Remove personal data from Message ID seed. (#880029/847107)
[squirrelmail.git] / functions / url_parser.php
CommitLineData
59177427 1<?php
43fcef5c 2
35586184 3/**
4 * url_parser.php
5 *
35586184 6 * This code provides various string manipulation functions that are
598294a7 7 * used by the rest of the SquirrelMail code.
35586184 8 *
d4e46166 9 * @copyright &copy; 1999-2009 The SquirrelMail Project Team
4b4abf93 10 * @license http://opensource.org/licenses/gpl-license.php GNU Public License
31841a9e 11 * @version $Id$
d6c32258 12 * @package squirrelmail
35586184 13 */
43fcef5c 14
d6c32258 15/**
16 * Undocumented - complain, then patch.
17 */
35586184 18function replaceBlock (&$in, $replace, $start, $end) {
19 $begin = substr($in,0,$start);
20 $end = substr($in,$end,strlen($in)-$end);
21 $in = $begin.$replace.$end;
22}
43fcef5c 23
01d27858 24/* Having this defined in just one spot could help when changes need
25 * to be made to the pattern
26 * Make sure that the expression is evaluated case insensitively
7e235a1a 27 *
ff18cccd 28 * RFC2822 (and RFC822) defines the left side of an email address as (roughly):
29 * 1*atext *("." 1*atext)
30 * where atext is: a-zA-Z0-9!#$%&'*+-/=?^_`{|}~
31 *
01d27858 32 * Here's pretty sophisticated IP matching:
33 * $IPMatch = '(2[0-5][0-9]|1?[0-9]{1,2})';
34 * $IPMatch = '\[?' . $IPMatch . '(\.' . $IPMatch . '){3}\]?';
35 */
36/* Here's enough: */
37global $IP_RegExp_Match, $Host_RegExp_Match, $Email_RegExp_Match;
38$IP_RegExp_Match = '\\[?[0-9]{1,3}(\\.[0-9]{1,3}){3}\\]?';
7e235a1a 39$Host_RegExp_Match = '(' . $IP_RegExp_Match .
01d27858 40 '|[0-9a-z]([-.]?[0-9a-z])*\\.[a-z][a-z]+)';
ff18cccd 41$atext = '([a-z0-9!#$&%*+/=?^_`{|}~-]|&amp;)';
42$dot_atom = $atext . '+(\.' . $atext . '+)*';
43$Email_RegExp_Match = $dot_atom . '(%' . $Host_RegExp_Match . ')?@' .
44 $Host_RegExp_Match;
7e235a1a 45
8b096f0a 46/**
47 * Parses a body and converts all found email addresses to clickable links.
48 *
49 * @param string body the body to process, by ref
50 * @return int the number of unique addresses found
51 */
01d27858 52function parseEmail (&$body) {
916669ad 53 global $Email_RegExp_Match;
cf0d436a 54 $sbody = $body;
55 $addresses = array();
56
57 /* Find all the email addresses in the body */
b7910e12 58 while (preg_match('/' . $Email_RegExp_Match . '/i', $sbody, $regs)) {
ff18cccd 59 $addresses[$regs[0]] = strtr($regs[0], array('&amp;' => '&'));
cf0d436a 60 $start = strpos($sbody, $regs[0]) + strlen($regs[0]);
61 $sbody = substr($sbody, $start);
62 }
916669ad 63
cf0d436a 64 /* Replace each email address with a compose URL */
ff18cccd 65 foreach ($addresses as $text => $email) {
66 $comp_uri = makeComposeLink('src/compose.php?send_to='.urlencode($email), $text);
67 $body = str_replace($text, $comp_uri, $body);
7e235a1a 68 }
916669ad 69
cf0d436a 70 /* Return number of unique addresses found */
71 return count($addresses);
01d27858 72}
43fcef5c 73
43fcef5c 74
01d27858 75/* We don't want to re-initialize this stuff for every line. Save work
76 * and just do it once here.
77 */
78global $url_parser_url_tokens;
79$url_parser_url_tokens = array(
80 'http://',
81 'https://',
82 'ftp://',
83 'telnet:', // Special case -- doesn't need the slashes
7efaee4f 84 'mailto:', // Special case -- doesn't use the slashes
01d27858 85 'gopher://',
86 'news://');
8f7163e7 87
01d27858 88global $url_parser_poss_ends;
62f7daa5 89$url_parser_poss_ends = array(' ', "\n", "\r", '<', '>', ".\r", ".\n",
90 '.&nbsp;', '&nbsp;', ')', '(', '&quot;', '&lt;', '&gt;', '.<',
01d27858 91 ']', '[', '{', '}', "\240", ', ', '. ', ",\n", ",\r");
8f7163e7 92
20a60f89 93
8b096f0a 94/**
7efaee4f 95 * rfc 2368 (mailto URL) preg_match() regexp
99554426 96 * @link http://www.ietf.org/rfc/rfc2368.txt
7efaee4f 97 * @global string MailTo_PReg_Match the encapsulated regexp for preg_match()
98 */
99global $MailTo_PReg_Match;
100$Mailto_Email_RegExp = '[0-9a-z%]([-_.+%]?[0-9a-z])*(%' . $Host_RegExp_Match . ')?@' . $Host_RegExp_Match;
101$MailTo_PReg_Match = '/((?:' . $Mailto_Email_RegExp . ')*)((?:\?(?:to|cc|bcc|subject|body)=[^\s\?&=,()]+)?(?:&amp;(?:to|cc|bcc|subject|body)=[^\s\?&=,()]+)*)/i';
102
103/**
8b096f0a 104 * Parses a body and converts all found URLs to clickable links.
105 *
106 * @param string body the body to process, by ref
107 * @return void
108 */
01d27858 109function parseUrl (&$body) {
99554426 110 global $url_parser_poss_ends, $url_parser_url_tokens;
cf0d436a 111 $start = 0;
112 $blength = strlen($body);
cf0d436a 113
7efaee4f 114 while ($start < $blength) {
8f7163e7 115 $target_token = '';
0d3ff000 116 $target_pos = $blength;
cf0d436a 117
01d27858 118 /* Find the first token to replace */
119 foreach ($url_parser_url_tokens as $the_token) {
120 $pos = strpos(strtolower($body), $the_token, $start);
6f1450f4 121 if (is_int($pos) && $pos < $target_pos) {
cf0d436a 122 $target_pos = $pos;
01d27858 123 $target_token = $the_token;
124 }
8f7163e7 125 }
cf0d436a 126
01d27858 127 /* Look for email addresses between $start and $target_pos */
cf0d436a 128 $check_str = substr($body, $start, $target_pos-$start);
129
01d27858 130 if (parseEmail($check_str)) {
131 replaceBlock($body, $check_str, $start, $target_pos);
cf0d436a 132 $blength = strlen($body);
01d27858 133 $target_pos = strlen($check_str) + $start;
8f7163e7 134 }
e2ef6f4b 135
879c8945 136 // rfc 2368 (mailto URL)
137 if ($target_token == 'mailto:') {
91e0dccc 138 $target_pos += 7; //skip mailto:
7efaee4f 139 $end = $blength;
140
141 $mailto = substr($body, $target_pos, $end-$target_pos);
142
143 global $MailTo_PReg_Match;
99554426 144 if ((preg_match($MailTo_PReg_Match, $mailto, $regs)) && ($regs[0] != '')) {
7efaee4f 145 //sm_print_r($regs);
146 $mailto_before = $target_token . $regs[0];
147 $mailto_params = $regs[10];
91e0dccc 148 if ($regs[1]) { //if there is an email addr before '?', we need to merge it with the params
7efaee4f 149 $to = 'to=' . $regs[1];
91e0dccc 150 if (strpos($mailto_params, 'to=') > -1) //already a 'to='
7efaee4f 151 $mailto_params = str_replace('to=', $to . '%2C%20', $mailto_params);
152 else {
91e0dccc 153 if ($mailto_params) //already some params, append to them
99554426 154 $mailto_params .= '&amp;' . $to;
7efaee4f 155 else
99554426 156 $mailto_params .= '?' . $to;
7efaee4f 157 }
158 }
1bc52de5 159 $url_str = preg_replace(array('/to=/i', '/(?<!b)cc=/i', '/bcc=/i'), array('send_to=', 'send_to_cc=', 'send_to_bcc='), $mailto_params);
7efaee4f 160 $comp_uri = makeComposeLink('src/compose.php' . $url_str, $mailto_before);
99554426 161 replaceBlock($body, $comp_uri, $target_pos - 7, $target_pos + strlen($regs[0]));
162 $target_pos += strlen($comp_uri) - 7;
7efaee4f 163 }
164 }
165 else
879c8945 166 /* If there was a token to replace, replace it */
01d27858 167 if ($target_token != '') {
168 /* Find the end of the URL */
cf0d436a 169 $end = $blength;
170 foreach ($url_parser_poss_ends as $val) {
171 $enda = strpos($body, $val, $target_pos);
01d27858 172 if (is_int($enda) && $enda < $end) {
173 $end = $enda;
174 }
175 }
cf0d436a 176
1f18e313 177 /* make sure that there are no 8bit chars between $target_pos and suspected end of URL */
178 if (!is_bool($first8bit=sq_strpos_8bit($body,$target_pos,$end))) {
179 $end = $first8bit;
180 }
181
01d27858 182 /* Extract URL */
183 $url = substr($body, $target_pos, $end-$target_pos);
cf0d436a 184
01d27858 185 /* Needed since lines are not passed with \n or \r */
b7910e12 186 while ( preg_match('/[,.]$/', $url) ) {
01d27858 187 $url = substr( $url, 0, -1 );
188 $end--;
189 }
1b3324b3 190
01d27858 191 /* Replace URL with HyperLinked Url, requires 1 char in link */
192 if ($url != '' && $url != $target_token) {
193 $url_str = "<a href=\"$url\" target=\"_blank\">$url</a>";
194 replaceBlock($body,$url_str,$target_pos,$end);
195 $target_pos += strlen($url_str);
cf0d436a 196 }
197 else {
198 // Not quite a valid link, skip ahead to next chance
199 $target_pos += strlen($target_token);
01d27858 200 }
8f7163e7 201 }
cf0d436a 202
01d27858 203 /* Move forward */
cf0d436a 204 $start = $target_pos;
205 $blength = strlen($body);
01d27858 206 }
62f7daa5 207}
916669ad 208
209/**
210 * Parses a string and returns the first e-mail address found.
211 *
212 * @param string string the string to process
213 * @return string the first e-mail address found
214 */
215function getEmail($string) {
216 global $Email_RegExp_Match;
217 $addresses = array();
218
219 /* Find all the email addresses in the body */
b7910e12 220 while (preg_match('/' . $Email_RegExp_Match . '/i', $string, $regs)) {
916669ad 221 $addresses[$regs[0]] = strtr($regs[0], array('&amp;' => '&'));
91e0dccc 222 $start = strpos($string, $regs[0]) + strlen($regs[0]);
223 $string = substr($string, $start);
916669ad 224 }
225
226 /* Return the first address, or an empty string if no address was found */
227 $addresses = array_values($addresses);
228 return (array_key_exists(0, $addresses) ? $addresses[0] : '');
229}
230
1f18e313 231/**
232 * Finds first occurrence of 8bit data in the string
233 *
234 * Function finds first 8bit symbol or html entity that represents 8bit character.
235 * Search start is defined by $offset argument. Search ends at $maxlength position.
236 * If $maxlength is not defined or bigger than provided string, search ends when
237 * string ends.
238 *
239 * Check returned data type in order to avoid confusion between bool(false)
240 * (not found) and int(0) (first char in the string).
241 * @param string $haystack
242 * @param integer $offset
243 * @param integer $maxlength
244 * @return mixed integer with first 8bit character position or boolean false
245 * @since 1.5.2
246 */
247function sq_strpos_8bit($haystack,$offset=0,$maxlength=false) {
248 $ret = false;
249
250 if ($maxlength===false || strlen($haystack) < $maxlength) {
251 $maxlength=strlen($haystack);
252 }
253
254 for($i=$offset;$i<$maxlength;$i++) {
255 /* rh7-8 compatibility. don't use full 8bit range in regexp */
256 if (preg_match('/[\200-\237]|\240|[\241-\377]/',$haystack[$i])) {
257 /* we have 8bit char. stop here and return position */
258 $ret = $i;
259 break;
260 } elseif ($haystack[$i]=='&') {
261 $substring = substr($haystack,$i);
262 /**
263 * 1. look for "&#(decimal number);" where decimal_number is bigger than 127
264 * 2. look for "&x(hexadecimal number);", where hex number is bigger than x7f
265 * 3. look for any html character entity that is not 7bit html special char. Use
266 * own sq_get_html_translation_table() function with 'utf-8' character set in
267 * order to get all html entities.
268 */
269 if ((preg_match('/^&#(\d+);/',$substring,$match) && $match[1]>127) ||
270 (preg_match('/^&x([0-9a-f]+);/i',$substring,$match) && $match[1]>"\x7f") ||
271 (preg_match('/^&([a-z]+);/i',$substring,$match) &&
272 !in_array($match[0],get_html_translation_table(HTML_SPECIALCHARS)) &&
273 in_array($match[0],sq_get_html_translation_table(HTML_ENTITIES,ENT_COMPAT,'utf-8')))) {
274 $ret = $i;
275 break;
276 }
277 }
278 }
279 return $ret;
280}