def get_privacy(self, tag_list):
api_names = []
api_descriptions = []
api2signature = {}
tp = 0
fp = 0
pkg_name = ""
classname = ""
for i in range(0, len(tag_list)):
tag = tag_list[i]
if tag.name == 'h2':
pkg_name = tag_list[i - 1].getText()
classname = tag.getText().split(" ")[-1]
is_method_section = False
if tag.name == 'h3':
des_text = tag.getText()
if "Public Member Functions" in des_text or "Method Detail" in des_text or "方法详细资料" in des_text or "メソッドの詳細" in des_text:
is_method_section = True
if not is_method_section:
continue
for j in range(i + 1, len(tag_list)):
next_tag = tag_list[j]
if next_tag.name == 'h4':
if j + 1 >= len(tag_list):
break
api_name = next_tag.getText()
pre_tag = tag_list[j + 1]
if pre_tag.name == "pre":
signature = pre_tag.getText()
else:
continue
api_names.append(api_name)
self.apis.append(api_name)
api2signature[api_name] = signature
if j + 2 < len(tag_list) and tag_list[j + 2].name == "div":
description = tag_list[j + 2].getText()
# print("*******************")
# print(api_name)
# print(description)
# print("===================")
api_descriptions.append(description)
else:
api_descriptions.append("")
for i in range(0, len(api_names)):
api_name = api_names[i]
api_description = api_descriptions[i]
is_sensitive, privacy_item = check_api_by_class(classname, api_name)
if is_sensitive:
api_description = api_description.replace("\n", " ")
self.sensitive_apis.append((pkg_name + "." + classname, api_name, privacy_item, api_description))
if pkg_name == "":
logger.error(classname)
fp = fp + 1
continue
tp = tp + 1
return tp, fp
def get_privacy(self, tag_list):
api_names = []
api_descriptions = []
tp = 0
fp = 0
for tag in tag_list:
if tag.name == 'table' and 'class' in tag.attrs.keys(
) and 'jd-inheritance-table' in tag.attrs['class']:
td_list = tag.find_all("td")
self.processing_class = td_list[-1].getText()
break
for i in range(0, len(tag_list)):
tag = tag_list[i]
is_method_section = False
if tag.name == 'h2':
des_text = tag.getText()
if "Public Methods" in des_text:
is_method_section = True
if not is_method_section:
continue
for j in range(i + 1, len(tag_list)):
next_tag = tag_list[j]
if next_tag.name == 'span' and 'class' in next_tag.attrs.keys(
) and 'sympad' in next_tag.attrs['class']:
if j + 1 >= len(tag_list):
break
api_name = next_tag.getText()
api_names.append(api_name)
if tag_list[j +
5].name == "div" and 'class' in tag_list[
j + 5].attrs.keys(
) and 'jd-tagdescr' in tag_list[
j + 5].attrs['class']:
api_description = tag_list[j + 5].getText()
api_descriptions.append(api_description)
else:
api_descriptions.append("")
for i in range(len(api_names)):
api = api_names[i]
self.apis.add(api)
api_description = api_descriptions[i]
is_sensitive, privacy_item = check_api_by_class(
self.processing_class, api)
if is_sensitive:
api_description = api_description.replace("\n", " ")
self.sensitive_apis.append((self.processing_class, api,
privacy_item, api_description))
fp = fp + 1
continue
tp = tp + 1
return tp, fp
def get_privacy(self, tag_list):
api2des = {}
tp = 0
fp = 0
method_tag = False
full_class_name = ""
for tag in tag_list:
if tag.name == "li":
txt = tag.getText()
if "com.facebook" in txt:
full_class_name = tag.getText()
for i in range(0, len(tag_list)):
tag = tag_list[i] # tag is a table class used to parse method.
if tag.name == "span" and "Method Summary" in tag.getText():
method_table = tag_list[i + 1].find_all()
first = True
for row in method_table:
if row.name != "tr":
continue
if first:
first = False
continue
columns = row.find_all()
for column in columns:
if column.name == "td":
api = column.getText()
left = api.find("(")
if left != -1:
self.apis.append(api[:left])
if "Method Detail" in tag.getText():
method_tag = True
if method_tag and tag.name == "h4" and tag.getText() in self.apis:
if len(tag_list[i + 3].getText()) > len(
tag_list[i + 2].getText()):
des = tag_list[i + 3].getText()
else:
des = tag_list[i + 2].getText()
api2des[tag.getText()] = des
for api_name, description in api2des.items():
is_sensitive, privacy_item = check_api_by_class(
full_class_name, api_name)
if is_sensitive:
description = description.replace("\n", " ")
self.sensitive_apis.append(
(full_class_name, api_name, description, privacy_item))
fp = fp + 1
continue
tp = tp + 1
return tp, fp
def get_privacy(self, tag_list):
api_names = set()
# api2des = {}
api2signature = {}
tp = 0
fp = 0
for i in range(0, len(tag_list)):
tag = tag_list[i]
is_method_section = False
if tag.name == 'h2':
# print(tag.getText())
des_text = tag.getText()
if "Methods" in des_text:
is_method_section = True
if not is_method_section:
continue
for j in range(i + 1, len(tag_list)):
next_tag = tag_list[j]
if next_tag.name == "dl" and "class" in next_tag.attrs.keys(
) and "method" in next_tag.attrs["class"]:
if j + 1 >= len(tag_list):
break
signature = tag_list[j + 1]["id"]
api_name = signature.split(".")[-1]
api_name = api_name[:api_name.rfind("(")]
api_names.add(api_name)
self.apis.append(api_name)
api2signature[api_name] = signature
# if j + 2 < len(tag_list) and tag_list[j + 2].name == "div":
# description = tag_list[j + 2].getText()
# api2des[api_name] = description
# print("first=" + str(len(self.apis)))
# print("second=" + str(len(api2des)))
for api in api_names:
is_sensitive, privacy_item = check_api_by_class(
self.processing_class, api)
if is_sensitive:
self.sensitive_apis.append(
(self.processing_class, api, privacy_item))
fp = fp + 1
continue
tp = tp + 1
return tp, fp
def run(self):
for method in self.methods:
clazz = method.cls
if "javax/" in clazz.package_name or "java/" in clazz.package_name or "android/" in clazz.package_name or "org/w3c" in clazz.package_name:
continue
if "<init>" == method.name or "<clinit>" == method.name or "toString" == method.name or "clone" == method.name:
continue
# Naive rules to filter obfuscated identifiers
# if len(method.name) == 1:
# continue
# para_list = method.prototype.parameters_type
# if len(para_list) == 2:
# if para_list[0] == "Ljava/lang/String;" and para_list[1] == "Ljava/lang/String;":
# print(method)
self.apis.append(method.name)
is_sensitive, privacy_item = check_api_by_class(
method.cls.fullname, method.name)
if is_sensitive:
self.sensitive_apis.append(
[method.cls.fullname, method.name, privacy_item])
def get_privacy(self, tag_list):
api_names = []
api_descriptions = []
api2signature = {}
tp = 0
fp = 0
self.processing_class = self.processing_class[5:-5]
self.processing_class = self.processing_class.replace("_1_1", ".")
class_info = ""
tag = False
for ch in self.processing_class:
if tag:
class_info = class_info + ch.upper()
tag = False
else:
if ch != "_":
class_info = class_info + ch
else:
tag = True
class_info = class_info.strip()
classname = class_info.split(".")[-1]
pkg_name = class_info[:-(len(classname) + 1)]
for i in range(0, len(tag_list)):
tag = tag_list[i]
# if "Class Reference" in tag.getText() and tag.name == "title":
# print(tag.getText().split(":")[1].strip().split(" ")[0] + "__" + tag.name)
# pkg_name = tag_list[i - 1].getText()
# classname = tag.getText().split(" ")[-1]
is_method_section = False
if tag.name == 'h2':
des_text = tag.getText()
if "Member Function Documentation" in des_text:
is_method_section = True
if not is_method_section:
continue
for j in range(i + 1, len(tag_list)):
next_tag = tag_list[j]
if next_tag.name == "td" and "class" in next_tag.attrs.keys() and "memname" in next_tag.attrs["class"]:
# print(next_tag.getText())
api_name = str(next_tag.getText()).strip().split(" ")[-1]
if "." in api_name:
api_name = api_name.split(".")[-1]
# print(api_name)
# if pre_tag.name == "pre":
# signature = pre_tag.getText()
# else:
# continue
api_names.append(api_name)
self.apis.append(api_name)
# api2signature[api_name] = signature
# if j + 2 < len(tag_list) and tag_list[j + 2].name == "div":
# description = tag_list[j + 2].getText()
# # print("*******************")
# # print(api_name)
# # print(description)
# # print("===================")
# api_descriptions.append(description)
# else:
# api_descriptions.append("")
for i in range(0, len(api_names)):
api_name = api_names[i]
# api_description = api_descriptions[i]
is_sensitive, privacy_item = check_api_by_class(classname, api_name)
if is_sensitive:
# api_description = api_description.replace("\n", " ")
self.sensitive_apis.append((pkg_name + "." + classname, api_name, privacy_item))
# self.sensitive_apis.append((pkg_name + "." + classname, api_name, privacy_item, api_description))
if pkg_name == "":
logger.error(classname + " no Class!")
fp = fp + 1
continue
tp = tp + 1
return tp, fp