application/LinkUtils.php

   1 <?php
   2
   3 use Shaarli\Bookmark\LinkDB;
   4
   5 /**
   6  * Get cURL callback function for CURLOPT_WRITEFUNCTION
   7  *
   8  * @param string $charset     to extract from the downloaded page (reference)
   9  * @param string $title       to extract from the downloaded page (reference)
  10  * @param string $curlGetInfo Optionnaly overrides curl_getinfo function
  11  *
  12  * @return Closure
  13  */
  14 function get_curl_download_callback(&$charset, &$title, $curlGetInfo = 'curl_getinfo')
  15 {
  16     $isRedirected = false;
  17     /**
  18      * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download).
  19      *
  20      * While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text'
  21      * Then we extract the title and the charset and stop the download when it's done.
  22      *
  23      * @param resource $ch   cURL resource
  24      * @param string   $data chunk of data being downloaded
  25      *
  26      * @return int|bool length of $data or false if we need to stop the download
  27      */
  28     return function (&$ch, $data) use ($curlGetInfo, &$charset, &$title, &$isRedirected) {
  29         $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE);
  30         if (!empty($responseCode) && in_array($responseCode, [301, 302])) {
  31             $isRedirected = true;
  32             return strlen($data);
  33         }
  34         if (!empty($responseCode) && $responseCode !== 200) {
  35             return false;
  36         }
  37         // After a redirection, the content type will keep the previous request value
  38         // until it finds the next content-type header.
  39         if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) {
  40             $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE);
  41         }
  42         if (!empty($contentType) && strpos($contentType, 'text/html') === false) {
  43             return false;
  44         }
  45         if (!empty($contentType) && empty($charset)) {
  46             $charset = header_extract_charset($contentType);
  47         }
  48         if (empty($charset)) {
  49             $charset = html_extract_charset($data);
  50         }
  51         if (empty($title)) {
  52             $title = html_extract_title($data);
  53         }
  54         // We got everything we want, stop the download.
  55         if (!empty($responseCode) && !empty($contentType) && !empty($charset) && !empty($title)) {
  56             return false;
  57         }
  58
  59         return strlen($data);
  60     };
  61 }
  62
  63 /**
  64  * Extract title from an HTML document.
  65  *
  66  * @param string $html HTML content where to look for a title.
  67  *
  68  * @return bool|string Extracted title if found, false otherwise.
  69  */
  70 function html_extract_title($html)
  71 {
  72     if (preg_match('!<title.*?>(.*?)</title>!is', $html, $matches)) {
  73         return trim(str_replace("\n", '', $matches[1]));
  74     }
  75     return false;
  76 }
  77
  78 /**
  79  * Extract charset from HTTP header if it's defined.
  80  *
  81  * @param string $header HTTP header Content-Type line.
  82  *
  83  * @return bool|string Charset string if found (lowercase), false otherwise.
  84  */
  85 function header_extract_charset($header)
  86 {
  87     preg_match('/charset="?([^; ]+)/i', $header, $match);
  88     if (! empty($match[1])) {
  89         return strtolower(trim($match[1]));
  90     }
  91
  92     return false;
  93 }
  94
  95 /**
  96  * Extract charset HTML content (tag <meta charset>).
  97  *
  98  * @param string $html HTML content where to look for charset.
  99  *
 100  * @return bool|string Charset string if found, false otherwise.
 101  */
 102 function html_extract_charset($html)
 103 {
 104     // Get encoding specified in HTML header.
 105     preg_match('#<meta .*charset=["\']?([^";\'>/]+)["\']? */?>#Usi', $html, $enc);
 106     if (!empty($enc[1])) {
 107         return strtolower($enc[1]);
 108     }
 109
 110     return false;
 111 }
 112
 113 /**
 114  * Count private links in given linklist.
 115  *
 116  * @param array|Countable $links Linklist.
 117  *
 118  * @return int Number of private links.
 119  */
 120 function count_private($links)
 121 {
 122     $cpt = 0;
 123     foreach ($links as $link) {
 124         if ($link['private']) {
 125             $cpt += 1;
 126         }
 127     }
 128
 129     return $cpt;
 130 }
 131
 132 /**
 133  * In a string, converts URLs to clickable links.
 134  *
 135  * @param string $text       input string.
 136  * @param string $redirector if a redirector is set, use it to gerenate links.
 137  * @param bool   $urlEncode  Use `urlencode()` on the URL after the redirector or not.
 138  *
 139  * @return string returns $text with all links converted to HTML links.
 140  *
 141  * @see Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722
 142  */
 143 function text2clickable($text, $redirector = '', $urlEncode = true)
 144 {
 145     $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[a-z0-9\(\)]/?)!si';
 146
 147     if (empty($redirector)) {
 148         return preg_replace($regex, '<a href="$1">$1</a>', $text);
 149     }
 150     // Redirector is set, urlencode the final URL.
 151     return preg_replace_callback(
 152         $regex,
 153         function ($matches) use ($redirector, $urlEncode) {
 154             $url = $urlEncode ? urlencode($matches[1]) : $matches[1];
 155             return '<a href="' . $redirector . $url .'">'. $matches[1] .'</a>';
 156         },
 157         $text
 158     );
 159 }
 160
 161 /**
 162  * Auto-link hashtags.
 163  *
 164  * @param string $description Given description.
 165  * @param string $indexUrl    Root URL.
 166  *
 167  * @return string Description with auto-linked hashtags.
 168  */
 169 function hashtag_autolink($description, $indexUrl = '')
 170 {
 171     /*
 172      * To support unicode: http://stackoverflow.com/a/35498078/1484919
 173      * \p{Pc} - to match underscore
 174      * \p{N} - numeric character in any script
 175      * \p{L} - letter from any language
 176      * \p{Mn} - any non marking space (accents, umlauts, etc)
 177      */
 178     $regex = '/(^|\s)#([\p{Pc}\p{N}\p{L}\p{Mn}]+)/mui';
 179     $replacement = '$1<a href="'. $indexUrl .'?addtag=$2" title="Hashtag $2">#$2</a>';
 180     return preg_replace($regex, $replacement, $description);
 181 }
 182
 183 /**
 184  * This function inserts &nbsp; where relevant so that multiple spaces are properly displayed in HTML
 185  * even in the absence of <pre>  (This is used in description to keep text formatting).
 186  *
 187  * @param string $text input text.
 188  *
 189  * @return string formatted text.
 190  */
 191 function space2nbsp($text)
 192 {
 193     return preg_replace('/(^| ) /m', '$1&nbsp;', $text);
 194 }
 195
 196 /**
 197  * Format Shaarli's description
 198  *
 199  * @param string $description shaare's description.
 200  * @param string $redirector  if a redirector is set, use it to gerenate links.
 201  * @param bool   $urlEncode  Use `urlencode()` on the URL after the redirector or not.
 202  * @param string $indexUrl    URL to Shaarli's index.
 203
 204  * @return string formatted description.
 205  */
 206 function format_description($description, $redirector = '', $urlEncode = true, $indexUrl = '')
 207 {
 208     return nl2br(space2nbsp(hashtag_autolink(text2clickable($description, $redirector, $urlEncode), $indexUrl)));
 209 }
 210
 211 /**
 212  * Generate a small hash for a link.
 213  *
 214  * @param DateTime $date Link creation date.
 215  * @param int      $id   Link ID.
 216  *
 217  * @return string the small hash generated from link data.
 218  */
 219 function link_small_hash($date, $id)
 220 {
 221     return smallHash($date->format(LinkDB::LINK_DATE_FORMAT) . $id);
 222 }