def _check_test(self, file_type):
"""Helper function to parse a binary file and check whether
the given string is in the parsed result"""
with tempfile.NamedTemporaryFile("w+b", suffix=file_type,
delete=False) as f:
if file_type == "out":
# write magic signature
f.write(b"\x7f\x45\x4c\x46\x02\x01\x01\x03\n")
f.seek(0)
self.assertTrue(is_binary(f.name))
else:
f.write(b"some other data\n")
f.seek(0)
self.assertFalse(is_binary(f.name))
os.remove(f.name)
def is_executable(self, filename):
"""check if file is an ELF binary file"""
output = None
if inpath("file"):
# use system file if available (for performance reasons)
output = subprocess.check_output(["file", filename])
output = output.decode(sys.stdout.encoding)
if "cannot open" in output:
self.logger.warning(
f"Unopenable file {filename} cannot be scanned")
return False, None
if (("LSB " not in output) and ("LSB shared" not in output)
and ("LSB executable" not in output)
and ("PE32 executable" not in output)
and ("PE32+ executable" not in output)
and ("Mach-O" not in output)
and ("PKG-INFO: " not in output)
and ("METADATA: " not in output)
and ("pom.xml" not in output)):
return False, None
# otherwise use python implementation of file
elif not is_binary(filename):
return False, None
return True, output
def scan_file(self, filename):
"""Scans a file to see if it contains any of the target libraries,
and whether any of those contain CVEs"""
self.logger.debug(f"Scanning file: {filename}")
self.total_scanned_files += 1
# Do not try to scan symlinks
if os.path.islink(filename):
return None
# Ensure filename is a file
if not os.path.isfile(filename):
self.logger.warning(f"Invalid file {filename} cannot be scanned")
return None
# step 1: check if it's an ELF binary file
if inpath("file"):
# use system file if available (for performance reasons)
o = subprocess.check_output(["file", filename])
o = o.decode(sys.stdout.encoding)
if "cannot open" in o:
self.logger.warning(
f"Unopenable file {filename} cannot be scanned")
return None
if (("LSB " not in o) and ("LSB shared" not in o)
and ("LSB executable" not in o)
and ("PE32 executable" not in o)
and ("PE32+ executable" not in o) and ("Mach-O" not in o)):
return None
# otherwise use python implementation of file
elif not is_binary(filename):
return None
# parse binary file's strings
if inpath("strings"):
# use "strings" on system if available (for performance)
o = subprocess.check_output(["strings", filename])
lines = o.decode("utf-8").splitlines()
else:
# Otherwise, use python implementation
s = Strings(filename)
lines = s.parse()
yield from self.run_checkers(filename, lines)
def test_windows(self):
"""file single-byte"""
self.assertTrue(is_binary(os.path.join(ASSETS_PATH, "windows.txt")))
def test_single_byte_file(self):
"""file single-byte"""
self.assertFalse(
is_binary(os.path.join(ASSETS_PATH, "single-byte.txt")))
def _check_test(self, filename, is_executable):
"""Helper function to parse a binary file and check whether
the given string is in the parsed result"""
self.assertEqual(is_binary(os.path.join(BINARIES_PATH, filename)),
is_executable)
