def test_end2end(self):
"""Test pypi-scan analysis from start to finish."""
package_names = get_all_packages()
top_packages = get_top_packages()
squat_candidates = create_suspicious_package_dict(
package_names, top_packages)
store_squatting_candidates(squat_candidates)
def scan_recent(max_distance, save_new_list=False):
"""Scan packages recently added to pypi for possible typosquatting.
Print recently added packages and any package names on which these
packages are potentially typosquatting.
Args:
max_distance (int): maximum edit distance to check for typosquatting
save_new_list (bool): flag to save new list
"""
# Download current list of PyPI packages and convert to set
current_packages_set = set(get_all_packages())
# If saving is requested, save new list with timestamped name
if save_new_list == True:
store_recent_scan_results(list(current_packages_set))
# Load most recent stored list of PyPI packages and concert to set
recent_packages_set = load_most_recent_packages()
# Find packages that are in newest list but not old list
# This is a set operation: find all elements in first not in second
new_packages = current_packages_set - recent_packages_set
# Check each new package and see if it is a potential typosquatter
squat_candidates = create_suspicious_package_dict(
current_packages_set, new_packages, max_distance
)
# TODO: Consider adding in length to avoid checking short package names
print_suspicious_packages(squat_candidates)
def test_create_suspicious_package_dict(self):
"""Test create_suspicious_package_dict function"""
# Check if misspelling and confusion attacks are detected
ALL_PACKAGES = ["eeny", "meeny", "miny", "moe", "cup-joe", "joe-cup"]
TOP_PACKAGE = ["eeny", "cup-joe"]
MAX_DISTANCE = 1
output = create_suspicious_package_dict(ALL_PACKAGES, TOP_PACKAGE, MAX_DISTANCE)
expected_output = collections.OrderedDict(
{"eeny": ["meeny"], "cup-joe": ["joe-cup"]}
)
self.assertEqual(output, expected_output)
def top_mods(max_distance, top_n, min_len, stored_json):
"""Check top packages for typosquatters.
Prints top packages and any potential typosquatters
Args:
max_distance (int): maximum edit distance to check for typosquatting
top_n (int): the number of top packages to retrieve
min_len (int): a minimum length of characters
stored_json (bool): a flag to denote whether to used stored top packages json
"""
# Get list of potential typosquatters
package_names = get_all_packages()
top_packages = get_top_packages(top_n=top_n, stored=stored_json)
filtered_package_list = filter_by_package_name_len(top_packages, min_len=min_len)
squat_candidates = create_suspicious_package_dict(
package_names, filtered_package_list, max_distance
)
post_whitelist_candidates = whitelist(squat_candidates)
store_squatting_candidates(post_whitelist_candidates)
print_suspicious_packages(post_whitelist_candidates)
def mod_squatters(module, max_distance):
"""Check if a particular package name has potential squatters.
Prints any potential typosquatters for specified module
Args:
module (str): name to check for typosquatting
max_distance (int): maximum edit distance to check for typosquatting
"""
module_in_list = [module]
package_names = get_all_packages()
squat_candidates = create_suspicious_package_dict(
package_names, module_in_list, max_distance
)
# Print results
print("Checking " + module + " for typosquatting candidates.")
# Check for no typosquatting candidates
if len(squat_candidates[module]) == 0:
print("No typosquatting candidates found.")
else:
for i, candidate in enumerate(squat_candidates[module]):
print(str(i) + ": " + candidate)