def run_python_package_checkers(self, filename, lines):
"""
This generator runs only for python packages.
There are no actual checkers.
The ProductInfo is computed without the help of any checkers from PKG-INFO or METADATA.
"""
try:
product = search(compile(r"^Name: (.+)$", MULTILINE),
lines).group(1)
version = search(compile(r"^Version: (.+)$", MULTILINE),
lines).group(1)
cve_db = CVEDB()
vendor_package_pair = cve_db.get_vendor_product_pairs(product)
if vendor_package_pair != []:
vendor = vendor_package_pair[0]["vendor"]
file_path = "".join(self.file_stack)
self.logger.info(f"{file_path} is {product} {version}")
yield ProductInfo(vendor, product, version), file_path
# There are packages with a METADATA file in them containing different data from what the tool expects
except AttributeError:
self.logger.debug(f"{filename} is an invalid METADATA/PKG-INFO")
self.logger.debug(f"Done scanning file: {filename}")
class SBOMManager:
SBOMtype = ["spdx", "cyclonedx", "swid"]
sbom_data: DefaultDict[ProductInfo, TriageData]
def __init__(
self, filename: str, sbom_type: str = "spdx", logger: Optional[Logger] = None
):
self.filename = filename
self.sbom_data = defaultdict(dict)
self.type = "unknown"
if sbom_type in self.SBOMtype:
self.type = sbom_type
self.logger = logger or LOGGER.getChild(self.__class__.__name__)
# Connect to the database
self.cvedb = CVEDB(version_check=False)
def scan_file(self) -> Dict[ProductInfo, TriageData]:
LOGGER.info(f"Processing SBOM {self.filename} of type {self.type.upper()}")
try:
if self.type == "spdx":
spdx = SPDXParser()
modules = spdx.parse(self.filename)
elif self.type == "cyclonedx":
cyclone = CycloneParser()
modules = cyclone.parse(self.filename)
elif self.type == "swid":
swid = SWIDParser()
modules = swid.parse(self.filename)
else:
modules = []
except (KeyError, FileNotFoundError, ET.ParseError) as e:
LOGGER.debug(e, exc_info=True)
modules = []
LOGGER.debug(
f"The number of modules identified in SBOM - {len(modules)}\n{modules}"
)
# Now process list of modules to create [vendor, product, version] tuples
parsed_data: List[ProductInfo] = []
for m in modules:
product, version = m[0], m[1]
if version != "":
# Now add vendor to create product record....
# print (f"Find vendor for {product} {version}")
vendor = self.get_vendor(product)
if vendor is not None:
parsed_data.append(ProductInfo(vendor, product, version))
# print(vendor,product,version)
for row in parsed_data:
self.sbom_data[row]["default"] = {
"remarks": Remarks.NewFound,
"comments": "",
"severity": "",
}
self.sbom_data[row]["paths"] = set(map(lambda x: x.strip(), "".split(",")))
LOGGER.debug(f"SBOM Data {self.sbom_data}")
return self.sbom_data
def get_vendor(self, product: str) -> Optional[str]:
vendor_package_pair = self.cvedb.get_vendor_product_pairs(product)
if vendor_package_pair != []:
vendor = vendor_package_pair[0]["vendor"]
return vendor
return None
def parse_list(self):
input_file = self.input_file
self.check_file()
if not input_file.endswith("requirements.txt"):
if distro.id() not in SUPPORTED_DISTROS:
LOGGER.warning(
f"Package list support only available on {','.join(SUPPORTED_DISTROS)}!"
)
return {}
system_packages = []
LOGGER.info(f"Scanning {distro.id().capitalize()} package list.")
if distro.id() in DEB_DISTROS:
installed_packages = run(
[
"dpkg-query",
"--show",
'--showformat={"name": "${binary:Package}", "version": "${Version}"}, ',
],
stdout=PIPE,
)
installed_packages = json.loads(
f"[{installed_packages.stdout.decode('utf-8')[0:-2]}]")
elif distro.id() in RPM_DISTROS:
installed_packages = run(
[
"rpm",
"--query",
"--all",
"--queryformat",
'{"name": "%{NAME}", "version": "%{VERSION}"\\}, ',
],
stdout=PIPE,
)
installed_packages = json.loads(
f"[{installed_packages.stdout.decode('utf-8')[0:-2]}]")
elif distro.id() in PACMAN_DISTROS:
installed_packages = []
installed_packages_output = run(
["pacman", "--query", "--explicit"],
stdout=PIPE,
)
installed_packages_output = installed_packages_output.stdout.decode(
"utf-8").splitlines()
dict_keys = ["name", "version"]
for installed_package in installed_packages_output:
package_details = installed_package.split(" ")
installed_package_dict = dict(
zip(dict_keys, package_details))
installed_packages.append(installed_package_dict)
with open(input_file) as req:
lines = req.readlines()
for line in lines:
system_packages.append(re.split("\n", line)[0])
for installed_package in installed_packages:
if installed_package["name"] in system_packages:
self.package_names_without_vendor.append(installed_package)
else:
LOGGER.info("Scanning python package list.")
txt_package_names = []
installed_packages_json = run(
["pip", "list", "--format", "json"],
stdout=PIPE,
)
installed_packages = json.loads(
installed_packages_json.stdout.decode("utf-8"))
with open(input_file) as txtfile:
lines = txtfile.readlines()
for line in lines:
txt_package_names.append(re.split(">|\\[|;|=|\n", line)[0])
for installed_package in installed_packages:
package_name = installed_package["name"].lower()
if package_name in txt_package_names:
self.package_names_without_vendor.append(
installed_package)
cve_db = CVEDB()
vendor_package_pairs = cve_db.get_vendor_product_pairs(
self.package_names_without_vendor)
self.add_vendor(vendor_package_pairs)
self.parse_data()
return self.parsed_data_with_vendor
class VersionScanner:
""" "Scans files for CVEs using CVE checkers"""
CHECKER_ENTRYPOINT = "cve_bin_tool.checker"
def __init__(
self,
should_extract=False,
exclude_folders=[],
checkers=None,
logger=None,
error_mode=ErrorMode.TruncTrace,
score=0,
):
self.logger = logger or LOGGER.getChild(self.__class__.__name__)
# Update egg if installed in development mode
if IS_DEVELOP():
self.logger.debug("Updating egg_info")
update_egg()
# Load checkers if not given
self.checkers = checkers or self.load_checkers()
self.score = score
self.total_scanned_files = 0
self.exclude_folders = exclude_folders + [".git"]
self.walker = DirWalk(folder_exclude_pattern=";".join(
exclude if exclude.endswith("*") else exclude + "*"
for exclude in exclude_folders)).walk
self.should_extract = should_extract
self.file_stack = []
self.error_mode = error_mode
self.cve_db = CVEDB()
# self.logger.info("Checkers loaded: %s" % (", ".join(self.checkers.keys())))
@classmethod
def load_checkers(cls):
"""Loads CVE checkers"""
checkers = dict(
map(
lambda checker: (checker.name, checker.load()),
importlib_metadata.entry_points()[cls.CHECKER_ENTRYPOINT],
))
return checkers
@classmethod
def available_checkers(cls):
checkers = importlib_metadata.entry_points()[cls.CHECKER_ENTRYPOINT]
checker_list = [item.name for item in checkers]
return checker_list
def remove_skiplist(self, skips):
# Take out any checkers that are on the skip list
# (string of comma-delimited checker names)
skiplist = skips
for skipme in skiplist:
if skipme in self.checkers:
del self.checkers[skipme]
self.logger.debug(f"Skipping checker: {skipme}")
else:
self.logger.error(
f"Checker {skipme} is not a valid checker name")
def print_checkers(self):
self.logger.info(f'Checkers: {", ".join(self.checkers.keys())}')
def number_of_checkers(self):
return len(self.checkers)
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 parse_strings(self, filename):
"""parse binary file's strings"""
if inpath("strings"):
# use "strings" on system if available (for performance)
lines = subprocess.check_output(["strings",
filename]).decode("utf-8")
else:
# Otherwise, use python implementation
s = Strings(filename)
lines = s.parse()
return lines
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.debug(f"Invalid file {filename} cannot be scanned")
return None
# check if it's an ELF binary file
is_exec, output = self.is_executable(filename)
if not is_exec:
return None
# parse binary file's strings
lines = self.parse_strings(filename)
# Check for Java package
if output and "pom.xml" in output:
java_lines = "\n".join(lines.splitlines())
yield from self.run_java_checker(filename, java_lines)
# If python package then strip the lines to avoid detecting other product strings
if output and ("PKG-INFO: " in output or "METADATA: " in output):
py_lines = "\n".join(lines.splitlines()[:3])
yield from self.run_python_package_checkers(filename, py_lines)
yield from self.run_checkers(filename, lines)
def find_java_vendor(self, product, version):
"""Find vendor for Java product"""
vendor_package_pair = self.cve_db.get_vendor_product_pairs(product)
# If no match, try alternative product name.
# Apache product names are stored as A_B in NVD database but often called A-B
# Some packages have -parent appended to product which is not in NVD database
if vendor_package_pair == [] and "-" in product:
self.logger.debug(f"Try alternative product {product}")
# Remove parent appendage
if "-parent" in product:
product = product.replace("-parent", "")
product = product.replace("-", "_")
vendor_package_pair = self.cve_db.get_vendor_product_pairs(product)
if vendor_package_pair != []:
vendor = vendor_package_pair[0]["vendor"]
file_path = "".join(self.file_stack)
self.logger.debug(f"{file_path} {product} {version} by {vendor}")
return ProductInfo(vendor, product, version), file_path
return None, None
def run_java_checker(self, filename, lines):
"""Process maven pom.xml file and extract product and dependency details"""
tree = ET.parse(filename)
# Find root element
root = tree.getroot()
# Extract schema
schema = root.tag[:root.tag.find("}") + 1]
parent = root.find(schema + "parent")
version = None
product = None
file_path = "".join(self.file_stack)
# Parent tag is optional.
if parent is None:
product = root.find(schema + "artifactId").text
version = root.find(schema + "version").text
if version is None:
version = parent.find(schema + "version").text
# Check valid version identifier (i.e. starts with a digit)
if not version[0].isdigit():
self.logger.debug(f"Invalid {version} detected in {filename}")
version = None
if product is None:
product = parent.find(schema + "artifactId").text
if product is not None and version is not None:
product_info, file_path = self.find_java_vendor(product, version)
if file_path is not None:
yield product_info, file_path
# Scan for any dependencies referenced in file
dependencies = root.find(schema + "dependencies")
if dependencies is not None:
for dependency in dependencies.findall(schema + "dependency"):
product = dependency.find(schema + "artifactId")
if product is not None:
version = dependency.find(schema + "version")
if version is not None:
version = version.text
self.logger.debug(
f"{file_path} {product.text} {version}")
if version[0].isdigit():
# Valid version identifier
product_info, file_path = self.find_java_vendor(
product.text, version)
if file_path is not None:
yield product_info, file_path
self.logger.debug(f"Done scanning file: {filename}")
def run_python_package_checkers(self, filename, lines):
"""
This generator runs only for python packages.
There are no actual checkers.
The ProductInfo is computed without the help of any checkers from PKG-INFO or METADATA.
"""
try:
product = search(compile(r"^Name: (.+)$", MULTILINE),
lines).group(1)
version = search(compile(r"^Version: (.+)$", MULTILINE),
lines).group(1)
cve_db = CVEDB()
vendor_package_pair = cve_db.get_vendor_product_pairs(product)
if vendor_package_pair != []:
vendor = vendor_package_pair[0]["vendor"]
file_path = "".join(self.file_stack)
self.logger.info(f"{file_path} is {product} {version}")
yield ProductInfo(vendor, product, version), file_path
# There are packages with a METADATA file in them containing different data from what the tool expects
except AttributeError:
self.logger.debug(f"{filename} is an invalid METADATA/PKG-INFO")
self.logger.debug(f"Done scanning file: {filename}")
def run_checkers(self, filename, lines):
# tko
for (dummy_checker_name, checker) in self.checkers.items():
checker = checker()
result = checker.get_version(lines, filename)
# do some magic so we can iterate over all results, even the ones that just return 1 hit
if "is_or_contains" in result:
results = [dict()]
results[0] = result
else:
results = result
for result in results:
if "is_or_contains" in result:
version = "UNKNOWN"
if "version" in result and result["version"] != "UNKNOWN":
version = result["version"]
elif result["version"] == "UNKNOWN":
file_path = "".join(self.file_stack)
self.logger.debug(
f"{dummy_checker_name} was detected with version UNKNOWN in file {file_path}"
)
else:
self.logger.error(
f"No version info for {dummy_checker_name}")
if version != "UNKNOWN":
file_path = "".join(self.file_stack)
self.logger.debug(
f'{file_path} {result["is_or_contains"]} {dummy_checker_name} {version}'
)
for vendor, product in checker.VENDOR_PRODUCT:
yield ProductInfo(vendor, product,
version), file_path
self.logger.debug(f"Done scanning file: {filename}")
@staticmethod
def clean_file_path(filepath):
"""Returns a cleaner filepath by removing temp path from filepath"""
# we'll recieve a filepath similar to
# /temp/anything/extractable_filename.extracted/folders/inside/file
# We'll return /folders/inside/file to be scanned
# start_point is the point from we want to start trimming
# len("extracted") = 9
start_point = filepath.find("extracted") + 9
return filepath[start_point:]
def scan_and_or_extract_file(self, ectx, filepath):
"""Runs extraction if possible and desired otherwise scans."""
# Scan the file
yield from self.scan_file(filepath)
# Attempt to extract the file and scan the contents
if ectx.can_extract(filepath):
if not self.should_extract:
LOGGER.warning(
f"{filepath} is an archive. Pass -x option to auto-extract"
)
return None
for filename in self.walker([ectx.extract(filepath)]):
clean_path = self.clean_file_path(filename)
self.file_stack.append(f" contains {clean_path}")
yield from self.scan_and_or_extract_file(ectx, filename)
self.file_stack.pop()
def recursive_scan(self, scan_path):
with Extractor(logger=self.logger, error_mode=self.error_mode) as ectx:
if os.path.isdir(scan_path):
for filepath in self.walker([scan_path]):
self.file_stack.append(filepath)
yield from self.scan_and_or_extract_file(ectx, filepath)
self.file_stack.pop()
elif os.path.isfile(scan_path):
self.file_stack.append(scan_path)
yield from self.scan_and_or_extract_file(ectx, scan_path)
self.file_stack.pop()