X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FContent%2FText%2FBBCode.php;h=cb375dcd21e7686b7018512f92ed3d2451a2dcf7;hb=ced4911c4ab268c9e502a3c3e1aa43ebda837044;hp=55fc24273da7954b3ae45701706c3e6fb384be2b;hpb=59d06b6c7809179339f3e8bf82467a85f17294b8;p=friendica.git diff --git a/src/Content/Text/BBCode.php b/src/Content/Text/BBCode.php index 55fc24273d..cb375dcd21 100644 --- a/src/Content/Text/BBCode.php +++ b/src/Content/Text/BBCode.php @@ -15,7 +15,9 @@ use Friendica\Core\Addon; use Friendica\Core\Cache; use Friendica\Core\Config; use Friendica\Core\L10n; +use Friendica\Core\Logger; use Friendica\Core\Protocol; +use Friendica\Core\Renderer; use Friendica\Core\System; use Friendica\Model\Contact; use Friendica\Model\Event; @@ -25,7 +27,8 @@ use Friendica\Util\Map; use Friendica\Util\Network; use Friendica\Util\ParseUrl; use Friendica\Util\Proxy as ProxyUtils; -use League\HTMLToMarkdown\HtmlConverter; +use Friendica\Util\Strings; +use Friendica\Util\XML; class BBCode extends BaseObject { @@ -348,7 +351,7 @@ class BBCode extends BaseObject */ public static function toPlaintext($text, $keep_urls = true) { - $naked_text = preg_replace('/\[(.+?)\]/','', $text); + $naked_text = preg_replace('/\[(.+?)\]\s*/','', $text); if (!$keep_urls) { $naked_text = preg_replace('#https?\://[^\s<]+[^\s\.\)]#i', '', $naked_text); } @@ -380,7 +383,7 @@ class BBCode extends BaseObject $c = preg_match_all('/\[img.*?\](.*?)\[\/img\]/ism', $s, $matches, PREG_SET_ORDER); if ($c) { foreach ($matches as $mtch) { - logger('scale_external_image: ' . $mtch[1]); + Logger::log('scale_external_image: ' . $mtch[1]); $hostname = str_replace('www.', '', substr(System::baseUrl(), strpos(System::baseUrl(), '://') + 3)); if (stristr($mtch[1], $hostname)) { @@ -415,7 +418,7 @@ class BBCode extends BaseObject $Image->scaleDown(640); $new_width = $Image->getWidth(); $new_height = $Image->getHeight(); - logger('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], LOGGER_DEBUG); + Logger::log('scale_external_images: ' . $orig_width . '->' . $new_width . 'w ' . $orig_height . '->' . $new_height . 'h' . ' match: ' . $mtch[0], Logger::DEBUG); $s = str_replace( $mtch[0], '[img=' . $new_width . 'x' . $new_height. ']' . $scaled . '[/img]' @@ -424,7 +427,7 @@ class BBCode extends BaseObject : ''), $s ); - logger('scale_external_images: new string: ' . $s, LOGGER_DEBUG); + Logger::log('scale_external_images: new string: ' . $s, Logger::DEBUG); } } } @@ -452,7 +455,7 @@ class BBCode extends BaseObject // than the maximum, then don't waste time looking for the images if ($maxlen && (strlen($body) > $maxlen)) { - logger('the total body length exceeds the limit', LOGGER_DEBUG); + Logger::log('the total body length exceeds the limit', Logger::DEBUG); $orig_body = $body; $new_body = ''; @@ -472,7 +475,7 @@ class BBCode extends BaseObject if (($textlen + $img_start) > $maxlen) { if ($textlen < $maxlen) { - logger('the limit happens before an embedded image', LOGGER_DEBUG); + Logger::log('the limit happens before an embedded image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); $textlen = $maxlen; } @@ -486,7 +489,7 @@ class BBCode extends BaseObject if (($textlen + $img_end) > $maxlen) { if ($textlen < $maxlen) { - logger('the limit happens before the end of a non-embedded image', LOGGER_DEBUG); + Logger::log('the limit happens before the end of a non-embedded image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); $textlen = $maxlen; } @@ -509,11 +512,11 @@ class BBCode extends BaseObject if (($textlen + strlen($orig_body)) > $maxlen) { if ($textlen < $maxlen) { - logger('the limit happens after the end of the last image', LOGGER_DEBUG); + Logger::log('the limit happens after the end of the last image', Logger::DEBUG); $new_body = $new_body . substr($orig_body, 0, $maxlen - $textlen); } } else { - logger('the text size with embedded images extracted did not violate the limit', LOGGER_DEBUG); + Logger::log('the text size with embedded images extracted did not violate the limit', Logger::DEBUG); $new_body = $new_body . $orig_body; } @@ -572,14 +575,17 @@ class BBCode extends BaseObject $return = sprintf('
', $data["type"]); } - if (!empty($data["image"])) { - $return .= sprintf('
', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]); - } elseif (!empty($data["preview"])) { - $return .= sprintf('
', $data["url"], self::proxyUrl($data["preview"], $simplehtml), $data["title"]); - } - if (!empty($data['title']) && !empty($data['url'])) { - $return .= sprintf('

%s

', $data['url'], $data['title']); + if (!empty($data["image"]) && empty($data["text"]) && ($data["type"] == "photo")) { + $return .= sprintf('', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]); + } else { + if (!empty($data["image"])) { + $return .= sprintf('
', $data["url"], self::proxyUrl($data["image"], $simplehtml), $data["title"]); + } elseif (!empty($data["preview"])) { + $return .= sprintf('
', $data["url"], self::proxyUrl($data["preview"], $simplehtml), $data["title"]); + } + $return .= sprintf('

%s

', $data['url'], $data['title']); + } } if (!empty($data["description"]) && $data["description"] != $data["title"]) { @@ -857,187 +863,140 @@ class BBCode extends BaseObject } /** - * Processes [share] tags + * This function converts a [share] block to text according to a provided callback function whose signature is: * - * Note: Can produce a [bookmark] tag in the output + * function(array $attributes, array $author_contact, string $content, boolean $is_quote_share): string * - * @brief Processes [share] tags - * @param array $share preg_match_callback result array - * @param bool|int $simplehtml - * @return string + * Where: + * - $attributes is an array of attributes of the [share] block itself. Missing keys will be completed by the contact + * data lookup + * - $author_contact is a contact record array + * - $content is the inner content of the [share] block + * - $is_quote_share indicates whether there's any content before the [share] block + * - Return value is the string that should replace the [share] block in the provided text + * + * This function is intended to be used by addon connector to format a share block like the target network is expecting it. + * + * @param string $text A BBCode string + * @param callable $callback + * @return string The BBCode string with all [share] blocks replaced */ - private static function convertShare($share, $simplehtml) + public static function convertShare($text, callable $callback) { - $attributes = $share[2]; - - $author = ""; - preg_match("/author='(.*?)'/ism", $attributes, $matches); - if (x($matches, 1)) { - $author = html_entity_decode($matches[1], ENT_QUOTES, 'UTF-8'); - } - - preg_match('/author="(.*?)"/ism', $attributes, $matches); - if (x($matches, 1)) { - $author = $matches[1]; - } - - $profile = ""; - preg_match("/profile='(.*?)'/ism", $attributes, $matches); - if (x($matches, 1)) { - $profile = $matches[1]; - } - - preg_match('/profile="(.*?)"/ism', $attributes, $matches); - if (x($matches, 1)) { - $profile = $matches[1]; - } - - $avatar = ""; - preg_match("/avatar='(.*?)'/ism", $attributes, $matches); - if (x($matches, 1)) { - $avatar = $matches[1]; - } - - preg_match('/avatar="(.*?)"/ism', $attributes, $matches); - if (x($matches, 1)) { - $avatar = $matches[1]; - } - - $link = ""; - preg_match("/link='(.*?)'/ism", $attributes, $matches); - if (x($matches, 1)) { - $link = $matches[1]; - } - - preg_match('/link="(.*?)"/ism', $attributes, $matches); - if (x($matches, 1)) { - $link = $matches[1]; - } - - $posted = ""; - - preg_match("/posted='(.*?)'/ism", $attributes, $matches); - if (x($matches, 1)) { - $posted = $matches[1]; - } - - preg_match('/posted="(.*?)"/ism', $attributes, $matches); - if (x($matches, 1)) { - $posted = $matches[1]; - } + $return = preg_replace_callback( + "/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism", + function ($match) use ($callback) { + $attribute_string = $match[2]; + + $attributes = []; + foreach(['author', 'profile', 'avatar', 'link', 'posted'] as $field) { + preg_match("/$field=(['\"])(.+?)\\1/ism", $attribute_string, $matches); + $attributes[$field] = html_entity_decode(defaults($matches, 2, ''), ENT_QUOTES, 'UTF-8'); + } - // We only call this so that a previously unknown contact can be added. - // This is important for the function "Model\Contact::getDetailsByURL()". - // This function then can fetch an entry from the contact table. - Contact::getIdForURL($profile, 0, true); + // We only call this so that a previously unknown contact can be added. + // This is important for the function "Model\Contact::getDetailsByURL()". + // This function then can fetch an entry from the contact table. + Contact::getIdForURL($attributes['profile'], 0, true); - $data = Contact::getDetailsByURL($profile); + $author_contact = Contact::getDetailsByURL($attributes['profile']); + $author_contact['addr'] = defaults($author_contact, 'addr' , Protocol::getAddrFromProfileUrl($attributes['profile'])); - if (x($data, "name") && x($data, "addr")) { - $userid_compact = $data["name"] . " (" . $data["addr"] . ")"; - } else { - $userid_compact = Protocol::getAddrFromProfileUrl($profile, $author); - } + $attributes['author'] = defaults($author_contact, 'name' , $attributes['author']); + $attributes['avatar'] = defaults($author_contact, 'micro', $attributes['avatar']); + $attributes['profile'] = defaults($author_contact, 'url' , $attributes['profile']); - if (x($data, "addr")) { - $userid = $data["addr"]; - } else { - $userid = Protocol::formatMention($profile, $author); - } + if ($attributes['avatar']) { + $attributes['avatar'] = ProxyUtils::proxifyUrl($attributes['avatar'], false, ProxyUtils::SIZE_THUMB); + } - if (x($data, "name")) { - $author = $data["name"]; - } + return $match[1] . $callback($attributes, $author_contact, $match[3], trim($match[1]) != ''); + }, + $text + ); - if (x($data, "micro")) { - $avatar = $data["micro"]; - } + return $return; + } - $preshare = trim($share[1]); - if ($preshare != "") { - $preshare .= "
"; - } + /** + * Default [share] tag conversion callback + * + * Note: Can produce a [bookmark] tag in the output + * + * @see BBCode::convertShare() + * @param array $attributes [share] block attribute values + * @param array $author_contact Contact row of the shared author + * @param string $content Inner content of the [share] block + * @param boolean $is_quote_share Whether there is content before the [share] block + * @param integer $simplehtml Mysterious integer value depending on the target network/formatting style + * @return string + */ + private static function convertShareCallback(array $attributes, array $author_contact, $content, $is_quote_share, $simplehtml) + { + $mention = Protocol::formatMention($attributes['profile'], $attributes['author']); switch ($simplehtml) { case 1: - $text = $preshare . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . ' ' . $userid . ":
»" . $share[3] . "«"; + $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $mention . ':

' . "\n" . '«' . $content . '»'; break; case 2: - $text = $preshare . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ":
" . $share[3]; + $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ':

' . "\n" . $content; break; case 3: // Diaspora - $headline = '' . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . $userid . ':
'; + $headline = '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . $mention . ':

' . "\n"; - $text = trim($share[1]); - - if ($text != "") { - $text .= "
"; - } - - if (stripos(normalise_link($link), 'http://twitter.com/') === 0) { - $text .= '
' . $link . ''; + if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0) { + $text = ($is_quote_share? '
' : '') . '

' . $attributes['link'] . '

' . "\n"; } else { - $text .= $headline . '
' . trim($share[3]) . "

"; + $text = ($is_quote_share? '
' : '') . $headline . '
' . trim($content) . '
' . "\n"; - if ($link != "") { - $text .= '
[l]'; + if ($attributes['link'] != '') { + $text .= '

[l]

' . "\n"; } } break; case 4: - $headline = '
' . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8'); - $headline .= L10n::t('%2$s %3$s', $link, $userid, $posted); - $headline .= ":
"; - - $text = trim($share[1]); + $headline = '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8'); + $headline .= L10n::t('%2$s %3$s', $attributes['link'], $mention, $attributes['posted']); + $headline .= ':

' . "\n"; - if ($text != "") { - $text .= "
"; - } - - $text .= $headline . '
' . trim($share[3]) . "

"; + $text = ($is_quote_share? '
' : '') . $headline . '
' . trim($content) . '
' . "\n"; break; case 5: - $text = $preshare . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ":
" . $share[3]; + $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ':

' . "\n" . $content; break; case 7: // statusnet/GNU Social - $text = $preshare . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . " @" . $userid_compact . ": " . $share[3]; - break; - case 8: // twitter - $text = $preshare . "RT @" . $userid_compact . ": " . $share[3]; + $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' @' . $author_contact['addr'] . ': ' . $content . '

' . "\n"; break; case 9: // Google+ - $text = $preshare . html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8') . ' ' . $userid_compact . ":
" . $share[3]; + $text = ($is_quote_share? '
' : '') . '

' . html_entity_decode('♲ ', ENT_QUOTES, 'UTF-8') . ' ' . $author_contact['addr'] . ':

' . "\n"; + $text .= '

' . $content . '

' . "\n"; - if ($link != "") { - $text .= "

" . $link; + if ($attributes['link'] != '') { + $text .= '

' . $attributes['link'] . '

'; } break; default: // Transforms quoted tweets in rich attachments to avoid nested tweets - if (stripos(normalise_link($link), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($link)) { + if (stripos(Strings::normaliseLink($attributes['link']), 'http://twitter.com/') === 0 && OEmbed::isAllowedURL($attributes['link'])) { try { - $oembed = OEmbed::getHTML($link, $preshare); + $text = ($is_quote_share? '
' : '') . OEmbed::getHTML($attributes['link']); } catch (Exception $e) { - $oembed = sprintf('[bookmark=%s]%s[/bookmark]', $link, $preshare); + $text = ($is_quote_share? '
' : '') . sprintf('[bookmark=%s]%s[/bookmark]', $attributes['link'], $content); } - - $text = $preshare . $oembed; } else { - $text = trim($share[1]) . "\n"; - - $avatar = ProxyUtils::proxifyUrl($avatar, false, ProxyUtils::SIZE_THUMB); - - $tpl = get_markup_template('shared_content.tpl'); - $text .= replace_macros($tpl, [ - '$profile' => $profile, - '$avatar' => $avatar, - '$author' => $author, - '$link' => $link, - '$posted' => $posted, - '$content' => trim($share[3]) + $text = ($is_quote_share? "\n" : ''); + + $tpl = Renderer::getMarkupTemplate('shared_content.tpl'); + $text .= Renderer::replaceMacros($tpl, [ + '$profile' => $attributes['profile'], + '$avatar' => $attributes['avatar'], + '$author' => $attributes['author'], + '$link' => $attributes['link'], + '$posted' => $attributes['posted'], + '$content' => trim($content) ]); } break; @@ -1058,11 +1017,11 @@ class BBCode extends BaseObject $ch = @curl_init($match[1]); @curl_setopt($ch, CURLOPT_NOBODY, true); @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); - @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent()); + @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent()); @curl_exec($ch); $curl_info = @curl_getinfo($ch); - $a->save_timestamp($stamp1, "network"); + $a->saveTimestamp($stamp1, "network"); if (substr($curl_info["content_type"], 0, 6) == "image/") { $text = "[url=" . $match[1] . "]" . $match[1] . "[/url]"; @@ -1117,11 +1076,11 @@ class BBCode extends BaseObject $ch = @curl_init($match[1]); @curl_setopt($ch, CURLOPT_NOBODY, true); @curl_setopt($ch, CURLOPT_RETURNTRANSFER, true); - @curl_setopt($ch, CURLOPT_USERAGENT, $a->get_useragent()); + @curl_setopt($ch, CURLOPT_USERAGENT, $a->getUserAgent()); @curl_exec($ch); $curl_info = @curl_getinfo($ch); - $a->save_timestamp($stamp1, "network"); + $a->saveTimestamp($stamp1, "network"); // if its a link to a picture then embed this picture if (substr($curl_info["content_type"], 0, 6) == "image/") { @@ -1161,21 +1120,6 @@ class BBCode extends BaseObject return $return; } - private static function textHighlightCallback($match) - { - // Fallback in case the language doesn't exist - $return = '[code]' . $match[2] . '[/code]'; - - if (in_array(strtolower($match[1]), - ['php', 'css', 'mysql', 'sql', 'abap', 'diff', 'html', 'perl', 'ruby', - 'vbscript', 'avrc', 'dtd', 'java', 'xml', 'cpp', 'python', 'javascript', 'js', 'sh', 'bash']) - ) { - $return = text_highlight($match[2], strtolower($match[1])); - } - - return $return; - } - /** * @brief Converts a BBCode message to HTML message * @@ -1224,6 +1168,22 @@ class BBCode extends BaseObject return $return; }; + // Extracting multi-line code blocks before the whitespace processing + $codeblocks = []; + + $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is", + function ($matches) use (&$codeblocks) { + $return = $matches[0]; + if (strpos($matches[2], "\n") !== false) { + $return = '#codeblock-' . count($codeblocks) . '#'; + + $codeblocks[] = '
' . trim($matches[2], "\n\r") . '
'; + } + return $return; + }, + $text + ); + // Hide all [noparse] contained bbtags by spacefying them // POSSIBLE BUG --> Will the 'preg' functions crash if there's an embedded image? @@ -1262,19 +1222,11 @@ class BBCode extends BaseObject $text = preg_replace("/\s?\[share(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "[share$1]$2[/share]", $text); $text = preg_replace("/\s?\[quote(.*?)\]\s?(.*?)\s?\[\/quote\]\s?/ism", "[quote$1]$2[/quote]", $text); - $text = preg_replace("/\n\[code\]/ism", "[code]", $text); - $text = preg_replace("/\[\/code\]\n/ism", "[/code]", $text); - // when the content is meant exporting to other systems then remove the avatar picture since this doesn't really look good on these systems if (!$try_oembed) { $text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism", "\n[share$1$2]$3[/share]", $text); } - // Check for [code] text here, before the linefeeds are messed with. - // The highlighter will unescape and re-escape the content. - if (strpos($text, '[code=') !== false) { - $text = preg_replace_callback("/\[code=(.*?)\](.*?)\[\/code\]/ism", 'self::textHighlightCallback', $text); - } // Convert new line chars to html
tags // nlbr seems to be hopelessly messed up @@ -1371,7 +1323,7 @@ class BBCode extends BaseObject // Handle Diaspora posts $text = preg_replace_callback( - "&\[url=/posts/([^\[\]]*)\](.*)\[\/url\]&Usi", + "&\[url=/?posts/([^\[\]]*)\](.*)\[\/url\]&Usi", function ($match) { return "[url=" . System::baseUrl() . "/display/" . $match[1] . "]" . $match[2] . "[/url]"; }, $text @@ -1389,15 +1341,21 @@ class BBCode extends BaseObject $expression = "=diaspora://.*?/post/([0-9A-Za-z\-_@.:]{15,254}[0-9A-Za-z])=ism"; $text = preg_replace($expression, System::baseUrl()."/display/$1", $text); - $text = preg_replace("/([#])\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", - '$1$3', $text); - - $text = preg_replace("/\[url\=([$URLSearchString]*)\]#(.*?)\[\/url\]/ism", - '#$2', $text); + /* Tag conversion + * Supports: + * - #[url=][/url] + * - [url=]#[/url] + */ + $text = preg_replace_callback("/(?:#\[url\=[$URLSearchString]*\]|\[url\=[$URLSearchString]*\]#)(.*?)\[\/url\]/ism", function($matches) { + return '#' + . XML::escape($matches[1]) + . ''; + }, $text); $text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '$1', $text); $text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '$2', $text); - //$Text = preg_replace("/\[url\=([$URLSearchString]*)\]([$URLSearchString]*)\[\/url\]/ism", '$2', $Text); // Red compatibility, though the link can't be authenticated on Friendica $text = preg_replace("/\[zrl\=([$URLSearchString]*)\](.*?)\[\/zrl\]/ism", '$2', $text); @@ -1413,7 +1371,7 @@ class BBCode extends BaseObject $text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '$2', $text); // leave open the posibility of [map=something] - // this is replaced in prepare_body() which has knowledge of the item location + // this is replaced in Item::prepareBody() which has knowledge of the item location if (strpos($text, '[/map]') !== false) { $text = preg_replace_callback( @@ -1524,7 +1482,7 @@ class BBCode extends BaseObject $text = str_replace('[hr]', '
', $text); - // This is actually executed in prepare_body() + // This is actually executed in Item::prepareBody() $text = str_replace('[nosmile]', '', $text); @@ -1626,10 +1584,12 @@ class BBCode extends BaseObject $text = preg_replace("/\[zmg\](.*?)\[\/zmg\]/ism", '' . L10n::t('Image/photo') . '', $text); // Shared content - $text = preg_replace_callback("/(.*?)\[share(.*?)\](.*?)\[\/share\]/ism", - function ($match) use ($simple_html) { - return self::convertShare($match, $simple_html); - }, $text); + $text = self::convertShare( + $text, + function (array $attributes, array $author_contact, $content, $is_quote_share) use ($simple_html) { + return self::convertShareCallback($attributes, $author_contact, $content, $is_quote_share, $simple_html); + } + ); $text = preg_replace("/\[crypt\](.*?)\[\/crypt\]/ism", '
' . L10n::t('Encrypted content') . '
', $text); $text = preg_replace("/\[crypt(.*?)\](.*?)\[\/crypt\]/ism", '
' . L10n::t('Encrypted content') . '
', $text); @@ -1719,18 +1679,6 @@ class BBCode extends BaseObject $text = Smilies::replace($text, false, true); } - // Replace inline code blocks - $text = preg_replace_callback("|(?!]*>)([^<]*)(?!]*>)|ism", - function ($match) use ($simple_html) { - $return = '' . $match[1] . ''; - // Use for Diaspora inline code blocks - if ($simple_html === 3) { - $return = '' . $match[1] . ''; - } - return $return; - } - , $text); - // Unhide all [noparse] contained bbtags unspacefying them // and triming the [noparse] tag. @@ -1768,6 +1716,18 @@ class BBCode extends BaseObject $text = self::interpolateSavedImagesIntoItemBody($text, $saved_image); } + // Restore code blocks + $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU', + function ($matches) use ($codeblocks) { + $return = $matches[0]; + if (isset($codeblocks[intval($matches[1])])) { + $return = $codeblocks[$matches[1]]; + } + return $return; + }, + $text + ); + // Clean up the HTML by loading and saving the HTML with the DOM. // Bad structured html can break a whole page. // For performance reasons do it only with ativated item cache or at export. @@ -1902,23 +1862,6 @@ class BBCode extends BaseObject // Converting images with size parameters to simple images. Markdown doesn't know it. $text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $text); - // Extracting multi-line code blocks before the whitespace processing/code highlighter in self::convert() - $codeblocks = []; - - $text = preg_replace_callback("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is", - function ($matches) use (&$codeblocks) { - $return = $matches[0]; - if (strpos($matches[2], "\n") !== false) { - $return = '#codeblock-' . count($codeblocks) . '#'; - - $prefix = '````' . $matches[1] . PHP_EOL; - $codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````'; - } - return $return; - }, - $text - ); - // Convert it to HTML - don't try oembed if ($for_diaspora) { $text = self::convert($text, false, 3); @@ -1948,13 +1891,12 @@ class BBCode extends BaseObject $stamp1 = microtime(true); // Now convert HTML to Markdown - $converter = new HtmlConverter(); - $text = $converter->convert($text); + $text = HTML::toMarkdown($text); // unmask the special chars back to HTML $text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['<', '>', '&'], $text); - $a->save_timestamp($stamp1, "parser"); + $a->saveTimestamp($stamp1, "parser"); // Libertree has a problem with escaped hashtags. $text = str_replace(['\#'], ['#'], $text); @@ -1972,20 +1914,82 @@ class BBCode extends BaseObject ); } - // Restore code blocks - $text = preg_replace_callback('/#codeblock-([0-9]+)#/iU', - function ($matches) use ($codeblocks) { - $return = ''; - if (isset($codeblocks[intval($matches[1])])) { - $return = $codeblocks[$matches[1]]; - } - return $return; - }, - $text - ); - Addon::callHooks('bb2diaspora', $text); return $text; } + + /** + * @brief Pull out all #hashtags and @person tags from $string. + * + * We also get @person@domain.com - which would make + * the regex quite complicated as tags can also + * end a sentence. So we'll run through our results + * and strip the period from any tags which end with one. + * Returns array of tags found, or empty array. + * + * @param string $string Post content + * + * @return array List of tag and person names + */ + public static function getTags($string) + { + $ret = []; + + // Convert hashtag links to hashtags + $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string); + + // ignore anything in a code block + $string = preg_replace('/\[code\](.*?)\[\/code\]/sm', '', $string); + + // Force line feeds at bbtags + $string = str_replace(['[', ']'], ["\n[", "]\n"], $string); + + // ignore anything in a bbtag + $string = preg_replace('/\[(.*?)\]/sm', '', $string); + + // Match full names against @tags including the space between first and last + // We will look these up afterward to see if they are full names or not recognisable. + + if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) { + foreach ($matches[1] as $match) { + if (strstr($match, ']')) { + // we might be inside a bbcode color tag - leave it alone + continue; + } + + if (substr($match, -1, 1) === '.') { + $ret[] = substr($match, 0, -1); + } else { + $ret[] = $match; + } + } + } + + // Otherwise pull out single word tags. These can be @nickname, @first_last + // and #hash tags. + + if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) { + foreach ($matches[1] as $match) { + if (strstr($match, ']')) { + // we might be inside a bbcode color tag - leave it alone + continue; + } + if (substr($match, -1, 1) === '.') { + $match = substr($match,0,-1); + } + // ignore strictly numeric tags like #1 + if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) { + continue; + } + // try not to catch url fragments + if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) { + continue; + } + $ret[] = $match; + } + } + + return $ret; + } }