X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;ds=sidebyside;f=include%2Fbb2diaspora.php;h=80ee1b50dcc0b505f338ee79f1257287ba023f5a;hb=44efdc3e1b01328a226c79e25834fe7adb2dbaf4;hp=eeed19d81686a150b9be58609243f2d5a684f74a;hpb=d3db9365cb76d6d174123056536423028c060311;p=friendica.git

diff --git a/include/bb2diaspora.php b/include/bb2diaspora.php
index eeed19d816..80ee1b50dc 100644
--- a/include/bb2diaspora.php
+++ b/include/bb2diaspora.php
@@ -1,52 +1,85 @@
 <?php
 
-require_once("include/oembed.php");
-require_once("include/event.php");
-require_once("library/markdown.php");
-require_once("include/html2bbcode.php");
-require_once("include/bbcode.php");
-require_once("include/markdownify/markdownify.php");
+use Friendica\App;
+use Friendica\Network\Probe;
+
+use League\HTMLToMarkdown\HtmlConverter;
+
+require_once 'include/oembed.php';
+require_once 'include/event.php';
+require_once 'library/markdown.php';
+require_once 'include/html2bbcode.php';
+require_once 'include/bbcode.php';
+
+/**
+ * @brief Callback function to replace a Diaspora style mention in a mention for Friendica
+ *
+ * @param array $match Matching values for the callback
+ * @return string Replaced mention
+ */
+function diaspora_mention2bb($match) {
+	if ($match[2] == '') {
+		return;
+	}
+
+	$data = get_contact_details_by_addr($match[2]);
 
+	$name = $match[1];
 
-// we don't want to support a bbcode specific markdown interpreter
-// and the markdown library we have is pretty good, but provides HTML output.
-// So we'll use that to convert to HTML, then convert the HTML back to bbcode,
-// and then clean up a few Diaspora specific constructs.
+	if ($name == '') {
+		$name = $data['name'];
+	}
+
+	return '@[url=' . $data['url'] . ']' . $name . '[/url]';
+}
 
+/*
+ * we don't want to support a bbcode specific markdown interpreter
+ * and the markdown library we have is pretty good, but provides HTML output.
+ * So we'll use that to convert to HTML, then convert the HTML back to bbcode,
+ * and then clean up a few Diaspora specific constructs.
+ */
 function diaspora2bb($s) {
 
-	$s = html_entity_decode($s,ENT_COMPAT,'UTF-8');
+	$s = html_entity_decode($s, ENT_COMPAT, 'UTF-8');
 
-	// Simply remove cr.
-	$s = str_replace("\r","",$s);
+	// Handles single newlines
+	$s = str_replace("\r\n", "\n", $s);
+	$s = str_replace("\n", " \n", $s);
+	$s = str_replace("\r", " \n", $s);
 
-	// <br/> is invalid. Replace it with the valid expression
-	$s = str_replace(array("<br/>", "</p>", "<p>"),array("<br />", "<br />", "<br />"),$s);
+	// Replace lonely stars in lines not starting with it with literal stars
+	$s = preg_replace('/^([^\*]+)\*([^\*]*)$/im', '$1\*$2', $s);
 
-	$s = preg_replace('/\@\{(.+?)\; (.+?)\@(.+?)\}/','@[url=https://$3/u/$2]$1[/url]',$s);
+	// The parser cannot handle paragraphs correctly
+	$s = str_replace(array('</p>', '<p>', '<p dir="ltr">'), array('<br>', '<br>', '<br>'), $s);
 
 	// Escaping the hash tags
-	$s = preg_replace('/\#([^\s\#])/','&#35;$1',$s);
+	$s = preg_replace('/\#([^\s\#])/', '&#35;$1', $s);
 
 	$s = Markdown($s);
 
-	$s = str_replace('&#35;','#',$s);
+	$regexp = "/@\{(?:([^\}]+?); )?([^\} ]+)\}/";
+	$s = preg_replace_callback($regexp, 'diaspora_mention2bb', $s);
+
+	$s = str_replace('&#35;', '#', $s);
 
 	$s = html2bbcode($s);
 
 	// protect the recycle symbol from turning into a tag, but without unescaping angles and naked ampersands
-	$s = str_replace('&#x2672;',html_entity_decode('&#x2672;',ENT_QUOTES,'UTF-8'),$s);
+	$s = str_replace('&#x2672;', html_entity_decode('&#x2672;', ENT_QUOTES, 'UTF-8'), $s);
 
 	// Convert everything that looks like a link to a link
-	$s = preg_replace("/([^\]\=]|^)(https?\:\/\/)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3]$2$3[/url]',$s);
+	$s = preg_replace('/([^\]=]|^)(https?\:\/\/)([a-zA-Z0-9:\/\-?&;.=_~#%$!+,@]+(?<!,))/ism', '$1[url=$2$3]$2$3[/url]', $s);
 
 	//$s = preg_replace("/([^\]\=]|^)(https?\:\/\/)(vimeo|youtu|www\.youtube|soundcloud)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3$4]$2$3$4[/url]',$s);
-	$s = bb_tag_preg_replace("/\[url\=?(.*?)\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/url\]/ism",'[youtube]$2[/youtube]','url',$s);
-	$s = bb_tag_preg_replace("/\[url\=https?:\/\/www.youtube.com\/watch\?v\=(.*?)\].*?\[\/url\]/ism",'[youtube]$1[/youtube]','url',$s);
-	$s = bb_tag_preg_replace("/\[url\=?(.*?)\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/url\]/ism",'[vimeo]$2[/vimeo]','url',$s);
-	$s = bb_tag_preg_replace("/\[url\=https?:\/\/vimeo.com\/([0-9]+)\](.*?)\[\/url\]/ism",'[vimeo]$1[/vimeo]','url',$s);
+	$s = bb_tag_preg_replace('/\[url\=?(.*?)\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/url\]/ism', '[youtube]$2[/youtube]', 'url', $s);
+	$s = bb_tag_preg_replace('/\[url\=https?:\/\/www.youtube.com\/watch\?v\=(.*?)\].*?\[\/url\]/ism'   , '[youtube]$1[/youtube]', 'url', $s);
+	$s = bb_tag_preg_replace('/\[url\=?(.*?)\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/url\]/ism'        , '[vimeo]$2[/vimeo]'    , 'url', $s);
+	$s = bb_tag_preg_replace('/\[url\=https?:\/\/vimeo.com\/([0-9]+)\](.*?)\[\/url\]/ism'              , '[vimeo]$1[/vimeo]'    , 'url', $s);
+
 	// remove duplicate adjacent code tags
-	$s = preg_replace("/(\[code\])+(.*?)(\[\/code\])+/ism","[code]$2[/code]", $s);
+	$s = preg_replace('/(\[code\])+(.*?)(\[\/code\])+/ism', '[code]$2[/code]', $s);
 
 	// Don't show link to full picture (until it is fixed)
 	$s = scale_external_images($s, false);
@@ -54,82 +87,148 @@ function diaspora2bb($s) {
 	return $s;
 }
 
-function bb2diaspora($Text,$preserve_nl = false, $fordiaspora = true) {
+/**
+ * @brief Callback function to replace a Friendica style mention in a mention for Diaspora
+ *
+ * @param array $match Matching values for the callback
+ * @return string Replaced mention
+ */
+function diaspora_mentions($match) {
+
+	$contact = get_contact_details_by_url($match[3]);
+
+	if (!x($contact, 'addr')) {
+		$contact = Probe::uri($match[3]);
+	}
+
+	if (!x($contact, 'addr')) {
+		return $match[0];
+	}
+
+	$mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
+	return $mention;
+}
+
+/**
+ * @brief Converts a BBCode text into Markdown
+ *
+ * This function converts a BBCode item body to be sent to Markdown-enabled
+ * systems like Diaspora and Libertree
+ *
+ * @param string $Text
+ * @param bool $preserve_nl Effects unclear, unused in Friendica
+ * @param bool $fordiaspora Diaspora requires more changes than Libertree
+ * @return string
+ */
+function bb2diaspora($Text, $preserve_nl = false, $fordiaspora = true) {
+	$a = get_app();
+
+	$OriginalText = $Text;
 
 	// Since Diaspora is creating a summary for links, this function removes them before posting
-	if ($fordiaspora)
+	if ($fordiaspora) {
 		$Text = bb_remove_share_information($Text);
+	}
 
 	/**
 	 * Transform #tags, strip off the [url] and replace spaces with underscore
 	 */
 	$URLSearchString = "^\[\]";
-	$Text = preg_replace_callback("/#\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i", create_function('$match',
-		'return \'#\'. str_replace(\' \', \'_\', $match[2]);'
-	), $Text);
-
+	$Text = preg_replace_callback("/#\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/i",
+		function ($matches) {
+			return '#' . str_replace(' ', '_', $matches[2]);
+		}
+	, $Text);
 
 	// 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 bbcode()
+	$codeblocks = [];
+	$Text = preg_replace_callback('#\[code(?:=([^\]]*))?\](?=\n)(.*?)\[\/code\]#is',
+		function ($matches) use (&$codeblocks) {
+			$return = '#codeblock-' . count($codeblocks) . '#';
+
+            $prefix = '````' . $matches[1] . PHP_EOL;
+			$codeblocks[] = $prefix . trim($matches[2]) . PHP_EOL . '````';
+			return $return;
+		}
+	, $Text);
+
 	// Convert it to HTML - don't try oembed
-	if ($fordiaspora)
+	if ($fordiaspora) {
 		$Text = bbcode($Text, $preserve_nl, false, 3);
-	else {
+
+		// Add all tags that maybe were removed
+		if (preg_match_all("/#\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $OriginalText, $tags)) {
+			$tagline = "";
+			foreach ($tags[2] as $tag) {
+				$tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
+				if (!strpos(html_entity_decode($Text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
+					$tagline .= '#' . $tag . ' ';
+				}
+			}
+			$Text = $Text." ".$tagline;
+		}
+	} else {
 		$Text = bbcode($Text, $preserve_nl, false, 4);
-		// Libertree doesn't convert a harizontal rule if there isn't a linefeed
-		$Text = str_replace("<hr />", "\n<hr />", $Text);
 	}
 
+	// mask some special HTML chars from conversation to markdown
+	$Text = str_replace(array('&lt;', '&gt;', '&amp;'), array('&_lt_;', '&_gt_;', '&_amp_;'), $Text);
+
+	// 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(array("</a><blockquote>"), array("</a><br><blockquote>"), $Text);
+
+	$stamp1 = microtime(true);
+
 	// Now convert HTML to Markdown
-	$md = new Markdownify(false, false, false);
-	$Text = $md->parseString($Text);
-
-	// The Markdownify converter converts underscores '_' in URLs to '\_', which
-	// messes up the URL. Manually fix these
-	$count = 1;
-	$pos = bb_find_open_close($Text, '[', ']', $count);
-	while($pos !== false) {
-		$start = substr($Text, 0, $pos['start']);
-		$subject = substr($Text, $pos['start'], $pos['end'] - $pos['start'] + 1);
-		$end = substr($Text, $pos['end'] + 1);
-
-		$subject = str_replace('\_', '_', $subject);
-		$Text = $start . $subject . $end;
-
-		$count++;
-		$pos = bb_find_open_close($Text, '[', ']', $count);
-	}
+	$converter = new HtmlConverter();
+	$Text = $converter->convert($Text);
+
+	// unmask the special chars back to HTML
+	$Text = str_replace(array('&_lt_;', '&_gt_;', '&_amp_;'), array('&lt;', '&gt;', '&amp;'), $Text);
 
-	// If the text going into bbcode() has a plain URL in it, i.e.
-	// with no [url] tags around it, it will come out of parseString()
-	// looking like: <http://url.com>, which gets removed by strip_tags().
-	// So take off the angle brackets of any such URL
-	$Text = preg_replace("/<http(.*?)>/is", "http$1", $Text);
+	$a->save_timestamp($stamp1, "parser");
 
-	// Remove all unconverted tags
-	$Text = strip_tags($Text);
+	// Libertree has a problem with escaped hashtags.
+	$Text = str_replace(array('\#'), array('#'), $Text);
 
 	// Remove any leading or trailing whitespace, as this will mess up
 	// the Diaspora signature verification and cause the item to disappear
 	$Text = trim($Text);
 
+	if ($fordiaspora) {
+		$URLSearchString = "^\[\]";
+		$Text = preg_replace_callback("/([@]\[(.*?)\])\(([$URLSearchString]*?)\)/ism", 'diaspora_mentions', $Text);
+	}
+
+	// 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;
+		}
+	, $Text);
+
 	call_hooks('bb2diaspora',$Text);
 
 	return $Text;
 }
 
 function unescape_underscores_in_links($m) {
-	$y = str_replace('\\_','_', $m[2]);
+	$y = str_replace('\\_', '_', $m[2]);
 	return('[' . $m[1] . '](' . $y . ')');
 }
 
 function format_event_diaspora($ev) {
-
-	$a = get_app();
-
-	if(! ((is_array($ev)) && count($ev)))
+	if (! ((is_array($ev)) && count($ev))) {
 		return '';
+	}
 
 	$bd_format = t('l F d, Y \@ g:i A') ; // Friday January 18, 2011 @ 8 AM
 
@@ -138,23 +237,25 @@ function format_event_diaspora($ev) {
 	$o .= '**' . (($ev['summary']) ? bb2diaspora($ev['summary']) : bb2diaspora($ev['desc'])) .  '**' . "\n";
 
 	$o .= t('Starts:') . ' ' . '['
-		. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC', 
+		. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
 			$ev['start'] , $bd_format ))
-			:  day_translate(datetime_convert('UTC', 'UTC', 
+			:  day_translate(datetime_convert('UTC', 'UTC',
 			$ev['start'] , $bd_format)))
-		.  '](' . $a->get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['start'])) . ")\n";
+		.  '](' . App::get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['start'])) . ")\n";
 
-	if(! $ev['nofinish'])
-		$o .= t('Finishes:') . ' ' . '[' 
-			. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC', 
+	if (! $ev['nofinish']) {
+		$o .= t('Finishes:') . ' ' . '['
+			. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
 				$ev['finish'] , $bd_format ))
-				:  day_translate(datetime_convert('UTC', 'UTC', 
+				:  day_translate(datetime_convert('UTC', 'UTC',
 				$ev['finish'] , $bd_format )))
-			. '](' . $a->get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['finish'])) . ")\n";
+			. '](' . App::get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['finish'])) . ")\n";
+	}
 
-	if(strlen($ev['location']))
-		$o .= t('Location:') . bb2diaspora($ev['location']) 
+	if (strlen($ev['location'])) {
+		$o .= t('Location:') . bb2diaspora($ev['location'])
 			. "\n";
+	}
 
 	$o .= "\n";
 	return $o;