friendica/include/bb2diaspora.php

260 lines
8.3 KiB
PHP
Raw Normal View History

<?php
2017-04-07 05:43:55 +02:00
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";
2011-08-26 01:37:27 +02:00
/**
* @brief Callback function to replace a Diaspora style mention in a mention for Friendica
*
* @param array $match Matching values for the callback
2017-02-05 08:02:20 +01:00
* @return string Replaced mention
*/
function diaspora_mention2bb($match) {
if ($match[2] == '') {
return;
}
$data = get_contact_details_by_addr($match[2]);
$name = $match[1];
if ($name == '') {
$name = $data['name'];
}
return '@[url=' . $data['url'] . ']' . $name . '[/url]';
}
/*
* we don't want to support a bbcode specific markdown interpreter
* and the markdown library we have is pretty good, but provides HTML output.
* So we'll use that to convert to HTML, then convert the HTML back to bbcode,
* and then clean up a few Diaspora specific constructs.
*/
2011-08-26 01:37:27 +02:00
function diaspora2bb($s) {
$s = html_entity_decode($s, ENT_COMPAT, 'UTF-8');
// Handles single newlines
$s = str_replace("\r\n", "\n", $s);
$s = str_replace("\n", " \n", $s);
$s = str_replace("\r", " \n", $s);
// Replace lonely stars in lines not starting with it with literal stars
$s = preg_replace('/^([^\*]+)\*([^\*]*)$/im', '$1\*$2', $s);
2015-02-24 20:39:02 +01:00
// The parser cannot handle paragraphs correctly
$s = str_replace(array('</p>', '<p>', '<p dir="ltr">'), array('<br>', '<br>', '<br>'), $s);
2011-11-30 12:22:01 +01:00
// Escaping the hash tags
$s = preg_replace('/\#([^\s\#])/', '&#35;$1', $s);
2012-02-02 02:09:53 +01:00
$s = Markdown($s);
2012-02-02 02:09:53 +01:00
$regexp = "/@\{(?:([^\}]+?); )?([^\} ]+)\}/";
$s = preg_replace_callback($regexp, 'diaspora_mention2bb', $s);
$s = str_replace('&#35;', '#', $s);
2012-02-02 02:09:53 +01:00
$s = html2bbcode($s);
2012-01-11 23:03:27 +01:00
// 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);
2012-02-20 01:53:22 +01:00
// 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]', $s);
2012-01-11 23:03:27 +01:00
2012-02-20 01:53:22 +01:00
//$s = preg_replace("/([^\]\=]|^)(https?\:\/\/)(vimeo|youtu|www\.youtube|soundcloud)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3$4]$2$3$4[/url]',$s);
2016-12-06 04:48:02 +01:00
$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);
2011-11-14 04:04:59 +01:00
// remove duplicate adjacent code tags
$s = preg_replace('/(\[code\])+(.*?)(\[\/code\])+/ism', '[code]$2[/code]', $s);
// Don't show link to full picture (until it is fixed)
2012-02-25 23:22:51 +01:00
$s = scale_external_images($s, false);
2011-08-26 01:37:27 +02:00
return $s;
}
2017-01-02 00:18:42 +01:00
/**
* @brief Callback function to replace a Friendica style mention in a mention for Diaspora
*
* @param array $match Matching values for the callback
2017-02-05 08:02:20 +01:00
* @return string Replaced mention
2017-01-02 00:18:42 +01:00
*/
function diaspora_mentions($match) {
$contact = get_contact_details_by_url($match[3]);
if (!x($contact, 'addr')) {
2017-01-02 00:18:42 +01:00
$contact = Probe::uri($match[3]);
}
if (!x($contact, 'addr')) {
2017-01-02 00:18:42 +01:00
return $match[0];
}
$mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
2017-01-02 00:18:42 +01:00
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) {
$Text = bb_remove_share_information($Text);
}
/**
* Transform #tags, strip off the [url] and replace spaces with underscore
*/
$URLSearchString = "^\[\]";
2017-04-07 05:43:55 +02:00
$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) {
$Text = bbcode($Text, $preserve_nl, false, 3);
// Add all tags that maybe were removed
if (preg_match_all("/#\[url\=([$URLSearchString]*)\](.*?)\[\/url\]/ism", $OriginalText, $tags)) {
$tagline = "";
foreach ($tags[2] as $tag) {
2015-11-28 14:29:14 +01:00
$tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2017-04-07 05:43:55 +02:00
if (!strpos(html_entity_decode($Text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
$tagline .= '#' . $tag . ' ';
}
2015-11-28 14:29:14 +01:00
}
$Text = $Text." ".$tagline;
}
} else {
$Text = bbcode($Text, $preserve_nl, false, 4);
}
2017-01-02 00:18:42 +01:00
// mask some special HTML chars from conversation to markdown
2017-04-07 05:43:55 +02:00
$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
$converter = new HtmlConverter();
$Text = $converter->convert($Text);
2017-01-02 00:18:42 +01:00
// unmask the special chars back to HTML
2017-04-07 05:43:55 +02:00
$Text = str_replace(array('&_lt_;', '&_gt_;', '&_amp_;'), array('&lt;', '&gt;', '&amp;'), $Text);
$a->save_timestamp($stamp1, "parser");
2015-02-24 20:39:02 +01:00
// 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);
2017-01-02 00:18:42 +01:00
if ($fordiaspora) {
$URLSearchString = "^\[\]";
2017-01-02 08:00:44 +01:00
$Text = preg_replace_callback("/([@]\[(.*?)\])\(([$URLSearchString]*?)\)/ism", 'diaspora_mentions', $Text);
2017-01-02 00:18:42 +01:00
}
// 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 $Text;
}
2011-09-03 16:13:12 +02:00
function unescape_underscores_in_links($m) {
2017-04-07 05:43:55 +02:00
$y = str_replace('\\_', '_', $m[2]);
return('[' . $m[1] . '](' . $y . ')');
}
2011-09-03 16:13:12 +02:00
function format_event_diaspora($ev) {
if (! ((is_array($ev)) && count($ev))) {
2011-09-03 16:13:12 +02:00
return '';
}
2011-09-03 16:13:12 +02:00
$bd_format = t('l F d, Y \@ g:i A') ; // Friday January 18, 2011 @ 8 AM
2012-01-26 00:16:43 +01:00
$o = 'Friendica event notification:' . "\n";
2011-09-03 16:13:12 +02:00
$o .= '**' . (($ev['summary']) ? bb2diaspora($ev['summary']) : bb2diaspora($ev['desc'])) . '**' . "\n";
2011-09-03 16:13:12 +02:00
$o .= t('Starts:') . ' ' . '['
. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
2011-09-03 16:13:12 +02:00
$ev['start'] , $bd_format ))
: day_translate(datetime_convert('UTC', 'UTC',
2011-09-03 16:13:12 +02:00
$ev['start'] , $bd_format)))
. '](' . App::get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['start'])) . ")\n";
2011-09-03 16:13:12 +02:00
if (! $ev['nofinish']) {
$o .= t('Finishes:') . ' ' . '['
. (($ev['adjust']) ? day_translate(datetime_convert('UTC', 'UTC',
2011-09-03 16:13:12 +02:00
$ev['finish'] , $bd_format ))
: day_translate(datetime_convert('UTC', 'UTC',
2011-09-03 16:13:12 +02:00
$ev['finish'] , $bd_format )))
. '](' . App::get_baseurl() . '/localtime/?f=&time=' . urlencode(datetime_convert('UTC','UTC',$ev['finish'])) . ")\n";
}
2011-09-03 16:13:12 +02:00
if (strlen($ev['location'])) {
$o .= t('Location:') . bb2diaspora($ev['location'])
2011-09-03 16:13:12 +02:00
. "\n";
}
2011-09-03 16:13:12 +02:00
$o .= "\n";
return $o;
}