1<?php
2
3/**
4 * URI normalization routines.
5 *
6 * @package OpenID
7 * @author JanRain, Inc. <openid@janrain.com>
8 * @copyright 2005-2008 Janrain, Inc.
9 * @license http://www.apache.org/licenses/LICENSE-2.0 Apache
10 */
11
12require_once 'Auth/Yadis/Misc.php';
13
14// from appendix B of rfc 3986 (http://www.ietf.org/rfc/rfc3986.txt)
15function Auth_OpenID_getURIPattern()
16{
17    return '&^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\?([^#]*))?(#(.*))?&';
18}
19
20function Auth_OpenID_getAuthorityPattern()
21{
22    return '/^([^@]*@)?([^:]*)(:.*)?/';
23}
24
25function Auth_OpenID_getEncodedPattern()
26{
27    return '/%([0-9A-Fa-f]{2})/';
28}
29
30# gen-delims  = ":" / "/" / "?" / "#" / "[" / "]" / "@"
31#
32# sub-delims  = "!" / "$" / "&" / "'" / "(" / ")"
33#                  / "*" / "+" / "," / ";" / "="
34#
35# unreserved  = ALPHA / DIGIT / "-" / "." / "_" / "~"
36function Auth_OpenID_getURLIllegalCharRE()
37{
38    return "/([^-A-Za-z0-9:\/\?#\[\]@\!\$&'\(\)\*\+,;=\._~\%])/";
39}
40
41function Auth_OpenID_getUnreserved()
42{
43    $_unreserved = array();
44    for ($i = 0; $i < 256; $i++) {
45        $_unreserved[$i] = false;
46    }
47
48    for ($i = ord('A'); $i <= ord('Z'); $i++) {
49        $_unreserved[$i] = true;
50    }
51
52    for ($i = ord('0'); $i <= ord('9'); $i++) {
53        $_unreserved[$i] = true;
54    }
55
56    for ($i = ord('a'); $i <= ord('z'); $i++) {
57        $_unreserved[$i] = true;
58    }
59
60    $_unreserved[ord('-')] = true;
61    $_unreserved[ord('.')] = true;
62    $_unreserved[ord('_')] = true;
63    $_unreserved[ord('~')] = true;
64
65    return $_unreserved;
66}
67
68function Auth_OpenID_getEscapeRE()
69{
70    $parts = array();
71    foreach (array_merge(Auth_Yadis_getUCSChars(),
72                         Auth_Yadis_getIPrivateChars()) as $pair) {
73        list($m, $n) = $pair;
74        $parts[] = sprintf("%s-%s", chr($m), chr($n));
75    }
76
77    return sprintf('[%s]', implode('', $parts));
78}
79
80function Auth_OpenID_pct_encoded_replace_unreserved($mo)
81{
82    $_unreserved = Auth_OpenID_getUnreserved();
83
84    $i = intval($mo[1], 16);
85    if ($_unreserved[$i]) {
86        return chr($i);
87    } else {
88        return strtoupper($mo[0]);
89    }
90
91    return $mo[0];
92}
93
94function Auth_OpenID_pct_encoded_replace($mo)
95{
96    return chr(intval($mo[1], 16));
97}
98
99function Auth_OpenID_remove_dot_segments($path)
100{
101    $result_segments = array();
102
103    while ($path) {
104        if (Auth_Yadis_startswith($path, '../')) {
105            $path = substr($path, 3);
106        } else if (Auth_Yadis_startswith($path, './')) {
107            $path = substr($path, 2);
108        } else if (Auth_Yadis_startswith($path, '/./')) {
109            $path = substr($path, 2);
110        } else if ($path == '/.') {
111            $path = '/';
112        } else if (Auth_Yadis_startswith($path, '/../')) {
113            $path = substr($path, 3);
114            if ($result_segments) {
115                array_pop($result_segments);
116            }
117        } else if ($path == '/..') {
118            $path = '/';
119            if ($result_segments) {
120                array_pop($result_segments);
121            }
122        } else if (($path == '..') ||
123                   ($path == '.')) {
124            $path = '';
125        } else {
126            $i = 0;
127            if ($path[0] == '/') {
128                $i = 1;
129            }
130            $i = strpos($path, '/', $i);
131            if ($i === false) {
132                $i = strlen($path);
133            }
134            $result_segments[] = substr($path, 0, $i);
135            $path = substr($path, $i);
136        }
137    }
138
139    return implode('', $result_segments);
140}
141
142function Auth_OpenID_urinorm($uri)
143{
144    $uri_matches = array();
145    preg_match(Auth_OpenID_getURIPattern(), $uri, $uri_matches);
146
147    if (count($uri_matches) < 9) {
148        for ($i = count($uri_matches); $i <= 9; $i++) {
149            $uri_matches[] = '';
150        }
151    }
152
153    $illegal_matches = array();
154    preg_match(Auth_OpenID_getURLIllegalCharRE(),
155               $uri, $illegal_matches);
156    if ($illegal_matches) {
157        return null;
158    }
159
160    $scheme = $uri_matches[2];
161    if ($scheme) {
162        $scheme = strtolower($scheme);
163    }
164
165    $scheme = $uri_matches[2];
166    if ($scheme === '') {
167        // No scheme specified
168        return null;
169    }
170
171    $scheme = strtolower($scheme);
172    if (!in_array($scheme, array('http', 'https'))) {
173        // Not an absolute HTTP or HTTPS URI
174        return null;
175    }
176
177    $authority = $uri_matches[4];
178    if ($authority === '') {
179        // Not an absolute URI
180        return null;
181    }
182
183    $authority_matches = array();
184    preg_match(Auth_OpenID_getAuthorityPattern(),
185               $authority, $authority_matches);
186    if (count($authority_matches) === 0) {
187        // URI does not have a valid authority
188        return null;
189    }
190
191    if (count($authority_matches) < 4) {
192        for ($i = count($authority_matches); $i <= 4; $i++) {
193            $authority_matches[] = '';
194        }
195    }
196
197    list($_whole, $userinfo, $host, $port) = $authority_matches;
198
199    if ($userinfo === null) {
200        $userinfo = '';
201    }
202
203    if (strpos($host, '%') !== -1) {
204        $host = strtolower($host);
205        $host = preg_replace_callback(
206                  Auth_OpenID_getEncodedPattern(),
207                  'Auth_OpenID_pct_encoded_replace', $host);
208        // NO IDNA.
209        // $host = unicode($host, 'utf-8').encode('idna');
210    } else {
211        $host = strtolower($host);
212    }
213
214    if ($port) {
215        if (($port == ':') ||
216            ($scheme == 'http' && $port == ':80') ||
217            ($scheme == 'https' && $port == ':443')) {
218            $port = '';
219        }
220    } else {
221        $port = '';
222    }
223
224    $authority = $userinfo . $host . $port;
225
226    $path = $uri_matches[5];
227    $path = preg_replace_callback(
228               Auth_OpenID_getEncodedPattern(),
229               'Auth_OpenID_pct_encoded_replace_unreserved', $path);
230
231    $path = Auth_OpenID_remove_dot_segments($path);
232    if (!$path) {
233        $path = '/';
234    }
235
236    $query = $uri_matches[6];
237    if ($query === null) {
238        $query = '';
239    }
240
241    $fragment = $uri_matches[8];
242    if ($fragment === null) {
243        $fragment = '';
244    }
245
246    return $scheme . '://' . $authority . $path . $query . $fragment;
247}
248
249
250