require_once("include/bbcode.php");
require_once("include/markdownify/markdownify.php");
+
// 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,
$s = str_replace("\r","",$s);
// <br/> is invalid. Replace it with the valid expression
- $s = str_replace("<br/>","<br />",$s);
+ logger("diaspora2bb: 1: ".$s);
+ $s = str_replace(array("<br/>", "</p>", "<p>"),array("<br />", "<br />", "<br />"),$s);
+ logger("diaspora2bb: 2: ".$s);
$s = preg_replace('/\@\{(.+?)\; (.+?)\@(.+?)\}/','@[url=https://$3/u/$2]$1[/url]',$s);
// $s = str_replace('*','*',$s);
// protect the recycle symbol from turning into a tag, but without unescaping angles and naked ampersands
- $s = str_replace('♲',html_entities_decode('♲',ENT_QUOTES,'UTF-8'),$s);
+ $s = str_replace('♲',html_entity_decode('♲',ENT_QUOTES,'UTF-8'),$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?\:\/\/)(vimeo|youtu|www\.youtube|soundcloud)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3$4]$2$3$4[/url]',$s);
- $s = preg_replace("/\[url\=?(.*?)\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/url\]/ism",'[youtube]$2[/youtube]',$s);
- $s = preg_replace("/\[url\=https?:\/\/www.youtube.com\/watch\?v\=(.*?)\].*?\[\/url\]/ism",'[youtube]$1[/youtube]',$s);
- $s = preg_replace("/\[url\=?(.*?)\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/url\]/ism",'[vimeo]$2[/vimeo]',$s);
- $s = preg_replace("/\[url\=https?:\/\/vimeo.com\/([0-9]+)\](.*?)\[\/url\]/ism",'[vimeo]$1[/vimeo]',$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]", $s);
}
-function bb2diaspora($Text,$preserve_nl = false) {
+function bb2diaspora($Text,$preserve_nl = false, $fordiaspora = true) {
-//////////////////////
-// An attempt was made to convert bbcode to html and then to markdown
-// consisting of the following lines.
-// I'm undoing this as we have a lot of bbcode constructs which
-// were simply getting lost, for instance bookmark, vimeo, video, youtube, events, etc.
-// We can try this again, but need a very good test sequence to verify
-// all the major bbcode constructs that we use are getting through.
-//////////////////////
-/*
- // bbcode() will convert "[*]" into "<li>" with no closing "</li>"
- // Markdownify() is unable to handle these, as it makes each new
- // "<li>" into a deeper nested element until it crashes. So pre-format
- // the lists as Diaspora lists before sending the $Text to bbcode()
- //
- // Note that to get nested lists to work for Diaspora, we would need
- // to define the closing tag for the list elements. So nested lists
- // are going to be flattened out in Diaspora for now
+ // Since Diaspora is creating a summary for links, this function removes them before posting
+ if ($fordiaspora)
+ $Text = bb_remove_share_information($Text);
- $endlessloop = 0;
- while ((((strpos($Text, "[/list]") !== false) && (strpos($Text, "[list") !== false)) ||
- ((strpos($Text, "[/ol]") !== false) && (strpos($Text, "[ol]") !== false)) ||
- ((strpos($Text, "[/ul]") !== false) && (strpos($Text, "[ul]") !== false))) && (++$endlessloop < 20)) {
- $Text = preg_replace_callback("/\[list\](.*?)\[\/list\]/is", 'diaspora_ul', $Text);
- $Text = preg_replace_callback("/\[list=1\](.*?)\[\/list\]/is", 'diaspora_ol', $Text);
- $Text = preg_replace_callback("/\[list=i\](.*?)\[\/list\]/s",'diaspora_ol', $Text);
- $Text = preg_replace_callback("/\[list=I\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
- $Text = preg_replace_callback("/\[list=a\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
- $Text = preg_replace_callback("/\[list=A\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
- $Text = preg_replace_callback("/\[ul\](.*?)\[\/ul\]/is", 'diaspora_ul', $Text);
- $Text = preg_replace_callback("/\[ol\](.*?)\[\/ol\]/is", 'diaspora_ol', $Text);
- }
+ // Re-enabling the converter again.
+ // The bbcode parser now handles youtube-links (and the other stuff) correctly.
+ // Additionally the html code is now fixed so that lists are now working.
-*/
+ /**
+ * Transform #tags, strip off the [url] and replace spaces with underscore
+ */
+ $Text = preg_replace_callback('/#\[url\=(\w+.*?)\](\w+.*?)\[\/url\]/i', create_function('$match',
+ 'return \'#\'. str_replace(\' \', \'_\', $match[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]', $Text);
+
+ // the following was added on 10-January-2012 due to an inability of Diaspora's
+ // new javascript markdown processor to handle links with images as the link "text"
+ // It is not optimal and may be removed if this ability is restored in the future
+ //if ($fordiaspora)
+ // $Text = preg_replace("/\[url\=([^\[\]]*)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\]/ism",
+ // "[url]$1[/url]\n[img]$2[/img]", $Text);
+
+ // Remove the avatar picture since that isn't looking good on the other side
+ //$Text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism","\n[share$1$2]$3[/share]",$Text);
// Convert it to HTML - don't try oembed
-// $Text = bbcode($Text, $preserve_nl, false);
+ if ($fordiaspora)
+ $Text = bbcode($Text, $preserve_nl, false, 3);
+ else
+ $Text = bbcode($Text, $preserve_nl, false, 4);
// Now convert HTML to Markdown
-// $md = new Markdownify(false, false, false);
-// $Text = $md->parseString($Text);
+ $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);
+ }
// 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);
+ $Text = preg_replace("/<http(.*?)>/is", "http$1", $Text);
// Remove all unconverted tags
-// $Text = strip_tags($Text);
-
-//////
-// end of bb->html->md conversion attempt
-//////
+ $Text = strip_tags($Text);
+/* Old routine
$ev = bbtoevent($Text);
// If we found an event earlier, strip out all the event code and replace with a reformatted version.
- if(x($ev,'desc') && x($ev,'start')) {
+ if(x($ev,'start')) {
$sub = format_event_diaspora($ev);
$Text = preg_replace_callback('/\[(.*?)\]\((.*?)\)/ism','unescape_underscores_in_links',$Text);
+*/
// Remove any leading or trailing whitespace, as this will mess up
// the Diaspora signature verification and cause the item to disappear
projects / friendica.git / blobdiff