def parse_symbols(file_num_to_consider: List[str]) -> Set[str]:
var.get(NL)
current_line = var.get(CL)
assert current_line == "# Address Size File Name"
var.get(NL)
symbol_line_pattern = re.compile(
".*?\[\s*(?P<file_num>\d+)\]\s*[\+-]\[(?P<obj_name>\S+)\s+(?P<method_name>\S+?)\]"
)
current_line = var.get(CL)
objects_name = set()
while current_line:
if current_line.startswith("#"):
break
else:
m = symbol_line_pattern.match(current_line)
if m and m.group('file_num') in file_num_to_consider:
objects_name.add(m.group('obj_name'))
var.get(NL)
current_line = var.get(CL)
return objects_name
def _move_to_seg_and_sect(seg, sect):
current_line = var.get(CL)
while current_line:
m = new_section_pattern.match(current_line)
if m and m.group('segname') == seg and m.group('sectname') == sect:
var.get(NL)
return
var.get(NL)
current_line = var.get(CL)
raise Exception(f'segment: {seg} section: {sect} not found')
def parse_link_map(lp, pod=None):
var.set(CF, open(lp, 'r'))
var.get(NL)
current_line = var.get(CL) #type:str
file_num_to_consider = None
object_names = None
while current_line:
if current_line.startswith("#"):
if current_line.replace(" ", '') == object_file_start_line:
file_num_to_consider = parse_object_files(pod=pod)
elif current_line.replace(" ", '') == symbols_start_line:
assert file_num_to_consider
object_names = parse_symbols(file_num_to_consider)
else:
var.get(NL)
else:
var.get(NL)
current_line = var.get(CL)
assert object_names
return object_names
def parse_object_files(pod=None) -> List:
object_file_line_pattern = re.compile(
"\[\s*(?P<file_num>\d+)\]\s+.+?\.a\((?P<file_name>\S+)\.o\)")
var.get(NL)
file_num_belongs_to_pod = []
current_line = var.get(CL) # type:str
while current_line:
if current_line.startswith("#"):
break
if pod and pod not in current_line:
pass
else:
m = object_file_line_pattern.match(current_line)
if m:
file_num_belongs_to_pod.append(m.group('file_num'))
var.get(NL)
current_line = var.get(CL)
return file_num_belongs_to_pod
def parse_class_list(ep) -> dict:
current_file = subprocess.Popen(['otool', '-oV', ep],
stdout=subprocess.PIPE).stdout
var.set(CF, current_file)
var.get(NL)
_move_to_seg_and_sect('__DATA', '__objc_classlist')
class_objects = {}
current_line = var.get(CL)
while current_line:
if new_class_pattern.match(current_line):
class_object = _parse_implemented_class()
class_objects[class_object.name] = class_object
elif new_section_pattern.match(current_line):
break
else:
var.get(NL)
current_line = var.get(CL)
return class_objects
def parse_referenced_methods(ep) -> List:
current_file = subprocess.Popen(
['otool', '-v', '-s', '__DATA', '__objc_selrefs', ep],
stdout=subprocess.PIPE).stdout
var.set(CF, current_file)
var.get(NL)
_move_to_seg_and_sect('__DATA', '__objc_selrefs')
referenced_methods = []
pattern = re.compile(
"[0-9a-f]{16}\s*__TEXT:__objc_methname:(?P<method>\S+)")
current_line = var.get(CL)
while current_line:
m = pattern.match(current_line)
if m:
referenced_methods.append(m.group("method"))
# logger.info(f"method {referenced_methods[-1]} referenced")
var.get(NL)
current_line = var.get(CL)
return referenced_methods
def _parse_method_list(class_name) -> list:
current_line = var.get(CL)
r = []
if current_line.startswith("entsize"):
var.get(NL)
current_line = var.get(CL)
assert current_line.startswith("count")
count = int(current_line.split(" ")[-1])
var.get(NL)
while count > 0:
current_line = var.get(CL)
assert current_line.startswith(
"name"), f"error with class {class_name}"
method_name = current_line.split()[-1]
r.append(method_name)
var.get(NL)
assert var.get(CL).startswith(
"types"), f"error with class {class_name}"
var.get(NL)
assert var.get(CL).startswith(
"imp"), f"error with class {class_name}"
var.get(NL)
count -= 1
return r
def _parse_implemented_class() -> ObjectEntity:
var.get(NL)
name = None
base_method_list = []
base_properties = []
base_protocols = []
while True:
current_line = var.get(CL) #type:str
if (not name) and current_line.startswith("name"):
name = current_line.split(' ')[-1]
var.get(NL)
elif current_line.startswith("baseMethods"):
assert name, f"error with line {current_line}"
var.get(NL)
base_method_list += _parse_method_list(name)
elif current_line.startswith("baseProperties"):
assert name, f"error with line {current_line}"
var.get(NL)
base_properties += _parse_property_list(name)
elif current_line.startswith(data_baseProtocols):
var.get(NL)
base_protocols = _parse_protocol_list(name)
elif new_class_pattern.match(
current_line) or new_section_pattern.match(current_line):
break
else:
var.get(NL)
return ObjectEntity(name=name,
base_method_list=base_method_list,
base_property_list=base_properties,
base_protocol_list=base_protocols)
def _parse_protocol_list(class_name: object) -> object:
current_line = var.get(CL)
r = []
if not current_line.startswith(data_baseProtocols_count):
return r
count = int(current_line.split(' ')[-1])
var.get(NL)
while count > 0:
var.get(NL)
var.get(NL)
current_line = var.get(CL)
assert current_line.startswith(data_baseProtocols_name)
r.append(current_line.split(' ')[-1])
while not current_line.startswith(
data_baseProtocols_instanceProperties):
var.get(NL)
current_line = var.get(CL)
var.get(NL)
current_line = var.get(CL)
count -= 1
assert current_line.startswith(ivars)
return r
def _parse_property_list(class_name) -> List:
current_line = var.get(CL)
r = []
if current_line.startswith("entsize"):
var.get(NL)
current_line = var.get(CL)
assert current_line.startswith("count")
count = int(current_line.split(" ")[-1])
var.get(NL)
while count > 0:
current_line = var.get(CL)
assert current_line.startswith(
"name"), f"error with class {class_name}"
property_name = current_line.split(' ')[-1]
r.append(property_name)
var.get(NL)
assert var.get(CL).startswith(
"attributes"), f"error with class {class_name}"
var.get(NL)
count -= 1
return r