X-Git-Url: https://git.immae.eu/?a=blobdiff_plain;f=application%2Fbookmark%2FLinkUtils.php;h=8fa2953a78d3e8f18bbcba3824801d859169c16e;hb=9ef8555ad298668bcb8537ccdd2ab6560f44177f;hp=35a5b290454b32870464c6457fea73059d613f4c;hpb=cc69aad4a903dc603f47cfa87aeb1865b03c8950;p=github%2Fshaarli%2FShaarli.git

diff --git a/application/bookmark/LinkUtils.php b/application/bookmark/LinkUtils.php
index 35a5b290..8fa2953a 100644
--- a/application/bookmark/LinkUtils.php
+++ b/application/bookmark/LinkUtils.php
@@ -1,64 +1,7 @@
 <?php
 
-use Shaarli\Bookmark\LinkDB;
-
-/**
- * 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 $curlGetInfo Optionally overrides curl_getinfo function
- *
- * @return Closure
- */
-function get_curl_download_callback(&$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);
-    };
-}
+use Shaarli\Bookmark\Bookmark;
+use Shaarli\Formatter\BookmarkDefaultFormatter;
 
 /**
  * Extract title from an HTML document.
@@ -84,7 +27,7 @@ function html_extract_title($html)
  */
 function header_extract_charset($header)
 {
-    preg_match('/charset="?([^; ]+)/i', $header, $match);
+    preg_match('/charset=["\']?([^; "\']+)/i', $header, $match);
     if (! empty($match[1])) {
         return strtolower(trim($match[1]));
     }
@@ -111,37 +54,63 @@ function html_extract_charset($html)
 }
 
 /**
- * Count private links in given linklist.
+ * Extract meta tag from HTML content in either:
+ *   - OpenGraph: <meta property="og:[tag]" ...>
+ *   - Meta tag: <meta name="[tag]" ...>
  *
- * @param array|Countable $links Linklist.
+ * @param string $tag  Name of the tag to retrieve.
+ * @param string $html HTML content where to look for charset.
  *
- * @return int Number of private links.
+ * @return bool|string Charset string if found, false otherwise.
  */
-function count_private($links)
+function html_extract_tag($tag, $html)
 {
-    $cpt = 0;
-    foreach ($links as $link) {
-        if ($link['private']) {
-            $cpt += 1;
-        }
+    $propertiesKey = ['property', 'name', 'itemprop'];
+    $properties = implode('|', $propertiesKey);
+    // We need a OR here to accept either 'property=og:noquote' or 'property="og:unrelated og:my-tag"'
+    $orCondition  = '["\']?(?:og:)?' . $tag . '["\']?|["\'][^\'"]*?(?:og:)?' . $tag . '[^\'"]*?[\'"]';
+    // Support quotes in double quoted content, and the other way around
+    $content = 'content=(["\'])((?:(?!\1).)*)\1';
+    // Try to retrieve OpenGraph tag.
+    $ogRegex = '#<meta[^>]+(?:' . $properties . ')=(?:' . $orCondition . ')[^>]*' . $content . '.*?>#';
+    // If the attributes are not in the order property => content (e.g. Github)
+    // New regex to keep this readable... more or less.
+    $ogRegexReverse = '#<meta[^>]+' . $content . '[^>]+(?:' . $properties . ')=(?:' . $orCondition . ').*?>#';
+
+    if (
+        preg_match($ogRegex, $html, $matches) > 0
+        || preg_match($ogRegexReverse, $html, $matches) > 0
+    ) {
+        return $matches[2];
     }
 
-    return $cpt;
+    return false;
 }
 
 /**
- * In a string, converts URLs to clickable links.
+ * In a string, converts URLs to clickable bookmarks.
  *
  * @param string $text       input string.
  *
- * @return string returns $text with all links converted to HTML links.
+ * @return string returns $text with all bookmarks converted to HTML bookmarks.
  *
  * @see Function inspired from http://www.php.net/manual/en/function.preg-replace.php#85722
  */
 function text2clickable($text)
 {
     $regex = '!(((?:https?|ftp|file)://|apt:|magnet:)\S+[a-z0-9\(\)]/?)!si';
-    return preg_replace($regex, '<a href="$1">$1</a>', $text);
+    $format = function (array $match): string {
+        return '<a href="' .
+            str_replace(
+                BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_OPEN,
+                '',
+                str_replace(BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_CLOSE, '', $match[1])
+            ) .
+            '">' . $match[1] . '</a>'
+        ;
+    };
+
+    return preg_replace_callback($regex, $format, $text);
 }
 
 /**
@@ -154,6 +123,9 @@ function text2clickable($text)
  */
 function hashtag_autolink($description, $indexUrl = '')
 {
+    $tokens = '(?:' . BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_OPEN . ')' .
+              '(?:' . BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_CLOSE . ')'
+    ;
     /*
      * To support unicode: http://stackoverflow.com/a/35498078/1484919
      * \p{Pc} - to match underscore
@@ -161,9 +133,20 @@ function hashtag_autolink($description, $indexUrl = '')
      * \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);
+    $regex = '/(^|\s)#([\p{Pc}\p{N}\p{L}\p{Mn}' . $tokens . ']+)/mui';
+    $format = function (array $match) use ($indexUrl): string {
+        $cleanMatch = str_replace(
+            BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_OPEN,
+            '',
+            str_replace(BookmarkDefaultFormatter::SEARCH_HIGHLIGHT_CLOSE, '', $match[2])
+        );
+        return $match[1] . '<a href="' . $indexUrl . './add-tag/' . $cleanMatch . '"' .
+            ' title="Hashtag ' . $cleanMatch . '">' .
+                '#' . $match[2] .
+        '</a>';
+    };
+
+    return preg_replace_callback($regex, $format, $description);
 }
 
 /**
@@ -184,12 +167,17 @@ function space2nbsp($text)
  *
  * @param string $description shaare's description.
  * @param string $indexUrl    URL to Shaarli's index.
-
+ * @param bool   $autolink    Turn on/off automatic linkifications of URLs and hashtags
+ *
  * @return string formatted description.
  */
-function format_description($description, $indexUrl = '')
+function format_description($description, $indexUrl = '', $autolink = true)
 {
-    return nl2br(space2nbsp(hashtag_autolink(text2clickable($description), $indexUrl)));
+    if ($autolink) {
+        $description = hashtag_autolink(text2clickable($description), $indexUrl);
+    }
+
+    return nl2br(space2nbsp($description));
 }
 
 /**
@@ -202,7 +190,7 @@ function format_description($description, $indexUrl = '')
  */
 function link_small_hash($date, $id)
 {
-    return smallHash($date->format(LinkDB::LINK_DATE_FORMAT) . $id);
+    return smallHash($date->format(Bookmark::LINK_DATE_FORMAT) . $id);
 }
 
 /**
@@ -217,3 +205,49 @@ function is_note($linkUrl)
 {
     return isset($linkUrl[0]) && $linkUrl[0] === '?';
 }
+
+/**
+ * Extract an array of tags from a given tag string, with provided separator.
+ *
+ * @param string|null $tags      String containing a list of tags separated by $separator.
+ * @param string      $separator Shaarli's default: ' ' (whitespace)
+ *
+ * @return array List of tags
+ */
+function tags_str2array(?string $tags, string $separator): array
+{
+    // For whitespaces, we use the special \s regex character
+    $separator = $separator === ' ' ? '\s' : $separator;
+
+    return preg_split('/\s*' . $separator . '+\s*/', trim($tags) ?? '', -1, PREG_SPLIT_NO_EMPTY);
+}
+
+/**
+ * Return a tag string with provided separator from a list of tags.
+ * Note that given array is clean up by tags_filter().
+ *
+ * @param array|null $tags      List of tags
+ * @param string     $separator
+ *
+ * @return string
+ */
+function tags_array2str(?array $tags, string $separator): string
+{
+    return implode($separator, tags_filter($tags, $separator));
+}
+
+/**
+ * Clean an array of tags: trim + remove empty entries
+ *
+ * @param array|null $tags List of tags
+ * @param string     $separator
+ *
+ * @return array
+ */
+function tags_filter(?array $tags, string $separator): array
+{
+    $trimDefault = " \t\n\r\0\x0B";
+    return array_values(array_filter(array_map(function (string $entry) use ($separator, $trimDefault): string {
+        return trim($entry, $trimDefault . $separator);
+    }, $tags ?? [])));
+}