X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=application%2FLinkUtils.php;h=d56e019f4246f40d94f7a2eb5b8f715d10c7f1a1;hb=1004742f09b55ff781c13745781b9a7e90986faa;hp=2df76ba8a4b457f4867efb9e960b7242d55a614d;hpb=ce7b0b6480aa854ee6893f5c889277b0e3b13efc;p=github%2Fshaarli%2FShaarli.git

diff --git a/application/LinkUtils.php b/application/LinkUtils.php
index 2df76ba8..d56e019f 100644
--- a/application/LinkUtils.php
+++ b/application/LinkUtils.php
@@ -1,5 +1,63 @@
 <?php
 
+/**
+ * Get cURL callback function for CURLOPT_WRITEFUNCTION
+ *
+ * @param string $charset     to extract from the downloaded page (reference)
+ * @param string $title       to extract from the downloaded page (reference)
+ * @param string $curlGetInfo Optionnaly overrides curl_getinfo function
+ *
+ * @return Closure
+ */
+function get_curl_download_callback(&$charset, &$title, $curlGetInfo = 'curl_getinfo')
+{
+    $isRedirected = false;
+    /**
+     * cURL callback function for CURLOPT_WRITEFUNCTION (called during the download).
+     *
+     * While downloading the remote page, we check that the HTTP code is 200 and content type is 'html/text'
+     * Then we extract the title and the charset and stop the download when it's done.
+     *
+     * @param resource $ch   cURL resource
+     * @param string   $data chunk of data being downloaded
+     *
+     * @return int|bool length of $data or false if we need to stop the download
+     */
+    return function (&$ch, $data) use ($curlGetInfo, &$charset, &$title, &$isRedirected) {
+        $responseCode = $curlGetInfo($ch, CURLINFO_RESPONSE_CODE);
+        if (!empty($responseCode) && in_array($responseCode, [301, 302])) {
+            $isRedirected = true;
+            return strlen($data);
+        }
+        if (!empty($responseCode) && $responseCode !== 200) {
+            return false;
+        }
+        // After a redirection, the content type will keep the previous request value
+        // until it finds the next content-type header.
+        if (! $isRedirected || strpos(strtolower($data), 'content-type') !== false) {
+            $contentType = $curlGetInfo($ch, CURLINFO_CONTENT_TYPE);
+        }
+        if (!empty($contentType) && strpos($contentType, 'text/html') === false) {
+            return false;
+        }
+        if (!empty($contentType) && empty($charset)) {
+            $charset = header_extract_charset($contentType);
+        }
+        if (empty($charset)) {
+            $charset = html_extract_charset($data);
+        }
+        if (empty($title)) {
+            $title = html_extract_title($data);
+        }
+        // We got everything we want, stop the download.
+        if (!empty($responseCode) && !empty($contentType) && !empty($charset) && !empty($title)) {
+            return false;
+        }
+
+        return strlen($data);
+    };
+}
+
 /**
  * Extract title from an HTML document.
  *
@@ -16,64 +74,147 @@ function html_extract_title($html)
 }
 
 /**
- * Determine charset from downloaded page.
- * Priority:
- *   1. HTTP headers (Content type).
- *   2. HTML content page (tag <meta charset>).
- *   3. Use a default charset (default: UTF-8).
+ * Extract charset from HTTP header if it's defined.
  *
- * @param array  $headers           HTTP headers array.
- * @param string $htmlContent       HTML content where to look for charset.
- * @param string $defaultCharset    Default charset to apply if other methods failed.
+ * @param string $header HTTP header Content-Type line.
  *
- * @return string Determined charset.
+ * @return bool|string Charset string if found (lowercase), false otherwise.
  */
-function get_charset($headers, $htmlContent, $defaultCharset = 'utf-8')
+function header_extract_charset($header)
 {
-    if ($charset = headers_extract_charset($headers)) {
-        return $charset;
+    preg_match('/charset="?([^; ]+)/i', $header, $match);
+    if (! empty($match[1])) {
+        return strtolower(trim($match[1]));
     }
 
-    if ($charset = html_extract_charset($htmlContent)) {
-        return $charset;
+    return false;
+}
+
+/**
+ * Extract charset HTML content (tag <meta charset>).
+ *
+ * @param string $html HTML content where to look for charset.
+ *
+ * @return bool|string Charset string if found, false otherwise.
+ */
+function html_extract_charset($html)
+{
+    // Get encoding specified in HTML header.
+    preg_match('#<meta .*charset=["\']?([^";\'>/]+)["\']? */?>#Usi', $html, $enc);
+    if (!empty($enc[1])) {
+        return strtolower($enc[1]);
     }
 
-    return $defaultCharset;
+    return false;
 }
 
 /**
- * Extract charset from HTTP headers if it's defined.
+ * Count private links in given linklist.
  *
- * @param array $headers HTTP headers array.
+ * @param array|Countable $links Linklist.
  *
- * @return bool|string Charset string if found (lowercase), false otherwise.
+ * @return int Number of private links.
  */
-function headers_extract_charset($headers)
+function count_private($links)
 {
-    if (! empty($headers['Content-Type']) && strpos($headers['Content-Type'], 'charset=') !== false) {
-        preg_match('/charset="?([^; ]+)/i', $headers['Content-Type'], $match);
-        if (! empty($match[1])) {
-            return strtolower(trim($match[1]));
+    $cpt = 0;
+    foreach ($links as $link) {
+        if ($link['private']) {
+            $cpt += 1;
         }
     }
 
-    return false;
+    return $cpt;
 }
 
 /**
- * Extract charset HTML content (tag <meta charset>).
+ * In a string, converts URLs to clickable links.
  *
- * @param string $html HTML content where to look for charset.
+ * @param string $text       input string.
+ * @param string $redirector if a redirector is set, use it to gerenate links.
+ * @param bool   $urlEncode  Use `urlencode()` on the URL after the redirector or not.
  *
- * @return bool|string Charset string if found, false otherwise.
+ * @return string returns $text with all links converted to HTML links.
+ *
+ * @see Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722
  */
-function html_extract_charset($html)
+function text2clickable($text, $redirector = '', $urlEncode = true)
 {
-    // Get encoding specified in HTML header.
-    preg_match('#<meta .*charset=["\']?([^";\'>/]+)["\']? */?>#Usi', $html, $enc);
-    if (!empty($enc[1])) {
-        return strtolower($enc[1]);
+    $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[a-z0-9\(\)]/?)!si';
+
+    if (empty($redirector)) {
+        return preg_replace($regex, '<a href="$1">$1</a>', $text);
     }
+    // Redirector is set, urlencode the final URL.
+    return preg_replace_callback(
+        $regex,
+        function ($matches) use ($redirector, $urlEncode) {
+            $url = $urlEncode ? urlencode($matches[1]) : $matches[1];
+            return '<a href="' . $redirector . $url .'">'. $matches[1] .'</a>';
+        },
+        $text
+    );
+}
 
-    return false;
+/**
+ * Auto-link hashtags.
+ *
+ * @param string $description Given description.
+ * @param string $indexUrl    Root URL.
+ *
+ * @return string Description with auto-linked hashtags.
+ */
+function hashtag_autolink($description, $indexUrl = '')
+{
+    /*
+     * To support unicode: http://stackoverflow.com/a/35498078/1484919
+     * \p{Pc} - to match underscore
+     * \p{N} - numeric character in any script
+     * \p{L} - letter from any language
+     * \p{Mn} - any non marking space (accents, umlauts, etc)
+     */
+    $regex = '/(^|\s)#([\p{Pc}\p{N}\p{L}\p{Mn}]+)/mui';
+    $replacement = '$1<a href="'. $indexUrl .'?addtag=$2" title="Hashtag $2">#$2</a>';
+    return preg_replace($regex, $replacement, $description);
+}
+
+/**
+ * This function inserts &nbsp; where relevant so that multiple spaces are properly displayed in HTML
+ * even in the absence of <pre>  (This is used in description to keep text formatting).
+ *
+ * @param string $text input text.
+ *
+ * @return string formatted text.
+ */
+function space2nbsp($text)
+{
+    return preg_replace('/(^| ) /m', '$1&nbsp;', $text);
+}
+
+/**
+ * Format Shaarli's description
+ *
+ * @param string $description shaare's description.
+ * @param string $redirector  if a redirector is set, use it to gerenate links.
+ * @param bool   $urlEncode  Use `urlencode()` on the URL after the redirector or not.
+ * @param string $indexUrl    URL to Shaarli's index.
+
+ * @return string formatted description.
+ */
+function format_description($description, $redirector = '', $urlEncode = true, $indexUrl = '')
+{
+    return nl2br(space2nbsp(hashtag_autolink(text2clickable($description, $redirector, $urlEncode), $indexUrl)));
+}
+
+/**
+ * Generate a small hash for a link.
+ *
+ * @param DateTime $date Link creation date.
+ * @param int      $id   Link ID.
+ *
+ * @return string the small hash generated from link data.
+ */
+function link_small_hash($date, $id)
+{
+    return smallHash($date->format(LinkDB::LINK_DATE_FORMAT) . $id);
 }