<?php
/**
- * @copyright Copyright (C) 2010-2021, the Friendica project
+ * @copyright Copyright (C) 2010-2022, the Friendica project
*
* @license GNU AGPL version 3 or any later version
*
use Friendica\Core\Protocol;
use Friendica\DI;
+use Friendica\Util\Network;
class Plaintext
{
+ // Assumed length of an URL when shortened via the network's own url shortener (e.g. Twitter)
+ const URL_LENGTH = 23;
+
/**
* Shortens message
*
* @param string $msg
* @param int $limit
+ * @param int $uid
* @return string
*
* @todo For Twitter URLs aren't shortened, but they have to be calculated as if.
*/
- public static function shorten($msg, $limit)
+ public static function shorten(string $msg, int $limit, int $uid = 0): string
{
+ $ellipsis = html_entity_decode("…", ENT_QUOTES, 'UTF-8');
+
+ if (!empty($uid) && DI::pConfig()->get($uid, 'system', 'simple_shortening')) {
+ return mb_substr(mb_substr(trim($msg), 0, $limit), 0, -3) . $ellipsis;
+ }
+
$lines = explode("\n", $msg);
$msg = "";
$recycle = html_entity_decode("♲ ", ENT_QUOTES, 'UTF-8');
- $ellipsis = html_entity_decode("…", ENT_QUOTES, 'UTF-8');
foreach ($lines as $row => $line) {
- if (iconv_strlen(trim($msg . "\n" . $line), "UTF-8") <= $limit) {
+ if (mb_strlen(trim($msg . "\n" . $line)) <= $limit) {
$msg = trim($msg . "\n" . $line);
} elseif (($msg == "") || (($row == 1) && (substr($msg, 0, 4) == $recycle))) {
// Is the new message empty by now or is it a reshared message?
- $msg = iconv_substr(iconv_substr(trim($msg . "\n" . $line), 0, $limit, "UTF-8"), 0, -3, "UTF-8") . $ellipsis;
+ $msg = mb_substr(mb_substr(trim($msg . "\n" . $line), 0, $limit), 0, -3) . $ellipsis;
} else {
break;
}
* @return array Same array structure than \Friendica\Content\Text\BBCode::getAttachedData
* @throws \Friendica\Network\HTTPException\InternalServerErrorException
* @see \Friendica\Content\Text\BBCode::getAttachedData
- *
*/
- public static function getPost($item, $limit = 0, $includedlinks = false, $htmlmode = BBCode::API, $target_network = '')
+ public static function getPost(array $item, int $limit = 0, bool $includedlinks = false, int $htmlmode = BBCode::MASTODON_API, string $target_network = '')
{
// Remove hashtags
$URLSearchString = '^\[\]';
$body = BBCode::stripAbstract($body);
// At first look at data that is attached via "type-..." stuff
- // This will hopefully replaced with a dedicated bbcode later
- //$post = self::getAttachedData($b['body']);
$post = BBCode::getAttachedData($body, $item);
if (($item['title'] != '') && ($post['text'] != '')) {
}
}
- $html = BBCode::convert($post['text'] . ($post['after'] ?? ''), false, $htmlmode);
+ $html = BBCode::convertForUriId($item['uri-id'], $post['text'] . ($post['after'] ?? ''), $htmlmode);
$msg = HTML::toPlaintext($html, 0, true);
$msg = trim(html_entity_decode($msg, ENT_QUOTES, 'UTF-8'));
+ $complete_msg = $msg;
+
$link = '';
if ($includedlinks) {
if ($post['type'] == 'link') {
// Will the text be shortened in the link?
// Or is the link the last item in the post?
- if (($limit > 0) && ($pos < $limit) && (($pos + 23 > $limit) || ($pos + strlen($link) == strlen($msg)))) {
+ if (($limit > 0) && ($pos < $limit) && (($pos + self::URL_LENGTH > $limit) || ($pos + mb_strlen($link) == mb_strlen($msg)))) {
$msg = trim(str_replace($link, '', $msg));
} elseif (($limit == 0) || ($pos < $limit)) {
// The limit has to be increased since it will be shortened - but not now
// Only do it with Twitter
- if (($limit > 0) && (strlen($link) > 23) && ($htmlmode == BBCode::TWITTER)) {
- $limit = $limit - 23 + strlen($link);
+ if (($limit > 0) && (mb_strlen($link) > self::URL_LENGTH) && ($htmlmode == BBCode::TWITTER)) {
+ $limit = $limit - self::URL_LENGTH + mb_strlen($link);
}
$link = '';
$msg = str_replace(' ', ' ', $msg);
}
+ if (!in_array($link, ['', $item['plink']]) && ($post['type'] != 'photo') && (strpos($complete_msg, $link) === false)) {
+ $complete_msg .= "\n" . $link;
+ }
+
+ $post['parts'] = self::getParts(trim($complete_msg), $limit);
+
// Twitter is using its own limiter, so we always assume that shortened links will have this length
- if (iconv_strlen($link, 'UTF-8') > 0) {
- $limit = $limit - 23;
+ if (mb_strlen($link) > 0) {
+ $limit = $limit - self::URL_LENGTH;
}
- if (iconv_strlen($msg, 'UTF-8') > $limit) {
+ if (mb_strlen($msg) > $limit) {
if (($post['type'] == 'text') && isset($post['url'])) {
$post['url'] = $item['plink'];
} elseif (!isset($post['url'])) {
- $limit = $limit - 23;
+ $limit = $limit - self::URL_LENGTH;
$post['url'] = $item['plink'];
} elseif (strpos($item['body'], '[share') !== false) {
$post['url'] = $item['plink'];
} elseif (DI::pConfig()->get($item['uid'], 'system', 'no_intelligent_shortening')) {
$post['url'] = $item['plink'];
}
- $msg = self::shorten($msg, $limit);
+ $msg = self::shorten($msg, $limit, $item['uid']);
}
}
return $post;
}
+
+ /**
+ * Split the message in parts
+ *
+ * @param string $message
+ * @param integer $baselimit
+ * @return array
+ */
+ private static function getParts(string $message, int $baselimit): array
+ {
+ $parts = [];
+ $part = '';
+
+ $limit = $baselimit;
+
+ while ($message) {
+ $pos1 = strpos($message, ' ');
+ $pos2 = strpos($message, "\n");
+
+ if (($pos1 !== false) && ($pos2 !== false)) {
+ $pos = min($pos1, $pos2) + 1;
+ } elseif ($pos1 !== false) {
+ $pos = $pos1 + 1;
+ } elseif ($pos2 !== false) {
+ $pos = $pos2 + 1;
+ } else {
+ $word = $message;
+ $message = '';
+ }
+
+ if (trim($message)) {
+ $word = substr($message, 0, $pos);
+ $message = trim(substr($message, $pos));
+ }
+
+ if (Network::isValidHttpUrl(trim($word))) {
+ $limit += mb_strlen(trim($word)) - self::URL_LENGTH;
+ }
+
+ if ((mb_strlen($part . $word) > $limit - 8) && ($parts || (mb_strlen($part . $word . $message) > $limit))) {
+ $parts[] = trim($part);
+ $part = '';
+ $limit = $baselimit;
+ }
+ $part .= $word;
+ }
+ $parts[] = trim($part);
+
+ if (count($parts) > 1) {
+ foreach ($parts as $key => $part) {
+ $parts[$key] .= ' (' . ($key + 1) . '/' . count($parts) . ')';
+ }
+ }
+
+ return $parts;
+ }
}