bazarr/libs/urllib3/util/ssl_match_hostname.py

"""The match_hostname() function from Python 3.3.3, essential when using SSL."""

# Note: This file is under the PSF license as the code comes from the python
# stdlib.   http://docs.python.org/3/license.html

import re
import sys

# ipaddress has been backported to 2.6+ in pypi.  If it is installed on the
# system, use it to handle IPAddress ServerAltnames (this was added in
# python-3.5) otherwise only do DNS matching.  This allows
# util.ssl_match_hostname to continue to be used in Python 2.7.
try:
    import ipaddress
except ImportError:
    ipaddress = None

__version__ = "3.5.0.1"


class CertificateError(ValueError):
    pass


def _dnsname_match(dn, hostname, max_wildcards=1):
    """Matching according to RFC 6125, section 6.4.3

    http://tools.ietf.org/html/rfc6125#section-6.4.3
    """
    pats = []
    if not dn:
        return False

    # Ported from python3-syntax:
    # leftmost, *remainder = dn.split(r'.')
    parts = dn.split(r".")
    leftmost = parts[0]
    remainder = parts[1:]

    wildcards = leftmost.count("*")
    if wildcards > max_wildcards:
        # Issue #17980: avoid denials of service by refusing more
        # than one wildcard per fragment.  A survey of established
        # policy among SSL implementations showed it to be a
        # reasonable choice.
        raise CertificateError(
            "too many wildcards in certificate DNS name: " + repr(dn)
        )

    # speed up common case w/o wildcards
    if not wildcards:
        return dn.lower() == hostname.lower()

    # RFC 6125, section 6.4.3, subitem 1.
    # The client SHOULD NOT attempt to match a presented identifier in which
    # the wildcard character comprises a label other than the left-most label.
    if leftmost == "*":
        # When '*' is a fragment by itself, it matches a non-empty dotless
        # fragment.
        pats.append("[^.]+")
    elif leftmost.startswith("xn--") or hostname.startswith("xn--"):
        # RFC 6125, section 6.4.3, subitem 3.
        # The client SHOULD NOT attempt to match a presented identifier
        # where the wildcard character is embedded within an A-label or
        # U-label of an internationalized domain name.
        pats.append(re.escape(leftmost))
    else:
        # Otherwise, '*' matches any dotless string, e.g. www*
        pats.append(re.escape(leftmost).replace(r"\*", "[^.]*"))

    # add the remaining fragments, ignore any wildcards
    for frag in remainder:
        pats.append(re.escape(frag))

    pat = re.compile(r"\A" + r"\.".join(pats) + r"\Z", re.IGNORECASE)
    return pat.match(hostname)


def _to_unicode(obj):
    if isinstance(obj, str) and sys.version_info < (3,):
        # ignored flake8 # F821 to support python 2.7 function
        obj = unicode(obj, encoding="ascii", errors="strict")  # noqa: F821
    return obj


def _ipaddress_match(ipname, host_ip):
    """Exact matching of IP addresses.

    RFC 6125 explicitly doesn't define an algorithm for this
    (section 1.7.2 - "Out of Scope").
    """
    # OpenSSL may add a trailing newline to a subjectAltName's IP address
    # Divergence from upstream: ipaddress can't handle byte str
    ip = ipaddress.ip_address(_to_unicode(ipname).rstrip())
    return ip == host_ip


def match_hostname(cert, hostname):
    """Verify that *cert* (in decoded format as returned by
    SSLSocket.getpeercert()) matches the *hostname*.  RFC 2818 and RFC 6125
    rules are followed, but IP addresses are not accepted for *hostname*.

    CertificateError is raised on failure. On success, the function
    returns nothing.
    """
    if not cert:
        raise ValueError(
            "empty or no certificate, match_hostname needs a "
            "SSL socket or SSL context with either "
            "CERT_OPTIONAL or CERT_REQUIRED"
        )
    try:
        # Divergence from upstream: ipaddress can't handle byte str
        host_ip = ipaddress.ip_address(_to_unicode(hostname))
    except ValueError:
        # Not an IP address (common case)
        host_ip = None
    except UnicodeError:
        # Divergence from upstream: Have to deal with ipaddress not taking
        # byte strings.  addresses should be all ascii, so we consider it not
        # an ipaddress in this case
        host_ip = None
    except AttributeError:
        # Divergence from upstream: Make ipaddress library optional
        if ipaddress is None:
            host_ip = None
        else:
            raise
    dnsnames = []
    san = cert.get("subjectAltName", ())
    for key, value in san:
        if key == "DNS":
            if host_ip is None and _dnsname_match(value, hostname):
                return
            dnsnames.append(value)
        elif key == "IP Address":
            if host_ip is not None and _ipaddress_match(value, host_ip):
                return
            dnsnames.append(value)
    if not dnsnames:
        # The subject is only checked when there is no dNSName entry
        # in subjectAltName
        for sub in cert.get("subject", ()):
            for key, value in sub:
                # XXX according to RFC 2818, the most specific Common Name
                # must be used.
                if key == "commonName":
                    if _dnsname_match(value, hostname):
                        return
                    dnsnames.append(value)
    if len(dnsnames) > 1:
        raise CertificateError(
            "hostname %r "
            "doesn't match either of %s" % (hostname, ", ".join(map(repr, dnsnames)))
        )
    elif len(dnsnames) == 1:
        raise CertificateError("hostname %r doesn't match %r" % (hostname, dnsnames[0]))
    else:
        raise CertificateError(
            "no appropriate commonName or subjectAltName fields were found"
        )