application/LinkUtils.php

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