file_size:
code = code[:settings._cropped_file_size]
grammar = self.latest_grammar if use_latest_grammar else self.grammar
return grammar.parse(code=code, path=path, file_io=file_io, **kwargs), code
def check_fs(path):
try:
f = open(path, 'rb')
except FileNotFoundError:
return
with f:
code = python_bytes_to_unicode(f.read(), errors='replace')
if name in code:
e_sys_path = evaluator.get_sys_path()
import_names = sys_path.transform_path_to_dotted(e_sys_path, path)
module = _load_module(
evaluator, path, code,
sys_path=e_sys_path,
import_names=import_names,
)
return module
def inline(script):
"""
:type script: api.Script
"""
new_lines = split_lines(python_bytes_to_unicode(script.source))
dct = {}
definitions = script.goto_assignments()
assert len(definitions) == 1
stmt = definitions[0]._definition
usages = script.usages()
inlines = [
r for r in usages
if not stmt.start_pos <= (r.line, r.column) <= stmt.end_pos
]
inlines = sorted(inlines,
key=lambda x: (x.module_path, x.line, x.column),
reverse=True)
expression_list = stmt.expression_list()
# don't allow multi-line refactorings for now.
assert stmt.start_pos[0] == stmt.end_pos[0]
index = stmt.start_pos[0] - 1
line = new_lines[index]
replace_str = line[expression_list[0].start_pos[1]:stmt.end_pos[1] + 1]
replace_str = replace_str.strip()
# tuples need parentheses
if expression_list and isinstance(expression_list[0], pr.Array):
arr = expression_list[0]
if replace_str[0] not in ['(', '[', '{'] and len(arr) > 1:
replace_str = '(%s)' % replace_str
# if it's the only assignment, remove the statement
if len(stmt.get_defined_names()) == 1:
line = line[:stmt.start_pos[1]] + line[stmt.end_pos[1]:]
dct = _rename(inlines, replace_str)
# remove the empty line
new_lines = dct[script.path][2]
if line.strip():
new_lines[index] = line
else:
new_lines.pop(index)
return Refactoring(dct)
def inline(script):
"""
:type script: api.Script
"""
new_lines = split_lines(python_bytes_to_unicode(script.source))
dct = {}
definitions = script.goto_assignments()
with common.ignored(AssertionError):
assert len(definitions) == 1
stmt = definitions[0]._definition
usages = script.usages()
inlines = [r for r in usages
if not stmt.start_pos <= (r.line, r.column) <= stmt.end_pos]
inlines = sorted(inlines, key=lambda x: (x.module_path, x.line, x.column),
reverse=True)
expression_list = stmt.expression_list()
# don't allow multiline refactorings for now.
assert stmt.start_pos[0] == stmt.end_pos[0]
index = stmt.start_pos[0] - 1
line = new_lines[index]
replace_str = line[expression_list[0].start_pos[1]:stmt.end_pos[1] + 1]
replace_str = replace_str.strip()
# tuples need parentheses
if expression_list and isinstance(expression_list[0], pr.Array):
arr = expression_list[0]
if replace_str[0] not in ['(', '[', '{'] and len(arr) > 1:
replace_str = '(%s)' % replace_str
# if it's the only assignment, remove the statement
if len(stmt.get_defined_names()) == 1:
line = line[:stmt.start_pos[1]] + line[stmt.end_pos[1]:]
dct = _rename(inlines, replace_str)
# remove the empty line
new_lines = dct[script.path][2]
if line.strip():
new_lines[index] = line
else:
new_lines.pop(index)
return Refactoring(dct)
def __init__(self,
source=None,
line=None,
column=None,
path=None,
encoding='utf-8',
sys_path=None):
self._orig_path = path
# An empty path (also empty string) should always result in no path.
self.path = os.path.abspath(path) if path else None
if source is None:
# TODO add a better warning than the traceback!
with open(path, 'rb') as f:
source = f.read()
# TODO do we really want that?
self._source = python_bytes_to_unicode(source,
encoding,
errors='replace')
self._code_lines = split_lines(self._source)
line = max(len(self._code_lines), 1) if line is None else line
if not (0 < line <= len(self._code_lines)):
raise ValueError('`line` parameter is not in a valid range.')
line_len = len(self._code_lines[line - 1])
column = line_len if column is None else column
if not (0 <= column <= line_len):
raise ValueError('`column` parameter is not in a valid range.')
self._pos = line, column
self._path = path
cache.clear_time_caches()
debug.reset_time()
# Load the Python grammar of the current interpreter.
self._grammar = parso.load_grammar()
if sys_path is None:
venv = os.getenv('VIRTUAL_ENV')
if venv:
sys_path = list(get_venv_path(venv))
self._evaluator = Evaluator(self._grammar, sys_path=sys_path)
debug.speed('init')
def _py__path__(self):
search_path = self.evaluator.get_sys_path()
init_path = self.py__file__()
if os.path.basename(init_path) == '__init__.py':
with open(init_path, 'rb') as f:
content = python_bytes_to_unicode(f.read(), errors='replace')
# these are strings that need to be used for namespace packages,
# the first one is ``pkgutil``, the second ``pkg_resources``.
options = ('declare_namespace(__name__)', 'extend_path(__path__')
if options[0] in content or options[1] in content:
# It is a namespace, now try to find the rest of the
# modules on sys_path or whatever the search_path is.
paths = set()
for s in search_path:
other = os.path.join(s, self.name.string_name)
if os.path.isdir(other):
paths.add(other)
if paths:
return list(paths)
# TODO I'm not sure if this is how nested namespace
# packages work. The tests are not really good enough to
# show that.
# Default to this.
return [self._get_init_directory()]