<?php
/**
- * @copyright Copyright (C) 2010-2023, the Friendica project
+ * @copyright Copyright (C) 2010-2024, the Friendica project
*
* @license GNU AGPL version 3 or any later version
*
const DIASPORA = 3;
const CONNECTORS = 4;
const TWITTER_API = 5;
+ const NPF = 6;
const OSTATUS = 7;
const TWITTER = 8;
const BACKLINK = 8;
const ACTIVITYPUB = 9;
+ const BLUESKY = 10;
- const TOP_ANCHOR = '<br class="top-anchor">';
+ const SHARED_ANCHOR = '<hr class="shared-anchor">';
+ const TOP_ANCHOR = '<br class="top-anchor">';
const BOTTOM_ANCHOR = '<br class="button-anchor">';
const PREVIEW_NONE = 0;
break;
case 'title':
- $value = self::convert(html_entity_decode($value, ENT_QUOTES, 'UTF-8'), false, true);
+ $value = self::toPlaintext(html_entity_decode($value, ENT_QUOTES, 'UTF-8'));
$value = html_entity_decode($value, ENT_QUOTES, 'UTF-8');
$value = str_replace(['[', ']'], ['[', ']'], $value);
$data['title'] = $value;
{
DI::profiler()->startRecording('rendering');
// Remove pictures in advance to avoid unneeded proxy calls
+ $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $text);
$text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $2 ', $text);
$text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
// Remove attachment
$text = self::replaceAttachment($text);
- $naked_text = HTML::toPlaintext(self::convert($text, false, 0, true), 0, !$keep_urls);
+ $naked_text = HTML::toPlaintext(self::convert($text, false, self::EXTERNAL, true), 0, !$keep_urls);
DI::profiler()->stopRecording();
return $naked_text;
}
+ /**
+ * Converts text into a format that can be used for the channel search and the language detection.
+ *
+ * @param string $text
+ * @param integer $uri_id
+ * @return string
+ */
+ public static function toSearchText(string $text, int $uri_id): string
+ {
+ // Removes attachments
+ $text = self::removeAttachment($text);
+
+ // Add images because of possible alt texts
+ if (!empty($uri_id)) {
+ $text = Post\Media::addAttachmentsToBody($uri_id, $text, [Post\Media::IMAGE]);
+ }
+
+ if (empty($text)) {
+ return '';
+ }
+
+ // Remove links without a link description
+ $text = preg_replace("~\[url\=.*\]https?:.*\[\/url\]~", ' ', $text);
+
+ // Remove pictures
+ $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $text);
+
+ // Replace picture with the alt description
+ $text = preg_replace("/\[img\=.*?\](.*?)\[\/img\]/ism", ' $1 ', $text);
+
+ // Remove the other pictures
+ $text = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $text);
+
+ // Removes mentions, remove links from hashtags
+ $text = preg_replace('/[@!]\[url\=.*?\].*?\[\/url\]/ism', ' ', $text);
+ $text = preg_replace('/[#]\[url\=.*?\](.*?)\[\/url\]/ism', ' #$1 ', $text);
+ $text = preg_replace('/[@!#]?\[url.*?\[\/url\]/ism', ' ', $text);
+ $text = preg_replace("/\[url=[^\[\]]*\](.*)\[\/url\]/Usi", ' $1 ', $text);
+
+ // Convert it to plain text
+ $text = self::toPlaintext($text, false);
+
+ // Remove possibly remaining links
+ $text = preg_replace(Strings::autoLinkRegEx(), '', $text);
+
+ // Remove all unneeded white space
+ do {
+ $oldtext = $text;
+ $text = str_replace([' ', "\n", "\r", '"'], ' ', $text);
+ } while ($oldtext != $text);
+
+ return trim($text);
+ }
+
private static function proxyUrl(string $image, int $simplehtml = self::INTERNAL, int $uriid = 0, string $size = ''): string
{
// Only send proxied pictures to API and for internal display
*/
private static function convertUrlForActivityPub(string $url): string
{
- return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, self::getStyledURL($url));
- }
-
- /**
- * Converts an URL in a nicer format (without the scheme and possibly shortened)
- *
- * @param string $url URL that is about to be reformatted
- * @return string reformatted link
- */
- private static function getStyledURL(string $url): string
- {
- $parts = parse_url($url);
- $scheme = $parts['scheme'] . '://';
- $styled_url = str_replace($scheme, '', $url);
-
- if (strlen($styled_url) > 30) {
- $styled_url = substr($styled_url, 0, 30) . "…";
- }
-
- return $styled_url;
+ return sprintf('<a href="%s" target="_blank" rel="noopener noreferrer">%s</a>', $url, Strings::getStyledURL($url));
}
/*
/*
* The previously spacefied [noparse][ i ]italic[ /i ][/noparse],
- * now turns back and the [noparse] tags are trimed
+ * now turns back and the [noparse] tags are trimmed
* returning [i]italic[/i]
*
* @param array $match
$network = $contact['network'] ?? Protocol::PHANTOM;
$tpl = Renderer::getMarkupTemplate('shared_content.tpl');
- $text .= Renderer::replaceMacros($tpl, [
+ $text .= self::SHARED_ANCHOR . Renderer::replaceMacros($tpl, [
'$profile' => $attributes['profile'],
'$avatar' => $attributes['avatar'],
'$author' => $attributes['author'],
public static function removeLinks(string $bbcode): string
{
DI::profiler()->startRecording('rendering');
+ $bbcode = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", ' ', $bbcode);
$bbcode = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/ism", ' $1 ', $bbcode);
$bbcode = preg_replace("/\[img.*?\[\/img\]/ism", ' ', $bbcode);
return $match[1] . '[url=' . $data['url'] . ']' . $data['nick'] . '[/url]';
}
+ /**
+ * Normalize links to Youtube and Vimeo to a unified format.
+ *
+ * @param string $text
+ * @return string
+ */
+ private static function normalizeVideoLinks(string $text): string
+ {
+ $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
+ $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
+ $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
+ $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
+
+ $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
+ $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
+
+ return $text;
+ }
+
+ /**
+ * Expand Youtube and Vimeo links to
+ *
+ * @param string $text
+ * @return string
+ */
+ public static function expandVideoLinks(string $text): string
+ {
+ $text = self::normalizeVideoLinks($text);
+ $text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '[url=https://www.youtube.com/watch?v=$1]https://www.youtube.com/watch?v=$1[/url]', $text);
+ $text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '[url=https://vimeo.com/$1]https://vimeo.com/$1[/url]', $text);
+
+ return $text;
+ }
+
/**
* Converts a BBCode message for a given URI-ID to a HTML message
*
"\n[tr]", "[tr]\n", " [tr]", "[tr] ", "\n[/tr]", "[/tr]\n", " [/tr]", "[/tr] ",
"\n[hr]", "[hr]\n", " [hr]", "[hr] ",
"\n[attachment ", " [attachment ", "\n[/attachment]", "[/attachment]\n", " [/attachment]", "[/attachment] ",
- "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] "
+ "[table]\n", "[table] ", " [table]", "\n[/table]", " [/table]", "[/table] ",
+ " \n", "\t\n", "[/li]\n", "\n[li]", "\n[*]",
];
$replace = [
"[th]", "[th]", "[th]", "[/th]", "[/th]", "[/th]",
"[tr]", "[tr]", "[tr]", "[tr]", "[/tr]", "[/tr]", "[/tr]", "[/tr]",
"[hr]", "[hr]", "[hr]", "[hr]",
"[attachment ", "[attachment ", "[/attachment]", "[/attachment]", "[/attachment]", "[/attachment]",
- "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]"
+ "[table]", "[table]", "[table]", "[/table]", "[/table]", "[/table]",
+ "\n", "\n", "[/li]", "[li]", "[*]",
];
do {
$oldtext = $text;
// removing multiplicated newlines
if (DI::config()->get('system', 'remove_multiplicated_lines')) {
$search = [
- "\n\n\n", "\n ", " \n", "[/quote]\n\n", "\n[/quote]", "[/li]\n", "\n[li]", "\n[*]", "\n[ul]", "[/ul]\n", "\n\n[share ", "[/attachment]\n",
+ "\n\n\n", "[/quote]\n\n", "\n[/quote]", "\n[ul]", "[/ul]\n", "\n[ol]", "[/ol]\n", "\n\n[share ", "[/attachment]\n",
"\n[h1]", "[/h1]\n", "\n[h2]", "[/h2]\n", "\n[h3]", "[/h3]\n", "\n[h4]", "[/h4]\n", "\n[h5]", "[/h5]\n", "\n[h6]", "[/h6]\n"
];
$replace = [
- "\n\n", "\n", "\n", "[/quote]\n", "[/quote]", "[/li]", "[li]", "[*]", "[ul]", "[/ul]", "\n[share ", "[/attachment]",
+ "\n\n", "[/quote]\n", "[/quote]", "[ul]", "[/ul]", "[ol]", "[/ol]", "\n[share ", "[/attachment]",
"[h1]", "[/h1]", "[h2]", "[/h2]", "[h3]", "[/h3]", "[h4]", "[/h4]", "[h5]", "[/h5]", "[h6]", "[/h6]"
];
do {
/// @todo Have a closer look at the different html modes
// Handle attached links or videos
- if (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
+ if ($simple_html == self::NPF) {
+ $text = self::removeAttachment($text);
+ } elseif (in_array($simple_html, [self::MASTODON_API, self::TWITTER_API, self::ACTIVITYPUB])) {
$text = self::replaceAttachment($text);
} elseif (!in_array($simple_html, [self::INTERNAL, self::EXTERNAL, self::CONNECTORS])) {
$text = self::replaceAttachment($text, true);
});
}
- // leave open the posibility of [map=something]
+ // leave open the possibility of [map=something]
// this is replaced in Item::prepareBody() which has knowledge of the item location
if (strpos($text, '[/map]') !== false) {
$text = preg_replace_callback(
}
// Check for headers
- $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '</p><h1>$1</h1><p>', $text);
- $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '</p><h2>$1</h2><p>', $text);
- $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '</p><h3>$1</h3><p>', $text);
- $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '</p><h4>$1</h4><p>', $text);
- $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '</p><h5>$1</h5><p>', $text);
- $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '</p><h6>$1</h6><p>', $text);
+
+ if ($simple_html == self::INTERNAL) {
+ //Ensure to always start with <h4> if possible
+ $heading_count = 0;
+ for ($level = 6; $level > 0; $level--) {
+ if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
+ $heading_count++;
+ }
+ }
+ if ($heading_count > 0) {
+ $heading = min($heading_count + 3, 6);
+ for ($level = 6; $level > 0; $level--) {
+ if (preg_match("(\[h$level\].*?\[\/h$level\])ism", $text)) {
+ $text = preg_replace("(\[h$level\](.*?)\[\/h$level\])ism", "</p><h$heading>$1</h$heading><p>", $text);
+ $heading--;
+ }
+ }
+ }
+ } else {
+ $text = preg_replace("(\[h1\](.*?)\[\/h1\])ism", '</p><h1>$1</h1><p>', $text);
+ $text = preg_replace("(\[h2\](.*?)\[\/h2\])ism", '</p><h2>$1</h2><p>', $text);
+ $text = preg_replace("(\[h3\](.*?)\[\/h3\])ism", '</p><h3>$1</h3><p>', $text);
+ $text = preg_replace("(\[h4\](.*?)\[\/h4\])ism", '</p><h4>$1</h4><p>', $text);
+ $text = preg_replace("(\[h5\](.*?)\[\/h5\])ism", '</p><h5>$1</h5><p>', $text);
+ $text = preg_replace("(\[h6\](.*?)\[\/h6\])ism", '</p><h6>$1</h6><p>', $text);
+ }
// Check for paragraph
$text = preg_replace("(\[p\](.*?)\[\/p\])ism", '<p>$1</p>', $text);
$text = preg_replace("(\[style=(.*?)\](.*?)\[\/style\])ism", '<span style="$1">$2</span>', $text);
// Mastodon Emoji (internal tag, do not document for users)
- $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
+ if ($simple_html == self::MASTODON_API) {
+ $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '$2', $text);
+ } else {
+ $text = preg_replace("(\[emoji=(.*?)](.*?)\[/emoji])ism", '<span class="mastodon emoji"><img src="$1" alt="$2" title="$2"/></span>', $text);
+ }
// Check for CSS classes
// @deprecated since 2021.12, left for backward-compatibility reasons
"/\[[iz]mg\=(.*?)\](.*?)\[\/[iz]mg\]/ism",
function ($matches) use ($simple_html, $uriid) {
$matches[1] = self::proxyUrl($matches[1], $simple_html, $uriid);
- $matches[2] = htmlspecialchars($matches[2], ENT_COMPAT);
- return '<img src="' . $matches[1] . '" alt="' . $matches[2] . '" title="' . $matches[2] . '">';
+ $alt = htmlspecialchars($matches[2], ENT_COMPAT);
+ // Fix for Markdown problems with Diaspora, see issue #12701
+ if (($simple_html != self::DIASPORA) || strpos($matches[2], '"') === false) {
+ return '<img src="' . $matches[1] . '" alt="' . $alt . '" title="' . $alt . '">';
+ } else {
+ return '<img src="' . $matches[1] . '" alt="' . $alt . '">';
+ }
},
$text
);
// Simplify "video" element
$text = preg_replace('(\[video[^\]]*?\ssrc\s?=\s?([^\s\]]+)[^\]]*?\].*?\[/video\])ism', '[video]$1[/video]', $text);
- if ($try_oembed) {
+ if ($simple_html == self::NPF) {
+ $text = preg_replace(
+ "/\[video\](.*?)\[\/video\]/ism",
+ '</p><video src="$1" controls width="100%" height="auto">$1</video><p>',
+ $text
+ );
+ $text = preg_replace(
+ "/\[audio\](.*?)\[\/audio\]/ism",
+ '</p><audio src="$1" controls>$1">$1</audio><p>',
+ $text
+ );
+ } elseif ($try_oembed) {
// html5 video and audio
$text = preg_replace(
"/\[video\](.*?\.(ogg|ogv|oga|ogm|webm|mp4).*?)\[\/video\]/ism",
// Backward compatibility, [iframe] support has been removed in version 2020.12
$text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/ism", '<a href="$1">$1</a>', $text);
- // Youtube extensions
- if ($try_oembed) {
- $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
- $text = preg_replace_callback("/\[youtube\](www.youtube.com\/watch\?v\=.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
- $text = preg_replace_callback("/\[youtube\](https?:\/\/www.youtube.com\/shorts\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
- $text = preg_replace_callback("/\[youtube\](https?:\/\/youtu.be\/.*?)\[\/youtube\]/ism", $try_oembed_callback, $text);
- }
-
- $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
- $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
- $text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/shorts\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
- $text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism", '[youtube]$1[/youtube]', $text);
+ $text = self::normalizeVideoLinks($text);
+ // Youtube extensions
if ($try_oembed) {
$text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://www.youtube.com/embed/$1" frameborder="0" ></iframe>', $text);
} else {
);
}
- if ($try_oembed) {
- $text = preg_replace_callback("/\[vimeo\](https?:\/\/player.vimeo.com\/video\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
- $text = preg_replace_callback("/\[vimeo\](https?:\/\/vimeo.com\/[0-9]+).*?\[\/vimeo\]/ism", $try_oembed_callback, $text);
- }
-
- $text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
- $text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism", '[vimeo]$1[/vimeo]', $text);
-
+ // Vimeo extensions
if ($try_oembed) {
$text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", '<iframe width="' . $a->getThemeInfoValue('videowidth') . '" height="' . $a->getThemeInfoValue('videoheight') . '" src="https://player.vimeo.com/video/$1" frameborder="0" ></iframe>', $text);
} else {
$text = preg_replace("/\[event\-id\](.*?)\[\/event\-id\]/ism", '', $text);
}
- if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA)) {
- $conv = html_entity_decode(str_replace([' ', "\n", "\r"], '', $text));
- // Emojis are always 4 byte Unicode characters
- if (!empty($conv) && (strlen($conv) / mb_strlen($conv) == 4)) {
- $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
- }
+ if (!$for_plaintext && DI::config()->get('system', 'big_emojis') && ($simple_html != self::DIASPORA) && Smilies::isEmojiPost($text)) {
+ $text = '<span style="font-size: xx-large; line-height: normal;">' . $text . '</span>';
}
// Handle mentions and hashtag links
$text
);
- if (in_array($simple_html, [self::OSTATUS, self::TWITTER])) {
+ if (in_array($simple_html, [self::OSTATUS, self::TWITTER, self::BLUESKY])) {
$text = preg_replace_callback("/([^#@!])\[url\=([^\]]*)\](.*?)\[\/url\]/ism", [self::class, 'expandLinksCallback'], $text);
//$text = preg_replace("/[^#@!]\[url\=([^\]]*)\](.*?)\[\/url\]/ism", ' $2 [url]$1[/url]', $text);
$text = preg_replace("/\[bookmark\=([^\]]*)\](.*?)\[\/bookmark\]/ism", ' $2 [url]$1[/url]', $text);
$text
);
- // sanitize href attributes (only allowlisted protocols URLs)
+ // sanitize href attributes (only whitelisted protocols URLs)
// default value for backward compatibility
$allowed_link_protocols = DI::config()->get('system', 'allowed_link_protocols', []);
{
DI::profiler()->startRecording('rendering');
- $text = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
+ $text = self::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) {
$text = preg_replace("/[\s|\n]*\[abstract\].*?\[\/abstract\][\s|\n]*/ism", ' ', $text);
$text = preg_replace("/[\s|\n]*\[abstract=.*?\].*?\[\/abstract][\s|\n]*/ism", ' ', $text);
return $text;
DI::profiler()->startRecording('rendering');
$addon = strtolower($addon);
- $abstract = BBCode::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
+ $abstract = self::performWithEscapedTags($text, ['code', 'noparse', 'nobb', 'pre'], function ($text) use ($addon) {
if ($addon && preg_match('#\[abstract=' . preg_quote($addon, '#') . '](.*?)\[/abstract]#ism', $text, $matches)) {
return $matches[1];
}
// Convert it to HTML - don't try oembed
if ($for_diaspora) {
- $text = self::convert($text, false, self::DIASPORA);
+ $text = self::convertForUriId(0, $text, self::DIASPORA);
// Add all tags that maybe were removed
if (preg_match_all("/#\[url\=([$url_search_string]*)\](.*?)\[\/url\]/ism", $original_text, $tags)) {
$text = $text . ' ' . $tagline;
}
} else {
- $text = self::convert($text, false, self::CONNECTORS);
+ $text = self::convertForUriId(0, $text, self::CONNECTORS);
}
// If a link is followed by a quote then there should be a newline before it
// Maybe we should make this newline at every time before a quote.
$text = str_replace(['</a><blockquote>'], ['</a><br><blockquote>'], $text);
+ // The converter doesn't convert these elements
+ $text = str_replace(['<div>', '</div>'], ['<p>', '</p>'], $text);
+
// Now convert HTML to Markdown
$text = HTML::toMarkdown($text);
projects / friendica.git / blobdiff