Replace pyopenssl with cryptography module and add parsing of SAN extension

CHANGES.rst

@@ -1,5 +1,8 @@
 Changelog
 =========
+* next release
+    * Replaced pyOpenSSL with cryptography
+    * Added parsing of subjectAltName X509 extension
 * 1.1.11
     * Added first version of WiFi results    
 * 1.1.10

README.rst

@@ -36,8 +36,7 @@ Troubleshooting
 
 Some setups (like MacOS) have trouble with building the dependencies required
 for reading SSL certificates. If you don't care about SSL stuff and only want to
-use sagan to say, parse traceroute or DNS results, then you can tell the
-installer to skip building ``pyOpenSSL`` by doing the following:
+use sagan to say, parse traceroute or DNS results, then you can do the following:
 
 .. code:: bash
 
@@ -175,9 +174,10 @@ What it requires
 As you might have guessed, with all of this magic going on under the hood, there
 are a few dependencies:
 
--  `pyOpenSSL`_ (Optional: see "Troubleshooting" above)
+-  `cryptography`_ (Optional: see "Troubleshooting" above)
 -  `python-dateutil`_
 -  `pytz`_
+-  `IPy`_
 
 Additionally, we recommend that you also install `ujson`_ as it will speed up
 the JSON-decoding step considerably, and `sphinx`_ if you intend to build the
@@ -229,9 +229,9 @@ But why "`Sagan`_"? The RIPE Atlas team decided to name all of its modules after
 explorers, and what better name for a parser than that of the man who spent
 decades reaching out to the public about the wonders of the cosmos?
 
-.. _pyOpenSSL: https://pypi.python.org/pypi/pyOpenSSL
 .. _python-dateutil: https://pypi.python.org/pypi/python-dateutil
 .. _pytz: https://pypi.python.org/pypi/pytz
+.. _IPy: https://pypi.python.org/pypi/IPy/
 .. _ujson: https://pypi.python.org/pypi/ujson
 .. _sphinx: https://pypi.python.org/pypi/Sphinx
 .. _Read the Docs: http://ripe-atlas-sagan.readthedocs.org/en/latest/

docs/installation.rst

@@ -11,15 +11,15 @@ Requirements
 As you might have guessed, with all of the magic going on under the hood, there
 are a few dependencies:
 
-* `pyOpenSSL`_
+* `cryptography`_
 * `python-dateutil`_
 * `pytz`_
 
 Additionally, we recommend that you also install `ujson`_ as it will speed up
 the JSON-decoding step considerably, and `sphinx`_ if you intend to build the
 documentation files for offline use.
 
-.. _pyOpenSSL: https://pypi.python.org/pypi/pyOpenSSL/
+.. _cryptography: https://pypi.python.org/pypi/cryptography
 .. _python-dateutil: https://pypi.python.org/pypi/python-dateutil/
 .. _pytz: https://pypi.python.org/pypi/pytz/
 .. _ujson: https://pypi.python.org/pypi/ujson/
@@ -75,8 +75,7 @@ Troubleshooting
 
 Some setups (like MacOS) have trouble with building the dependencies required
 for reading SSL certificates.  If you don't care about SSL stuff and only want
-to use sagan to say, parse traceroute or DNS results, then you can tell the
-installer to skip building ``pyOpenSSL`` by doing the following::
+to use sagan to say, parse traceroute or DNS results, then you can do the following:
 
      $ SAGAN_WITHOUT_SSL=1 pip install ripe.atlas.sagan
 

docs/types.rst

@@ -696,6 +696,7 @@ checksum_md5           str       The md5 checksum
 checksum_sha1          str       The sha1 checksum
 checksum_sha256        str       The sha256 checksum
 has_expired            bool      Set to ``True`` if the certificate is no longer valid
+extensions             dict      Parsed extensions. For now it can only be subjectAltName, which is a list of names contained in the SAN extension, if that exists.
 =====================  ========  ===================================================================================
 
 

ripe/atlas/sagan/ssl.py

@@ -15,16 +15,17 @@
 
 import logging
 import pytz
-import re
+import codecs
 
 from datetime import datetime
-from dateutil.relativedelta import relativedelta
 
 try:
-    import OpenSSL
+    from cryptography import x509
+    from cryptography.hazmat.backends import openssl
+    from cryptography.hazmat.primitives import hashes
 except ImportError:
     logging.warning(
-        "pyOpenSSL is not installed, without it you cannot parse SSL "
+        "cryptography module is not installed, without it you cannot parse SSL "
         "certificate measurement results"
     )
 
@@ -33,11 +34,6 @@
 
 class Certificate(ParsingDict):
 
-    TIME_FORMAT = "%Y%m%d%H%M%SZ"
-    TIME_REGEX = re.compile(
-        "(\d\d\d\d)(\d\d)(\d\d)(\d\d)(\d\d)(\d\d)(\+|\-)(\d\d)(\d\d)"
-    )
-
     def __init__(self, data, **kwargs):
 
         ParsingDict.__init__(self, **kwargs)
@@ -49,6 +45,7 @@ def __init__(self, data, **kwargs):
         self.issuer_cn = None
         self.issuer_o = None
         self.issuer_c = None
+
         self.valid_from = None
         self.valid_until = None
 
@@ -58,32 +55,59 @@ def __init__(self, data, **kwargs):
 
         self.has_expired = None
 
-        # Clean up the certificate data and use OpenSSL to parse it
-        x509 = OpenSSL.crypto.load_certificate(
-            OpenSSL.crypto.FILETYPE_PEM,
-            data.replace("\\/", "/").replace("\n\n", "\n")
-        )
-        subject = dict(x509.get_subject().get_components())
-        issuer = dict(x509.get_issuer().get_components())
+        self.extensions = {}
+
+        cert = x509.load_pem_x509_certificate(data.encode('ascii'), openssl.backend)
+
+        if cert:
+            self.checksum_md5 = self._colonify(cert.fingerprint(hashes.MD5()))
+            self.checksum_sha1 = self._colonify(cert.fingerprint(hashes.SHA1()))
+            self.checksum_sha256 = self._colonify(cert.fingerprint(hashes.SHA256()))
+
+            self.valid_from = pytz.utc.localize(cert.not_valid_before)
+            self.valid_until = pytz.utc.localize(cert.not_valid_after)
 
-        if x509 and subject and issuer:
+            self.has_expired = self._has_expired()
 
-            self.subject_cn = self._string_from_dict_or_none(subject, b"CN")
-            self.subject_o = self._string_from_dict_or_none(subject, b"O")
-            self.subject_c = self._string_from_dict_or_none(subject, b"C")
-            self.issuer_cn = self._string_from_dict_or_none(issuer, b"CN")
-            self.issuer_o = self._string_from_dict_or_none(issuer, b"O")
-            self.issuer_c = self._string_from_dict_or_none(issuer, b"C")
+            self._add_extensions(cert)
 
-            self.checksum_md5 = x509.digest("md5").decode()
-            self.checksum_sha1 = x509.digest("sha1").decode()
-            self.checksum_sha256 = x509.digest("sha256").decode()
+        if cert and cert.subject:
+            self.subject_cn, self.subject_o, self.subject_c = \
+                self._parse_x509_name(cert.subject)
 
-            self.has_expired = bool(x509.has_expired())
+        if cert and cert.issuer:
+            self.issuer_cn, self.issuer_o, self.issuer_c = \
+                self._parse_x509_name(cert.issuer)
 
-            self.valid_from = None
-            self.valid_until = None
-            self._process_validation_times(x509)
+    def _add_extensions(self, cert):
+        for ext in cert.extensions:
+            if ext.oid._name == 'subjectAltName':
+                self.extensions['subjectAltName'] = []
+                for san in ext.value:
+                    self.extensions['subjectAltName'].append(san.value)
+
+    @staticmethod
+    def _colonify(bytes):
+        hex = codecs.getencoder('hex_codec')(bytes)[0].decode('ascii').upper()
+        return ':'.join(a+b for a,b in zip(hex[::2], hex[1::2]))
+
+    @staticmethod
+    def _parse_x509_name(name):
+        cn = None
+        o = None
+        c = None
+        for attr in name:
+            if attr.oid.dotted_string == '2.5.4.6': # country
+                c = attr.value
+            elif attr.oid.dotted_string == '2.5.4.10': # organisation
+                o = attr.value
+            elif attr.oid.dotted_string == '2.5.4.3': # common name
+                cn = attr.value
+        return cn, o, c
+
+    def _has_expired(self):
+        now = pytz.utc.localize(datetime.utcnow())
+        return self.valid_from <= now <= self.valid_until
 
     @property
     def cn(self):
@@ -113,72 +137,6 @@ def country(self):
     def checksum(self):
         return self.checksum_sha256
 
-    def _process_validation_times(self, x509):
-        """
-        PyOpenSSL uses a kooky date format that *usually* parses out quite
-        easily but on the off chance that it's not in UTC, a lot of work needs
-        to be done.
-        """
-
-        valid_from = x509.get_notBefore()
-        valid_until = x509.get_notAfter()
-
-        try:
-            self.valid_from = pytz.UTC.localize(datetime.strptime(
-                valid_from.decode(),
-                self.TIME_FORMAT
-            ))
-        except ValueError:
-            self.valid_from = self._process_nonstandard_time(valid_from)
-
-        try:
-            self.valid_until = pytz.UTC.localize(datetime.strptime(
-                valid_until.decode(),
-                self.TIME_FORMAT
-            ))
-        except ValueError:
-            self.valid_until = self._process_nonstandard_time(valid_until)
-
-    def _process_nonstandard_time(self, string):
-        """
-        In addition to `YYYYMMDDhhmmssZ`, PyOpenSSL can also use timestamps
-        in `YYYYMMDDhhmmss+hhmm` or `YYYYMMDDhhmmss-hhmm`.
-        """
-
-        match = re.match(self.TIME_REGEX, string)
-
-        if not match:
-            raise ResultParseError(
-                "Unrecognised time format: {s}".format(s=string)
-            )
-
-        r = datetime(
-            int(match.group(1)),
-            int(match.group(2)),
-            int(match.group(3)),
-            int(match.group(4)),
-            int(match.group(5)),
-            int(match.group(6)),
-            0,
-            pytz.UTC
-        )
-        delta = relativedelta(
-            hours=int(match.group(8)),
-            minutes=int(match.group(9))
-        )
-        if match.group(7) == "-":
-            return r - delta
-        return r + delta
-
-    @staticmethod
-    def _string_from_dict_or_none(dictionary, key):
-        """
-        Created to make nice with the Python3 problem.
-        """
-        if key not in dictionary:
-            return None
-        return dictionary[key].decode("UTF-8")
-
 
 class Alert(ParsingDict):
 

setup.py

@@ -6,15 +6,15 @@
 
 name = "ripe.atlas.sagan"
 install_requires = [
+    "IPy",
     "python-dateutil",
     "pytz",
 ]
 
 tests_require = ["nose"]
 
-# pyOpenSSL support is flaky on some systems (I'm looking at you Apple)
 if "SAGAN_WITHOUT_SSL" not in os.environ:
-    install_requires.append("pyOpenSSL")
+    install_requires.append("cryptography")
 
 # Allow setup.py to be run from any path
 os.chdir(os.path.normpath(os.path.join(os.path.abspath(__file__), os.pardir)))

tests/ssl.py

@@ -386,3 +386,9 @@ def test_alert():
     result = Result.get('{"af":4,"cert":["-----BEGIN CERTIFICATE-----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\\nbMZdcMur9E+ovrS58lIKGCvDEPSUDXHzr0tpb4A13TTnxW6pclqUyJk=\\n-----END CERTIFICATE-----"],"dst_addr":"80.79.115.54","dst_name":"pretical.ee","dst_port":"https","from":"77.95.64.18","fw":4480,"method":"SSL","msm_id":1006864,"prb_id":517,"src_addr":"77.95.64.18","timestamp":1362454627,"type":"sslcert","ver":"3.0"}')
     assert(result.alert is None)
     assert(result.is_error is False)
+
+def test_san_extension():
+    result = Result.get('{"af":4,"cert":["-----BEGIN CERTIFICATE-----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\nrum/OChG\n-----END CERTIFICATE-----"],"dst_addr":"193.0.6.158","dst_name":"atlas.ripe.net","dst_port":"443","from":"86.82.178.27","fw":4760,"lts":133,"method":"TLS","msm_id":14002,"msm_name":"SSLCert","prb_id":10951,"rt":51.558465,"src_addr":"192.168.180.22","timestamp":1490659208,"ttc":14.88238,"type":"sslcert","ver":"1.2"}')
+    ext = result.certificates[0].extensions
+    assert(ext and len(ext['subjectAltName'])==54)
+