]>
git.immae.eu Git - github/shaarli/Shaarli.git/blob - application/bookmark/LinkUtils.php
3 use Shaarli\Bookmark\LinkDB
;
6 * Get cURL callback function for CURLOPT_WRITEFUNCTION
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 $curlGetInfo Optionally overrides curl_getinfo function
14 function get_curl_download_callback(&$charset, &$title, $curlGetInfo = 'curl_getinfo')
16 $isRedirected = false;
18 * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download).
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.
23 * @param resource $ch cURL resource
24 * @param string $data chunk of data being downloaded
26 * @return int|bool length of $data or false if we need to stop the download
28 return function (&$ch, $data) use ($curlGetInfo, &$charset, &$title, &$isRedirected) {
29 $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE
);
30 if (!empty($responseCode) && in_array($responseCode, [301, 302])) {
34 if (!empty($responseCode) && $responseCode !== 200) {
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
);
42 if (!empty($contentType) && strpos($contentType, 'text/html') === false) {
45 if (!empty($contentType) && empty($charset)) {
46 $charset = header_extract_charset($contentType);
48 if (empty($charset)) {
49 $charset = html_extract_charset($data);
52 $title = html_extract_title($data);
54 // We got everything we want, stop the download.
55 if (!empty($responseCode) && !empty($contentType) && !empty($charset) && !empty($title)) {
64 * Extract title from an HTML document.
66 * @param string $html HTML content where to look for a title.
68 * @return bool|string Extracted title if found, false otherwise.
70 function html_extract_title($html)
72 if (preg_match('!<title.*?>(.*?)</title>!is', $html, $matches)) {
73 return trim(str_replace("\n", '', $matches[1]));
79 * Extract charset from HTTP header if it's defined.
81 * @param string $header HTTP header Content-Type line.
83 * @return bool|string Charset string if found (lowercase), false otherwise.
85 function header_extract_charset($header)
87 preg_match('/charset="?([^; ]+)/i', $header, $match);
88 if (! empty($match[1])) {
89 return strtolower(trim($match[1]));
96 * Extract charset HTML content (tag <meta charset>).
98 * @param string $html HTML content where to look for charset.
100 * @return bool|string Charset string if found, false otherwise.
102 function html_extract_charset($html)
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]);
114 * Count private links in given linklist.
116 * @param array|Countable $links Linklist.
118 * @return int Number of private links.
120 function count_private($links)
123 foreach ($links as $link) {
124 if ($link['private']) {
133 * In a string, converts URLs to clickable links.
135 * @param string $text input string.
137 * @return string returns $text with all links converted to HTML links.
139 * @see Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722
141 function text2clickable($text)
143 $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[a-z0-9\(\)]/?)!si';
144 return preg_replace($regex, '<a href="$1">$1</a>', $text);
148 * Auto-link hashtags.
150 * @param string $description Given description.
151 * @param string $indexUrl Root URL.
153 * @return string Description with auto-linked hashtags.
155 function hashtag_autolink($description, $indexUrl = '')
158 * To support unicode: http://stackoverflow.com/a/35498078/1484919
159 * \p{Pc} - to match underscore
160 * \p{N} - numeric character in any script
161 * \p{L} - letter from any language
162 * \p{Mn} - any non marking space (accents, umlauts, etc)
164 $regex = '/(^|\s)#([\p{Pc}\p{N}\p{L}\p{Mn}]+)/mui';
165 $replacement = '$1<a href="'. $indexUrl .'?addtag=$2" title="Hashtag $2">#$2</a>';
166 return preg_replace($regex, $replacement, $description);
170 * This function inserts where relevant so that multiple spaces are properly displayed in HTML
171 * even in the absence of <pre> (This is used in description to keep text formatting).
173 * @param string $text input text.
175 * @return string formatted text.
177 function space2nbsp($text)
179 return preg_replace('/(^| ) /m', '$1 ', $text);
183 * Format Shaarli's description
185 * @param string $description shaare's description.
186 * @param string $indexUrl URL to Shaarli's index.
188 * @return string formatted description.
190 function format_description($description, $indexUrl = '')
192 return nl2br(space2nbsp(hashtag_autolink(text2clickable($description), $indexUrl)));
196 * Generate a small hash for a link.
198 * @param DateTime $date Link creation date.
199 * @param int $id Link ID.
201 * @return string the small hash generated from link data.
203 function link_small_hash($date, $id)
205 return smallHash($date->format(LinkDB
::LINK_DATE_FORMAT
) . $id);
209 * Returns whether or not the link is an internal note.
210 * Its URL starts by `?` because it's actually a permalink.
212 * @param string $linkUrl
214 * @return bool true if internal note, false otherwise.
216 function is_note($linkUrl)
218 return isset($linkUrl[0]) && $linkUrl[0] === '?';