3 require_once("include/oembed.php");
4 require_once("include/event.php");
5 require_once("library/markdown.php");
6 require_once("include/html2bbcode.php");
7 require_once("include/bbcode.php");
8 require_once("include/markdownify/markdownify.php");
11 // we don't want to support a bbcode specific markdown interpreter
12 // and the markdown library we have is pretty good, but provides HTML output.
13 // So we'll use that to convert to HTML, then convert the HTML back to bbcode,
14 // and then clean up a few Diaspora specific constructs.
16 function diaspora2bb($s) {
18 // for testing purposes: Collect raw markdown articles
19 // $file = tempnam("/tmp/friendica/", "markdown");
20 // file_put_contents($file, $s);
22 $s = html_entity_decode($s,ENT_COMPAT,'UTF-8');
24 // Too many new lines. So deactivated the following line
25 // $s = str_replace("\r","\n",$s);
27 $s = str_replace("\r","",$s);
29 // <br/> is invalid. Replace it with the valid expression
30 logger("diaspora2bb: 1: ".$s);
31 $s = str_replace(array("<br/>", "</p>", "<p>"),array("<br />", "<br />", "<br />"),$s);
32 logger("diaspora2bb: 2: ".$s);
34 $s = preg_replace('/\@\{(.+?)\; (.+?)\@(.+?)\}/','@[url=https://$3/u/$2]$1[/url]',$s);
36 // Escaping the hash tags - doesn't always seem to work
37 // $s = preg_replace('/\#([^\s\#])/','\\#$1',$s);
39 $s = preg_replace('/\#([^\s\#])/','#$1',$s);
43 $s = str_replace('#','#',$s);
44 // we seem to have double linebreaks
45 // $s = str_replace("\n",'<br />',$s);
48 // $s = str_replace('*','*',$s);
50 // protect the recycle symbol from turning into a tag, but without unescaping angles and naked ampersands
51 $s = str_replace('♲',html_entity_decode('♲',ENT_QUOTES,'UTF-8'),$s);
53 // Convert everything that looks like a link to a link
54 $s = preg_replace("/([^\]\=]|^)(https?\:\/\/)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3]$2$3[/url]',$s);
56 //$s = preg_replace("/([^\]\=]|^)(https?\:\/\/)(vimeo|youtu|www\.youtube|soundcloud)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3$4]$2$3$4[/url]',$s);
57 $s = bb_tag_preg_replace("/\[url\=?(.*?)\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/url\]/ism",'[youtube]$2[/youtube]','url',$s);
58 $s = bb_tag_preg_replace("/\[url\=https?:\/\/www.youtube.com\/watch\?v\=(.*?)\].*?\[\/url\]/ism",'[youtube]$1[/youtube]','url',$s);
59 $s = bb_tag_preg_replace("/\[url\=?(.*?)\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/url\]/ism",'[vimeo]$2[/vimeo]','url',$s);
60 $s = bb_tag_preg_replace("/\[url\=https?:\/\/vimeo.com\/([0-9]+)\](.*?)\[\/url\]/ism",'[vimeo]$1[/vimeo]','url',$s);
61 // remove duplicate adjacent code tags
62 $s = preg_replace("/(\[code\])+(.*?)(\[\/code\])+/ism","[code]$2[/code]", $s);
64 // Don't show link to full picture (until it is fixed)
65 $s = scale_external_images($s, false);
71 function stripdcode_br_cb($s) {
72 return '[code]' . str_replace('<br />', "\n\t", $s[1]) . '[/code]';
76 //////////////////////
77 // The following "diaspora_ul" and "diaspora_ol" are only appropriate for the
78 // pre-Markdownify conversion. If Markdownify isn't used, use the non-Markdownify
80 //////////////////////
82 function diaspora_ul($s) {
83 // Replace "[*]" followed by any number (including zero) of
84 // spaces by "* " to match Diaspora's list format
85 if( strpos($s[0], "[list]") === 0 )
86 return '<ul class="listbullet" style="list-style-type: circle;">' . preg_replace("/\[\*\]( *)/", "* ", $s[1]) . '</ul>';
87 elseif( strpos($s[0], "[ul]") === 0 )
88 return '<ul class="listbullet" style="list-style-type: circle;">' . preg_replace("/\[\*\]( *)/", "* ", $s[1]) . '</ul>';
94 function diaspora_ol($s) {
95 // A hack: Diaspora will create a properly-numbered ordered list even
96 // if you use '1.' for each element of the list, like:
100 if( strpos($s[0], "[list=1]") === 0 )
101 return '<ul class="listdecimal" style="list-style-type: decimal;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
102 elseif( strpos($s[0], "[list=i]") === 0 )
103 return '<ul class="listlowerroman" style="list-style-type: lower-roman;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
104 elseif( strpos($s[0], "[list=I]") === 0 )
105 return '<ul class="listupperroman" style="list-style-type: upper-roman;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
106 elseif( strpos($s[0], "[list=a]") === 0 )
107 return '<ul class="listloweralpha" style="list-style-type: lower-alpha;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
108 elseif( strpos($s[0], "[list=A]") === 0 )
109 return '<ul class="listupperalpha" style="list-style-type: upper-alpha;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
110 elseif( strpos($s[0], "[ol]") === 0 )
111 return '<ul class="listdecimal" style="list-style-type: decimal;">' . preg_replace("/\[\*\]( *)/", "1. ", $s[1]) . '</ul>';
117 //////////////////////
118 // Non-Markdownify versions of "diaspora_ol" and "diaspora_ul"
119 //////////////////////
120 function diaspora_ul($s) {
121 // Replace "[\\*]" followed by any number (including zero) of
122 // spaces by "* " to match Diaspora's list format
123 return preg_replace("/\[\\\\\*\]( *)/", "* ", $s[1]);
126 function diaspora_ol($s) {
127 // A hack: Diaspora will create a properly-numbered ordered list even
128 // if you use '1.' for each element of the list, like:
132 return preg_replace("/\[\\\\\*\]( *)/", "1. ", $s[1]);
136 function bb2diaspora($Text,$preserve_nl = false, $fordiaspora = true) {
138 // Since Diaspora is creating a summary for links, this function removes them before posting
140 $Text = bb_remove_share_information($Text);
142 // Re-enabling the converter again.
143 // The bbcode parser now handles youtube-links (and the other stuff) correctly.
144 // Additionally the html code is now fixed so that lists are now working.
147 * Transform #tags, strip off the [url] and replace spaces with underscore
149 $Text = preg_replace_callback('/#\[url\=(\w+.*?)\](\w+.*?)\[\/url\]/i', create_function('$match',
150 'return \'#\'. str_replace(\' \', \'_\', $match[2]);'
154 // Converting images with size parameters to simple images. Markdown doesn't know it.
155 $Text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/ism", '[img]$3[/img]', $Text);
157 // the following was added on 10-January-2012 due to an inability of Diaspora's
158 // new javascript markdown processor to handle links with images as the link "text"
159 // It is not optimal and may be removed if this ability is restored in the future
161 // $Text = preg_replace("/\[url\=([^\[\]]*)\]\s*\[img\](.*?)\[\/img\]\s*\[\/url\]/ism",
162 // "[url]$1[/url]\n[img]$2[/img]", $Text);
164 // Remove the avatar picture since that isn't looking good on the other side
165 //$Text = preg_replace("/\[share(.*?)avatar\s?=\s?'.*?'\s?(.*?)\]\s?(.*?)\s?\[\/share\]\s?/ism","\n[share$1$2]$3[/share]",$Text);
167 // Convert it to HTML - don't try oembed
169 $Text = bbcode($Text, $preserve_nl, false, 3);
171 $Text = bbcode($Text, $preserve_nl, false, 4);
173 // Now convert HTML to Markdown
174 $md = new Markdownify(false, false, false);
175 $Text = $md->parseString($Text);
177 // The Markdownify converter converts underscores '_' in URLs to '\_', which
178 // messes up the URL. Manually fix these
180 $pos = bb_find_open_close($Text, '[', ']', $count);
181 while($pos !== false) {
182 $start = substr($Text, 0, $pos['start']);
183 $subject = substr($Text, $pos['start'], $pos['end'] - $pos['start'] + 1);
184 $end = substr($Text, $pos['end'] + 1);
186 $subject = str_replace('\_', '_', $subject);
187 $Text = $start . $subject . $end;
190 $pos = bb_find_open_close($Text, '[', ']', $count);
193 // If the text going into bbcode() has a plain URL in it, i.e.
194 // with no [url] tags around it, it will come out of parseString()
195 // looking like: <http://url.com>, which gets removed by strip_tags().
196 // So take off the angle brackets of any such URL
197 $Text = preg_replace("/<http(.*?)>/is", "http$1", $Text);
199 // Remove all unconverted tags
200 $Text = strip_tags($Text);
205 $ev = bbtoevent($Text);
207 // Replace any html brackets with HTML Entities to prevent executing HTML or script
208 // Don't use strip_tags here because it breaks [url] search by replacing & with amp
210 $Text = str_replace("<", "<", $Text);
211 $Text = str_replace(">", ">", $Text);
213 // If we find any event code, turn it into an event.
214 // After we're finished processing the bbcode we'll
215 // replace all of the event code with a reformatted version.
218 $Text = str_replace(array("\n","\r"), array('',''),$Text);
220 // Remove the "return" character, as Diaspora uses only the "newline"
221 // character, so having the "return" character can cause signature
223 $Text = str_replace("\r", "", $Text);
226 // Set up the parameters for a URL search string
227 $URLSearchString = "^\[\]";
228 // Set up the parameters for a MAIL search string
229 $MAILSearchString = $URLSearchString;
231 // Perform URL Search
233 // [img]pathtoimage[/img]
235 // the following was added on 10-January-2012 due to an inability of Diaspora's
236 // new javascript markdown processor to handle links with images as the link "text"
237 // It is not optimal and may be removed if this ability is restored in the future
239 $Text = preg_replace("/\[url\=([$URLSearchString]*)\]\[img\](.*?)\[\/img\]\[\/url\]/ism",
240 '![' . t('image/photo') . '](' . '$2' . ')' . "\n" . '[' . t('link') . '](' . '$1' . ')', $Text);
242 $Text = preg_replace("/\[bookmark\]([$URLSearchString]*)\[\/bookmark\]/ism", '[$1]($1)', $Text);
243 $Text = preg_replace("/\[bookmark\=([$URLSearchString]*)\](.*?)\[\/bookmark\]/ism", '[$2]($1)', $Text);
245 $Text = preg_replace("/\[url\]([$URLSearchString]*)\[\/url\]/ism", '[$1]($1)', $Text);
246 $Text = preg_replace("/\#\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[#$2]($1)', $Text);
247 $Text = preg_replace("/\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", '[$2]($1)', $Text);
250 $Text = preg_replace("/\[img\](.*?)\[\/img\]/", '![' . t('image/photo') . '](' . '$1' . ')', $Text);
251 $Text = preg_replace("/\[img\=(.*?)\](.*?)\[\/img\]/", '![' . t('image/photo') . '](' . '$2' . ')', $Text);
253 // Perform MAIL Search
254 $Text = preg_replace("(\[mail\]([$MAILSearchString]*)\[/mail\])", '[$1](mailto:$1)', $Text);
255 $Text = preg_replace("/\[mail\=([$MAILSearchString]*)\](.*?)\[\/mail\]/", '[$2](mailto:$1)', $Text);
257 $Text = str_replace('*', '\\*', $Text);
258 $Text = str_replace('_', '\\_', $Text);
260 $Text = str_replace('`','\\`', $Text);
262 // Check for bold text
263 $Text = preg_replace("(\[b\](.*?)\[\/b\])is",'**$1**',$Text);
265 // Check for italics text
266 $Text = preg_replace("(\[i\](.*?)\[\/i\])is",'_$1_',$Text);
268 // Check for underline text
269 // Replace with italics since Diaspora doesn't have underline
270 $Text = preg_replace("(\[u\](.*?)\[\/u\])is",'_$1_',$Text);
272 // Check for strike-through text
273 $Text = preg_replace("(\[s\](.*?)\[\/s\])is",'**[strike]**$1**[/strike]**',$Text);
275 // Check for over-line text
276 // $Text = preg_replace("(\[o\](.*?)\[\/o\])is",'<span class="overline">$1</span>',$Text);
278 // Check for colored text
279 // Remove color since Diaspora doesn't support it
280 $Text = preg_replace("(\[color=(.*?)\](.*?)\[\/color\])is","$2",$Text);
282 // Check for sized text
283 // Remove it since Diaspora doesn't support sizes very well
284 $Text = preg_replace("(\[size=(.*?)\](.*?)\[\/size\])is","$2",$Text);
286 // Check for list text
288 while ((((strpos($Text, "[/list]") !== false) && (strpos($Text, "[list") !== false)) ||
289 ((strpos($Text, "[/ol]") !== false) && (strpos($Text, "[ol]") !== false)) ||
290 ((strpos($Text, "[/ul]") !== false) && (strpos($Text, "[ul]") !== false)) ||
291 ((strpos($Text, "[/li]") !== false) && (strpos($Text, "[li]") !== false))) && (++$endlessloop < 20)) {
292 $Text = preg_replace_callback("/\[list\](.*?)\[\/list\]/is", 'diaspora_ul', $Text);
293 $Text = preg_replace_callback("/\[list=1\](.*?)\[\/list\]/is", 'diaspora_ol', $Text);
294 $Text = preg_replace_callback("/\[list=i\](.*?)\[\/list\]/s",'diaspora_ol', $Text);
295 $Text = preg_replace_callback("/\[list=I\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
296 $Text = preg_replace_callback("/\[list=a\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
297 $Text = preg_replace_callback("/\[list=A\](.*?)\[\/list\]/s", 'diaspora_ol', $Text);
298 $Text = preg_replace_callback("/\[ul\](.*?)\[\/ul\]/is", 'diaspora_ul', $Text);
299 $Text = preg_replace_callback("/\[ol\](.*?)\[\/ol\]/is", 'diaspora_ol', $Text);
300 $Text = preg_replace("/\[li\]( *)(.*?)\[\/li\]/s", '* $2' ,$Text);
303 // Just get rid of table tags since Diaspora doesn't support tables
304 $Text = preg_replace("/\[th\](.*?)\[\/th\]/s", '$1' ,$Text);
305 $Text = preg_replace("/\[td\](.*?)\[\/td\]/s", '$1' ,$Text);
306 $Text = preg_replace("/\[tr\](.*?)\[\/tr\]/s", '$1' ,$Text);
307 $Text = preg_replace("/\[table\](.*?)\[\/table\]/s", '$1' ,$Text);
309 $Text = preg_replace("/\[table border=(.*?)\](.*?)\[\/table\]/s", '$2' ,$Text);
310 // $Text = preg_replace("/\[table border=0\](.*?)\[\/table\]/s", '<table border="0" >$1</table>' ,$Text);
313 // $Text = str_replace("[*]", "<li>", $Text);
315 // Check for font change text
316 // $Text = preg_replace("(\[font=(.*?)\](.*?)\[\/font\])","<span style=\"font-family: $1;\">$2</span>",$Text);
319 $Text = preg_replace_callback("/\[code\](.*?)\[\/code\]/is",'stripdcode_br_cb',$Text);
321 // Check for [code] text
322 $Text = preg_replace("/(\[code\])+(.*?)(\[\/code\])+/is","\t$2\n", $Text);
327 // Declare the format for [quote] layout
328 // $QuoteLayout = '<blockquote>$1</blockquote>';
329 // Check for [quote] text
330 $Text = preg_replace("/\[quote\](.*?)\[\/quote\]/is",">$1\n\n", $Text);
331 $Text = preg_replace("/\[quote=(.*?)\](.*?)\[\/quote\]/is",">$2\n\n", $Text);
335 // html5 video and audio
337 $Text = preg_replace("/\[video\](.*?)\[\/video\]/", '$1', $Text);
339 $Text = preg_replace("/\[audio\](.*?)\[\/audio\]/", '$1', $Text);
341 // $Text = preg_replace("/\[iframe\](.*?)\[\/iframe\]/", '<iframe src="$1" width="425" height="350"><a href="$1">$1</a></iframe>', $Text);
343 // [img=widthxheight]image source[/img]
344 // $Text = preg_replace("/\[img\=([0-9]*)x([0-9]*)\](.*?)\[\/img\]/", '<img src="$3" style="height:{$2}px; width:{$1}px;" >', $Text);
346 $Text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/watch\?v\=(.*?)\[\/youtube\]/ism",'http://www.youtube.com/watch?v=$1',$Text);
347 $Text = preg_replace("/\[youtube\]https?:\/\/www.youtube.com\/embed\/(.*?)\[\/youtube\]/ism",'http://www.youtube.com/watch?v=$1',$Text);
348 $Text = preg_replace("/\[youtube\]https?:\/\/youtu.be\/(.*?)\[\/youtube\]/ism",'http://www.youtube.com/watch?v=$1',$Text);
349 $Text = preg_replace("/\[youtube\]([A-Za-z0-9\-_=]+)(.*?)\[\/youtube\]/ism", 'http://www.youtube.com/watch?v=$1', $Text);
351 $Text = preg_replace("/\[vimeo\]https?:\/\/player.vimeo.com\/video\/([0-9]+)(.*?)\[\/vimeo\]/ism",'http://vimeo.com/$1',$Text);
352 $Text = preg_replace("/\[vimeo\]https?:\/\/vimeo.com\/([0-9]+)(.*?)\[\/vimeo\]/ism",'http://vimeo.com/$1',$Text);
353 $Text = preg_replace("/\[vimeo\]([0-9]+)(.*?)\[\/vimeo\]/ism", 'http://vimeo.com/$1',$Text);
356 $Text = str_replace('[nosmile]','',$Text);
359 // $Text = oembed_bbcode2html($Text);
361 // If we found an event earlier, strip out all the event code and replace with a reformatted version.
365 $sub = format_event_diaspora($ev);
367 $Text = preg_replace("/\[event\-summary\](.*?)\[\/event\-summary\]/is",'',$Text);
368 $Text = preg_replace("/\[event\-description\](.*?)\[\/event\-description\]/is",'',$Text);
369 $Text = preg_replace("/\[event\-start\](.*?)\[\/event\-start\]/is",$sub,$Text);
370 $Text = preg_replace("/\[event\-finish\](.*?)\[\/event\-finish\]/is",'',$Text);
371 $Text = preg_replace("/\[event\-location\](.*?)\[\/event\-location\]/is",'',$Text);
372 $Text = preg_replace("/\[event\-adjust\](.*?)\[\/event\-adjust\]/is",'',$Text);
375 $Text = preg_replace("/\<(.*?)(src|href)=(.*?)\&\;(.*?)\>/ism",'<$1$2=$3&$4>',$Text);
377 $Text = preg_replace_callback('/\[(.*?)\]\((.*?)\)/ism','unescape_underscores_in_links',$Text);
381 // Remove any leading or trailing whitespace, as this will mess up
382 // the Diaspora signature verification and cause the item to disappear
385 call_hooks('bb2diaspora',$Text);
390 function unescape_underscores_in_links($m) {
391 $y = str_replace('\\_','_', $m[2]);
392 return('[' . $m[1] . '](' . $y . ')');
395 function format_event_diaspora($ev) {
399 if(! ((is_array($ev)) && count($ev)))
402 $bd_format = t('l F d, Y \@ g:i A') ; // Friday January 18, 2011 @ 8 AM
404 $o = 'Friendica event notification:' . "\n";
406 $o .= '**' . (($ev['summary']) ? bb2diaspora($ev['summary']) : bb2diaspora($ev['desc'])) . '**' . "\n";
408 $o .= t('Starts:') . ' ' . '['
409 . (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
410 $ev['start'] , $bd_format ))
411 : day_translate(datetime_convert('UTC', 'UTC',
412 $ev['start'] , $bd_format)))
413 . '](' . $a->get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['start'])) . ")\n";
415 if(! $ev['nofinish'])
416 $o .= t('Finishes:') . ' ' . '['
417 . (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
418 $ev['finish'] , $bd_format ))
419 : day_translate(datetime_convert('UTC', 'UTC',
420 $ev['finish'] , $bd_format )))
421 . '](' . $a->get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['finish'])) . ")\n";
423 if(strlen($ev['location']))
424 $o .= t('Location:') . bb2diaspora($ev['location'])