summaryrefslogtreecommitdiff
path: root/openid2rp.py
blob: 63bb81579020822e949a7b22f56ed41be5a4db4f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
# -*- coding: utf-8 -*-
# OpenID relying party library
# Copyright Martin v. Löwis, 2009
# Licensed under the Academic Free License, version 3

# This library implements OpenID Authentication 2.0,
# in the role of a relying party

import urlparse, urllib, httplib, time, cgi, htmllib, formatter
import cStringIO, base64, hmac, hashlib, datetime, re, random
import itertools, cPickle, sys

try:
    from xml.etree import ElementTree
except ImportError:
    from elementtree import ElementTree

# Don't use urllib2, since it breaks in 2.5
# for https://login.launchpad.net//+xrds

# Don't use urllib, since it sometimes selects HTTP/1.1 (e.g. in PyPI)
# and then fails to parse chunked responses.

# 3.x portability

if sys.version_info < (3,):
    def b(s):
        return s
else:
    def b(s):
        return s.encode('ascii')

def normalize_uri(uri):
    """Normalize an uri according to OpenID section 7.2. Return a pair
    type,value, where type can be either 'xri' or 'uri'."""
    
    # 7.2 Normalization
    if uri.startswith('xri://'):
        uri = uri[6:]
    if uri[0] in ("=", "@", "+", "$", "!", ")"):
        return 'xri', uri
    if not uri.startswith('http'):
        uri = 'http://' + uri
    # RFC 3986, section 6

    # 6.2.2.1 case normalization
    parts = urlparse.urlparse(uri) # already lower-cases scheme
    if '@' in parts[1]: #netloc
        userinfo,hostname = parts[1].rsplit('@', 1)
    else:
        userinfo,hostname = None,parts[1]
    if ':' in hostname:
        host,port = hostname.rsplit(':', 1)
        if ']' in port:
            # IPv6
            host,port = hostname,None
    else:
        host,port = hostname,None
    netloc = hostname = host.lower()
    if port:
        netloc = hostname = host+':'+port
    if userinfo:
        netloc = userinfo + '@' + hostname
    parts = list(parts)
    parts[1] = netloc
    uri = urlparse.urlunparse(parts)

    # 6.2.2.2. normalize case in % escapes
    # XXX should restrict search to parts that can be pct-encoded
    for match in re.findall('%[0-9a-fA-F][0-9a-fA-F]', uri):
        m2 = match.upper()
        if m2 != match:
            uri = uri.replace(match, m2)

    # 6.2.2.3 remove dot segments
    parts = urlparse.urlparse(uri)
    path = parts[2] #path
    newpath = ''
    while path:
        if path.startswith('../'):
            path = path[3:]
        elif path.startswith('./'):
            path = path[2:]
        elif path.startswith('/./'):
            newpath += '/'; path = path[3:]
        elif path == '/.':
            newpath += '/'; path = ''
        elif path.startswith('/../'):
            newpath = newpath.rsplit('/', 1)[0]
            path = path[3:] # leave /
        elif path == '/..':
            newpath = newpath.rsplit('/', 1)[0]
            path = '/'
        elif path == '.' or path=='..':
            path = ''
        else:
            pos = path.find('/', 1)
            if pos == -1:
                pos = len(path)
            newpath += path[:pos]
            path = path[pos:]
    parts = list(parts)
    parts[2] = newpath
    uri = urlparse.urlunparse(parts)

    # 6.2.3 scheme based normalization

    parts = urlparse.urlparse(uri)    
    netloc = parts[1]
    if netloc.endswith(':'):
        netloc = netloc[:-1]
    elif parts[0] == 'http' and netloc.endswith(':80'):
        netloc = netloc[:-3]
    elif parts[0] == 'https' and netloc.endswith(':443'):
        netloc = netloc[:-4]
    # other default ports not considered here

    path = parts[2]
    if parts[0] in ('http', 'https') and parts[2]=='':
        path = '/'

    # 6.2.5 protocol-based normalization not done, as it
    # is not appropriate to resolve the URL just for normalization
    # it seems like a bug in the OpenID spec that it doesn't specify
    # which normalizations exactly should be performed

    parts = list(parts)
    parts[1] = netloc
    parts[2] = path
    return 'uri', urlparse.urlunparse(parts)


def parse_response(s):
    '''Parse a key-value form (OpenID section 4.1.1) into a dictionary'''
    res = {}
    for line in s.decode('utf-8').splitlines():
        k,v = line.split(':', 1)
        res[k] = v
    return res

class OpenIDParser(htmllib.HTMLParser):
    def __init__(self):
        htmllib.HTMLParser.__init__(self, formatter.NullFormatter())
        self.links = {}
        self.xrds_location=None

    def do_link(self, attrs):
        attrs = dict(attrs)
        self.links[attrs['rel']] = attrs['href']

    def do_meta(self, attrs):
        attrs = dict(attrs)
        # Yadis 6.2.5 option 1: meta tag
        if attrs['http-equiv'].lower() == 'x-xrds-location':
            self.xrds_location = attrs['content']

def discover(url):
    '''Perform service discovery on the OP URL.
    Return list of service types, and the auth/2.0 URL,
    or None if discovery fails.'''
    scheme, netloc, path, query, fragment = urlparse.urlsplit(url)
    assert not fragment
    if scheme == 'https':
        conn = httplib.HTTPSConnection(netloc)
    elif scheme == 'http':
        conn = httplib.HTTPConnection(netloc)
    else:
        raise ValueError, "Unsupported scheme "+scheme
    # conn.set_debuglevel(1)
    if query:
        path += '?'+query
    try:
        conn.connect()
    except:
        # DNS or TCP error
        return None
    # httplib in 2.5 incorrectly sends https port in Host
    # header even if it is 443
    conn.putrequest("GET", path, skip_host=1)
    conn.putheader('Host', netloc)
    conn.putheader('Accept', "text/html; q=0.3, "+
                   "application/xhtml+xml; q=0.5, "+
                   "application/xrds+xml")
    conn.endheaders()

    res = conn.getresponse()
    data = res.read()
    conn.close()

    if res.status in (301, 302, 303, 307):
        return discover(res.msg.get('location'))

    if sys.version_info < (3,0):
        content_type = res.msg.gettype()
    else:
        content_type = res.msg.get_content_type()

    # Yadis 6.2.5 option 2 and 3: header includes x-xrds-location
    xrds_loc = res.msg.get('x-xrds-location')
    if xrds_loc and content_type != 'application/xrds+xml':
        return discover(xrds_loc)

    if content_type in ('text/html', 'application/xhtml+xml'):
        parser = OpenIDParser()
        parser.feed(data)
        parser.close()
        # Yadis 6.2.5 option 1: meta tag
        if parser.xrds_location:
            return discover(parser.xrds_location)
        # OpenID 7.3.3: attempt html based discovery
        op_endpoint = parser.links.get('openid2.provider')
        if op_endpoint:
            op_local = parser.links.get('openid2.local_id')
            return ['http://specs.openid.net/auth/2.0/signon'], op_endpoint, op_local
        # 14.2.1: 1.1 compatibility
        op_endpoint = parser.links.get('openid.server')
        if op_endpoint:
            op_local = parser.links.get('openid.delegate')
            return ['http://openid.net/signon/1.1'], op_endpoint, op_local
        # Discovery failed
        return None

    elif content_type == 'application/xrds+xml':
        # Yadis 6.2.5 option 4
        doc = ElementTree.fromstring(data)
        for svc in doc.findall(".//{xri://$xrd*($v*2.0)}Service"):
            services = [x.text for x in svc.findall("{xri://$xrd*($v*2.0)}Type")]
            if 'http://specs.openid.net/auth/2.0/server' in services:
                # 7.3.2.1.1 OP Identifier Element
                uri = svc.find("{xri://$xrd*($v*2.0)}URI")
                if uri is not None:
                    op_local = None
                    op_endpoint = uri.text
                    break
            elif 'http://specs.openid.net/auth/2.0/signon' in services:
                # 7.3.2.1.2.  Claimed Identifier Element
                op_local = svc.find("{xri://$xrd*($v*2.0)}LocalID")
                if op_local is not None:
                    op_local = op_local.text
                uri = svc.find("{xri://$xrd*($v*2.0)}URI")
                if uri is not None:
                    op_endpoint = uri.text
                    break
            elif 'http://openid.net/server/1.0' in services or \
                 'http://openid.net/server/1.1' in services or \
                 'http://openid.net/signon/1.0' in services or \
                 'http://openid.net/signon/1.1' in services:
                # 14.2.1 says we also need to check for the 1.x types;
                # XXX should check 1.x only if no 2.0 service is found
                op_local = svc.find("{http://openid.net/xmlns/1.0}Delegate")
                if op_local is not None:
                    op_local = op_local.text
                uri = svc.find("{xri://$xrd*($v*2.0)}URI")
                if uri is not None:
                    op_endpoint = uri.text
                    break
        else:
            return None # No OpenID 2.0 service found
    else:
        # unknown content type
        return None
    return services, op_endpoint, op_local

def is_compat_1x(services):
    for uri in ('http://specs.openid.net/auth/2.0/signon',
                'http://specs.openid.net/auth/2.0/server'):
        if uri in services:
            return False
    for uri in ('http://openid.net/signon/1.0',
                'http://openid.net/signon/1.1',
                'http://openid.net/server/1.0',
                'http://openid.net/server/1.1'):
        if uri in services:
            return True
    raise ValueError, "Neither 1.x nor 2.0 service found"

def is_op_endpoint(services):
    for uri in ('http://specs.openid.net/auth/2.0/server',
                'http://openid.net/server/1.0',
                'http://openid.net/server/1.1'):
        if uri in services:
            return True
    return False
is_op_identifier = is_op_endpoint

# 4.1.3: Binary two's complement
def btwoc(l):
    res = cPickle.dumps(l, 2)
    # Pickle result: proto 2, long1 (integer < 256 bytes)
    # number of bytes, little-endian integer, stop
    assert res[:3] == '\x80\x02\x8a' 
    # btwoc ought to produce the shortest representation in two's
    # complement. Fortunately, pickle already does that.
    return res[3+ord(res[3]):3:-1]

def unbtwoc(b):
    return cPickle.loads('\x80\x02\x8a'+chr(len(b))+b[::-1]+'.')

# Appendix B; DH default prime
dh_prime = """
DCF93A0B883972EC0E19989AC5A2CE310E1D37717E8D9571BB7623731866E61E
F75A2E27898B057F9891C2E27A639C3F29B60814581CD3B2CA3986D268370557
7D45C2E7E52DC81C7A171876E5CEA74B1448BFDFAF18828EFD2519F14E45E382
6634AF1949E5B535CC829A483B8A76223E5D490A257F05BDFF16F2FB22C583AB
"""
dh_prime = long("".join(dh_prime.split()), 16)

def string_xor(s1, s2):
    res = []
    for c1, c2 in itertools.izip(s1, s2):
        res.append(chr(ord(c1) ^ ord(c2)))
    return ''.join(res)

def associate(services, url):
    '''Create an association (OpenID section 8) between RP and OP.
    Return response as a dictionary.'''
    data = {
        'openid.ns':"http://specs.openid.net/auth/2.0",
        'openid.mode':"associate",
        'openid.assoc_type':"HMAC-SHA1",
        'openid.session_type':"no-encryption",
        }
    if url.startswith('http:'):
        # Use DH exchange
        data['openid.session_type'] = "DH-SHA1"
        # Private key: random number between 1 and dh_prime-1
        priv = random.randrange(1, dh_prime-1)
        # Public key: 2^priv mod prime
        pubkey = pow(2L, priv, dh_prime)
        dh_public_base64 = base64.b64encode(btwoc(pubkey))
        # No need to send key and generator
        data['openid.dh_consumer_public'] = dh_public_base64
    if is_compat_1x(services):
        # 14.2.1: clear session_type in 1.1 compatibility mode
        if data['openid.session_type'] == "no-encryption":
            data['openid.session_type'] = ''
        del data['openid.ns']
    res = urllib.urlopen(url, urllib.urlencode(data))
    data = parse_response(res.read())
    if 'error' in data:
        raise ValueError, "associate failed: "+data['error']
    if url.startswith('http:'):
        enc_mac_key = base64.b64decode(data['enc_mac_key'])
        dh_server_public = unbtwoc(base64.b64decode(data['dh_server_public']))
        # shared secret: sha1(2^(server_priv*priv) mod prime) xor enc_mac_key
        shared_secret = btwoc(pow(dh_server_public, priv, dh_prime))
        shared_secret = hashlib.sha1(shared_secret).digest()
        if len(shared_secret) != len(enc_mac_key):
            raise ValueError, "incorrect DH key size"
        # Fake mac_key result
        data['mac_key'] = base64.b64encode(string_xor(enc_mac_key, shared_secret))
    return data

class _AX:
    def __init__(self):
        self.__dict__['_reverse'] = {}
    def __setattr__(self, name, value):
        self.__dict__[name] = value
        self._reverse[value] = name
    def lookup(self, value):
        try:
            return self._reverse[value]
        except KeyError:
            return 'a%d' % (hash(value) % 1000000000)

AX = _AX()
AX.nickname  = "http://axschema.org/namePerson/friendly"
AX.email     = "http://axschema.org/contact/email"
AX.fullname  = "http://axschema.org/namePerson"
AX.dob       = "http://axschema.org/birthDate"
AX.gender    = "http://axschema.org/person/gender"
AX.postcode  = "http://axschema.org/contact/postalCode/home"
AX.country   = "http://axschema.org/contact/country/home"
AX.language  = "http://axschema.org/pref/language"
AX.timezone  = "http://axschema.org/pref/timezone"
AX.first     = "http://axschema.org/namePerson/first"
AX.last      = "http://axschema.org/namePerson/last"

def request_authentication(services, url, assoc_handle, return_to,
                           claimed=None, op_local=None, realm=None,
                           sreg = (('nickname', 'email'), ()),
                           ax = ((AX.email, AX.first, AX.last), ())):
    '''Request authentication (OpenID section 9).
    services is the list of discovered service types,
    url the OP service URL, assoc_handle the established session
    dictionary, and return_to the return URL.

    The return_to URL will also be passed as realm, and the
    OP may perform RP discovery on it; always request these
    data through SREG 1.0 as well.

    If AX or SREG 1.1 are supported, request email address,
    first/last name, or nickname.

    Return the URL that the browser should be redirected to.'''

    if is_op_endpoint(services):
        # claimed is an OP identifier
        claimed = op_local = None

    if claimed is None:
        claimed = "http://specs.openid.net/auth/2.0/identifier_select"
    if op_local is None:
        op_local = "http://specs.openid.net/auth/2.0/identifier_select"
    if realm is None:
        realm = return_to
    data = {
        'openid.ns':"http://specs.openid.net/auth/2.0",
        'openid.mode':"checkid_setup",
        'openid.assoc_handle':assoc_handle,
        'openid.return_to':return_to,
        'openid.claimed_id':claimed,
        'openid.identity':op_local,
        'openid.realm':realm,
        'openid.sreg.required':'nickname,email',
        }
    sreg_req, sreg_opt = sreg
    sreg11 = {}
    if sreg_req or sreg_opt:
        data['openid.ns.sreg'] = "http://openid.net/sreg/1.0"
        if sreg_req:
            data['openid.sreg.required'] = sreg11['openid.sreg11.required'] = ','.join(sreg_req)
        if sreg_opt:
            data['openid.sreg.optional'] =  sreg11['openid.sreg11.optional'] =','.join(sreg_opt)
    if is_compat_1x(services):
        del data['openid.ns']
        del data['openid.claimed_id']
        del data['openid.realm']
        data['openid.trust_root'] = return_to
    ax_req, ax_opt = ax
    if "http://openid.net/srv/ax/1.0" in services and (ax_req or ax_opt):
        data.update({
            'openid.ns.ax':"http://openid.net/srv/ax/1.0",
            'openid.ax.mode':'fetch_request',
            })
        for uri in ax_req + ax_opt:
            data['openid.ax.type.'+AX.lookup(uri)] = uri
        if ax_req:
            data['openid.ax.required'] = ','.join(AX.lookup(uri) for uri in ax_req)
        if ax_opt:
            data['openid.ax.if_available'] = ','.join(AX.lookup(uri) for uri in ax_req)
    if "http://openid.net/extensions/sreg/1.1" in services and sreg11:
        data.update({
            'openid.ns.sreg11':"http://openid.net/extensions/sreg/1.1",
            })
        data.update(sreg11)
    if '?' in url:
        return url+'&'+urllib.urlencode(data)
    else:
        return url+"?"+urllib.urlencode(data)

class NotAuthenticated(Exception):
    pass

def _prepare_response(response):
    if isinstance(response, str):
        return cgi.parse_qs(response)
    # backwards compatibility: allow caller to pass parse_qs result
    # already
    pass
    return response

def authenticate(session, response):
    '''Process an authentication response.  session must be the
    established session (minimally including assoc_handle and
    mac_key), response the query string as given in the original URL
    (i.e. as the CGI variable QUERY_STRING).  If authentication
    succeeds, return the list of signed fields.  If the user was not
    authenticated, NotAuthenticated is raised.  If the HTTP request is
    invalid (missing parameters, failure to validate signature),
    different exceptions will be raised, typically ValueError.

    Callers must check openid.response_nonce for replay attacks.
    '''

    response = _prepare_response(response)

    # 1.1 compat: openid.ns may not be sent
    # if response['openid.ns'][0] != 'http://specs.openid.net/auth/2.0':
    #    raise ValueError('missing openid.ns')
    if session['assoc_handle'] != response['openid.assoc_handle'][0]:
        raise ValueError('incorrect session')
    if response['openid.mode'][0] == 'cancel':
        raise NotAuthenticated('provider did not authenticate user (cancelled)')
    if response['openid.mode'][0] != 'id_res':
        raise ValueError('invalid openid.mode')
    if  'openid.identity' not in response:
        raise ValueError('missing openid.identity')

    # Will not check nonce value - caller must verify this is not a replay

    signed = response['openid.signed'][0].split(',')
    query = []
    for name in signed:
        value = response['openid.'+name][0]
        value = '%s:%s\n' % (name, value)
        if sys.version_info >= (3,):
            value = value.encode('utf-8')
        query.append(value)
    query = b('').join(query)

    mac_key = base64.decodestring(b(session['mac_key']))
    transmitted_sig = base64.decodestring(b(response['openid.sig'][0]))
    computed_sig = hmac.new(mac_key, query, hashlib.sha1).digest()

    if transmitted_sig != computed_sig:
        raise ValueError('Invalid signature')

    # Check that all critical fields are signed. OpenID 2.0 says
    # that in a positive assertion, op_endpoint, return_to,
    # response_nonce and assoc_handle must be signed, and claimed_id
    # and identity if present in the response. 1.1 compatibility
    # says that response_nonce and op_endpoint may be missing.
    # In addition, OpenID 1.1 providers apparently fail to sign
    # assoc_handle often.
    if response['openid.mode'][0] == 'id_res':
        if 'return_to' not in signed or \
           ('openid.identity' in response and 'identity' not in signed) or \
           ('openid.claimed_id' in response and 'claimed_id' not in signed):
            raise ValueError, "Critical field missing in signature"

    return signed

def parse_nonce(nonce):
    '''Split a nonce into a (timestamp, ID) pair'''
    stamp = nonce.split('Z', 1)[0]
    stamp = time.strptime(stamp, "%Y-%m-%dT%H:%M:%S")[:6]
    stamp = datetime.datetime(*stamp)
    return stamp

def get_namespaces(resp):
    resp = _prepare_response(resp)
    res = {}
    for k, v in resp.items():
        if k.startswith('openid.ns.'):
            k = k.rsplit('.', 1)[1]
            res[v[0]] = k
    return res

def get_ax(resp, ns, validated):
    if "http://openid.net/srv/ax/1.0" not in ns:
        return {}
    ax = ns["http://openid.net/srv/ax/1.0"]+"."
    oax = "openid."+ax
    res = {}
    resp = _prepare_response(resp)
    for k, v in resp.items():
        if k.startswith(oax+"type."):
            k = k.rsplit('.',1)[1]
            value_name = oax+"value."+k
            if ax+"value."+k not in validated:
                continue
            res[v[0]] = resp[value_name][0]
    return res

def get_sreg(resp, validated):
    """Return the dictionary of simple registration parameters in resp,
    with the openid.sreg. prefix stripped."""
    res = {}
    resp = _prepare_response(resp)
    for k, v in resp.items():
        if k.startswith('openid.sreg.'):
            k = k[len('openid.sreg.'):]
            if 'sreg.'+k in validated:
                res[k] = v[0]
    return res

def get_email(resp):
    "Return the email address embedded response, or None."

    resp = _prepare_response(resp)
    validated = resp['openid.signed'][0]

    # SREG 1.0; doesn't require namespace, as the protocol doesn't
    # specify one
    if 'openid.sreg.email' in resp and \
       'sreg.email' in validated:
        return resp['openid.sreg.email'][0]

    ns = get_namespaces(resp)

    ax = get_ax(resp, ns, validated)
    if "http://axschema.org/contact/email" in ax:
        return ax["http://axschema.org/contact/email"]

    # TODO: SREG 1.1
    return None

def get_username(resp):
    """Return either nickname or (first, last) or None.
    This function is deprecated; use get_ax and get_sreg instead.
    """

    resp = _prepare_response(resp)
    validated = resp['openid.signed'][0]
    if 'openid.sreg.nickname' in resp and \
       'sreg.nickname' in validated:
        return resp['openid.sreg.nickname'][0]

    ns = get_namespaces(resp)

    ax = get_ax(resp, ns, validated)
    if "http://axschema.org/namePerson/first" in ax and \
       "http://axschema.org/namePerson/last" in ax:
        return (ax["http://axschema.org/namePerson/first"],
                ax["http://axschema.org/namePerson/last"])

    # TODO: SREG 1.1
    return