friendica/include/bb2diaspora.php

267 lines
8.5 KiB
PHP
Raw Normal View History

<?php
use Friendica\Content\Text\BBCode;
2018-01-14 18:59:08 -05:00
use Friendica\Content\Text\Markdown;
use Friendica\Core\Addon;
use Friendica\Core\L10n;
2017-08-26 02:04:21 -04:00
use Friendica\Core\System;
2017-12-07 09:04:24 -05:00
use Friendica\Model\Contact;
2017-05-07 14:44:30 -04:00
use Friendica\Network\Probe;
use Friendica\Util\DateTimeFormat;
2017-04-06 23:43:55 -04:00
use League\HTMLToMarkdown\HtmlConverter;
2017-05-07 14:40:23 -04:00
require_once 'include/event.php';
require_once 'include/html2bbcode.php';
require_once 'include/bbcode.php';
2011-08-25 19:37:27 -04:00
/**
* @brief Callback function to replace a Diaspora style mention in a mention for Friendica
*
* @param array $match Matching values for the callback
2017-02-05 02:02:20 -05:00
* @return string Replaced mention
*/
function diaspora_mention2bb($match) {
if ($match[2] == '') {
return;
}
$data = Contact::getDetailsByAddr($match[2]);
$name = $match[1];
if ($name == '') {
$name = $data['name'];
}
return '@[url=' . $data['url'] . ']' . $name . '[/url]';
}
/*
* we don't want to support a bbcode specific markdown interpreter
* and the markdown library we have is pretty good, but provides HTML output.
* So we'll use that to convert to HTML, then convert the HTML back to bbcode,
* and then clean up a few Diaspora specific constructs.
*/
2011-08-25 19:37:27 -04:00
function diaspora2bb($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 14:39:02 -05:00
// The parser cannot handle paragraphs correctly
$s = str_replace(['</p>', '<p>', '<p dir="ltr">'], ['<br>', '<br>', '<br>'], $s);
2011-11-30 06:22:01 -05:00
// Escaping the hash tags
$s = preg_replace('/\#([^\s\#])/', '&#35;$1', $s);
2012-02-01 20:09:53 -05:00
2018-01-14 18:59:08 -05:00
$s = Markdown::convert($s);
2012-02-01 20:09:53 -05:00
$regexp = "/@\{(?:([^\}]+?); )?([^\} ]+)\}/";
$s = preg_replace_callback($regexp, 'diaspora_mention2bb', $s);
$s = str_replace('&#35;', '#', $s);
2012-02-01 20:09:53 -05:00
$s = html2bbcode($s);
2012-01-11 17:03:27 -05:00
// 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);
2012-02-19 19:53:22 -05:00
// 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]', $s);
2012-01-11 17:03:27 -05:00
2012-02-19 19:53:22 -05:00
//$s = preg_replace("/([^\]\=]|^)(https?\:\/\/)(vimeo|youtu|www\.youtube|soundcloud)([a-zA-Z0-9\:\/\-\?\&\;\.\=\_\~\#\%\$\!\+\,]+)/ism", '$1[url=$2$3$4]$2$3$4[/url]',$s);
2016-12-05 22:48:02 -05:00
$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);
2011-11-13 22:04:59 -05:00
// 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)
$s = BBCode::scaleExternalImages($s, false);
2011-08-25 19:37:27 -04:00
return $s;
}
2017-01-01 18:18:42 -05:00
/**
* @brief Callback function to replace a Friendica style mention in a mention for Diaspora
*
* @param array $match Matching values for the callback
2017-02-05 02:02:20 -05:00
* @return string Replaced mention
2017-01-01 18:18:42 -05:00
*/
function diaspora_mentions($match) {
$contact = Contact::getDetailsByURL($match[3]);
2017-01-01 18:18:42 -05:00
if (!x($contact, 'addr')) {
2017-01-01 18:18:42 -05:00
$contact = Probe::uri($match[3]);
}
if (!x($contact, 'addr')) {
2017-01-01 18:18:42 -05:00
return $match[0];
}
$mention = '@{' . $match[2] . '; ' . $contact['addr'] . '}';
2017-01-01 18:18:42 -05:00
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 = "^\[\]";
2017-04-06 23:43:55 -04:00
$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("#\[code(?:=([^\]]*))?\](.*?)\[\/code\]#is",
function ($matches) use (&$codeblocks) {
$return = $matches[0];
if (strpos($matches[2], "\n") !== false) {
$return = '#codeblock-' . count($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 08:29:14 -05:00
$tag = html_entity_decode($tag, ENT_QUOTES, 'UTF-8');
2017-04-06 23:43:55 -04:00
if (!strpos(html_entity_decode($Text, ENT_QUOTES, 'UTF-8'), '#' . $tag)) {
$tagline .= '#' . $tag . ' ';
}
2015-11-28 08:29:14 -05:00
}
$Text = $Text." ".$tagline;
}
} else {
$Text = bbcode($Text, $preserve_nl, false, 4);
}
2017-01-01 18:18:42 -05:00
// mask some special HTML chars from conversation to markdown
$Text = str_replace(['&lt;', '&gt;', '&amp;'], ['&_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(["</a><blockquote>"], ["</a><br><blockquote>"], $Text);
$stamp1 = microtime(true);
// Now convert HTML to Markdown
$converter = new HtmlConverter();
$Text = $converter->convert($Text);
2017-01-01 18:18:42 -05:00
// unmask the special chars back to HTML
$Text = str_replace(['&\_lt\_;', '&\_gt\_;', '&\_amp\_;'], ['&lt;', '&gt;', '&amp;'], $Text);
$a->save_timestamp($stamp1, "parser");
2015-02-24 14:39:02 -05:00
// Libertree has a problem with escaped hashtags.
$Text = str_replace(['\#'], ['#'], $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);
2017-01-01 18:18:42 -05:00
if ($fordiaspora) {
$URLSearchString = "^\[\]";
2017-01-02 02:00:44 -05:00
$Text = preg_replace_callback("/([@]\[(.*?)\])\(([$URLSearchString]*?)\)/ism", 'diaspora_mentions', $Text);
2017-01-01 18:18:42 -05:00
}
// 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;
}
, $Text);
Addon::callHooks('bb2diaspora',$Text);
return $Text;
}
2011-09-03 10:13:12 -04:00
function unescape_underscores_in_links($m) {
2017-04-06 23:43:55 -04:00
$y = str_replace('\\_', '_', $m[2]);
return('[' . $m[1] . '](' . $y . ')');
}
2011-09-03 10:13:12 -04:00
function format_event_diaspora($ev) {
if (! ((is_array($ev)) && count($ev))) {
2011-09-03 10:13:12 -04:00
return '';
}
2011-09-03 10:13:12 -04:00
$bd_format = L10n::t('l F d, Y \@ g:i A') ; // Friday January 18, 2011 @ 8 AM
2011-09-03 10:13:12 -04:00
2012-01-25 18:16:43 -05:00
$o = 'Friendica event notification:' . "\n";
2011-09-03 10:13:12 -04:00
$o .= '**' . (($ev['summary']) ? bb2diaspora($ev['summary']) : bb2diaspora($ev['desc'])) . '**' . "\n";
2011-09-03 10:13:12 -04:00
// @todo What. Is. Going. On. With. This. Useless. Ternary. Operator? - mrpetovan
$o .= L10n::t('Starts:') . ' ' . '[' . day_translate(
$ev['adjust'] ? DateTimeFormat::utc($ev['start'], $bd_format) : DateTimeFormat::utc($ev['start'], $bd_format)
)
. '](' . System::baseUrl() . '/localtime/?f=&time=' . urlencode(DateTimeFormat::utc($ev['start'])) . ")\n";
2011-09-03 10:13:12 -04:00
if (! $ev['nofinish']) {
$o .= L10n::t('Finishes:') . ' ' . '[' . day_translate(
$ev['adjust'] ? DateTimeFormat::utc($ev['finish'], $bd_format) : DateTimeFormat::utc($ev['finish'], $bd_format)
)
. '](' . System::baseUrl() . '/localtime/?f=&time=' . urlencode(DateTimeFormat::utc($ev['finish'])) . ")\n";
}
2011-09-03 10:13:12 -04:00
if (strlen($ev['location'])) {
$o .= L10n::t('Location:') . bb2diaspora($ev['location'])
2011-09-03 10:13:12 -04:00
. "\n";
}
2011-09-03 10:13:12 -04:00
$o .= "\n";
return $o;
}