<?php
/**
* Helper to do common string operations, such as word wrap and censoring.
*
* @package XenForo_Helper
*/
class XenForo_Helper_String
{
/**
* Censor cache for default words/censor string option.
*
* @var array|null Null when not set up
*/
protected static $_censorCache = null;
/**
* Internal value of html encoded variable for auto-linking. Only used
* to pass to callback.
*
* @var boolean
*/
protected static $_alptHtmlEncoded = true;
protected static $_htmlReplacementPlaceholders = array();
/**
* Private constructor. Use statically.
*/
private function __construct() {}
/**
* Adds spaces to the given string after the given amount of
* contiguous, non-whitespace characters.
*
* @param string $string
* @param integer $breakLength Number of characters before break; if null, use option
*
* @return string
*/
public static function wordWrapString($string, $breakLength)
{
$breakLength = intval($breakLength);
if ($breakLength < 1 || $breakLength > strlen($string)) // strlen isn't completely accurate, but this is an optimization
{
return $string;
}
return preg_replace('#[^\s]{' . $breakLength . '}(?=[^\s])#u', '$0 ', $string);
}
/**
* Helper to allow templates to pass the censor string to self::censorString()
*
* @param string $string
* @param string $censorString
*
* @return string
*/
public static function censorStringTemplateHelper($string, $censorString = null)
{
return self::censorString($string, null, $censorString);
}
/**
* Censors the given string.
*
* @param string $string
* @param array|null $words Words to censor. Null to use option value.
* @param string|null $censorString String to censor each character with. Null to use option value.
*
* @return string
*/
public static function censorString($string, array $words = null, $censorString = null)
{
$allowCache = ($words === null && $censorString === null); // ok to use cache for default
$censorCache = ($allowCache ? self::$_censorCache : null);
if ($censorCache === null)
{
if ($words === null)
{
$words = XenForo_Application::get('options')->censorWords;
}
if (!$words)
{
if ($allowCache)
{
self::$_censorCache = array();
}
return $string;
}
if ($censorString === null)
{
$censorString = XenForo_Application::get('options')->censorCharacter;
}
$censorCache = self::buildCensorArray($words, $censorString);
if ($allowCache)
{
self::$_censorCache = $censorCache;
}
}
if ($censorCache)
{
$string = preg_replace(array_keys($censorCache), $censorCache, $string);
}
return $string;
}
public static function stripCensoredText($string, array $words = null)
{
if ($words === null)
{
$words = XenForo_Application::get('options')->censorWords;
}
foreach ($words AS $key => &$word)
{
if (is_string($key) || !isset($word['word']) || !isset($word['replace']))
{
// old format or broken
continue;
}
$word['replace'] = utf8_strlen($word['word']);
}
return self::censorString($string, $words, '');
}
/**
* Builds the censorship array.
*
* @param array $words List of words (from option format)
* @param string $censorString String to replace each character with if no replacement map
*
* @return array Possible keys: exact, any with key-value search/replace pairs
*/
public static function buildCensorArray(array $words, $censorString)
{
$censorCache = array();
foreach ($words AS $key => $word)
{
if (is_string($key) || !isset($word['regex']) || !isset($word['replace']))
{
// old format or broken
continue;
}
$regex = $word['regex'];
$replace = $word['replace'];
$censorCache[$regex] = is_int($replace) ? str_repeat($censorString, $replace) : $replace;
}
return $censorCache;
}
/**
* Returns a string that is snipped to $maxLength characters at the nearest space,
* Appends an elipsis if the string is snipped.
*
* @param string
* @param integer Max length of returned string, excluding elipsis
* @param integer Offset from string start - will add leading elipsis
* @param string Elipses string (default: '...')
*
* @return string
*/
public static function wholeWordTrim($string, $maxLength, $offset = 0, $elipses = '...')
{
//TODO: this may need a handler for language independence and some form of error correction for bbcode
$elipsesLen = strlen($elipses);
if ($offset)
{
$string = preg_replace('/^\S*\s+/s', '', utf8_substr($string, $offset));
if ($maxLength > 0)
{
$maxLength = max(1, $maxLength - $elipsesLen);
}
}
$strLength = utf8_strlen($string);
if ($maxLength > 0 && $strLength > $maxLength)
{
$maxLength -= $elipsesLen;
if ($maxLength > 0)
{
$string = utf8_substr($string, 0, $maxLength);
$string = strrev(preg_replace('/^\S*\s+/s', '', strrev($string))) . $elipses;
}
else if ($maxLength <= 0)
{
// too short with the elipses, can't really display anything
$string = $elipses;
$offset = 0;
}
}
if ($offset)
{
$string = $elipses . $string;
}
return $string;
}
public static function moveHtmlToPlaceholders($string)
{
self::$_htmlReplacementPlaceholders = array();
return preg_replace_callback(
'#<[^>]*>#si',
array('XenForo_Helper_String', 'htmlPlaceHolderHandlerInternal'),
$string
);
}
public static function htmlPlaceHolderHandlerInternal(array $match)
{
$placeholder = "\x1A" . count(self::$_htmlReplacementPlaceholders) . "\x1A";
self::$_htmlReplacementPlaceholders[$placeholder] = $match[0];
return $placeholder;
}
public static function restoreHtmlPlaceholders($string)
{
$result = strtr($string, self::$_htmlReplacementPlaceholders);
self::$_htmlReplacementPlaceholders = array();
return $result;
}
public static function removeHtmlPlaceholders($string)
{
return preg_replace("#\x1A\\d+\x1A#", '', $string);
}
public static function linkTaggedPlainText($string)
{
$string = self::moveHtmlToPlaceholders($string);
$string = preg_replace_callback(
'#@\[(\d+):(\'|"|"|)(.*)\\2\]#iU',
array('self', '_linkTaggedPlainTextCallback'),
$string
);
$string = self::restoreHtmlPlaceholders($string);
return $string;
}
protected static function _linkTaggedPlainTextCallback(array $match)
{
$userId = intval($match[1]);
$username = self::removeHtmlPlaceholders($match[3]);
$username = htmlspecialchars($username, ENT_QUOTES, 'utf-8', false);
$link = XenForo_Link::buildPublicLink('full:members', array('user_id' => $userId));
return '<a href="' . htmlspecialchars($link) . '" class="username" data-user="' . $userId . ', ' . $username . '">'
. $username . '</a>';
}
/**
* Automatically links URLs/email in the given string of BB code. Output will be original
* BB code input with [url] or [email] tags inserted where ncessary.
*
* @param string $string BB code string
* @param boolean $autoEmbedIfEnabled Do auto embed of media if enabled in options
*
* @return string
*/
public static function autoLinkBbCode($string, $autoEmbedIfEnabled = true)
{
$formatter = XenForo_BbCode_Formatter_Base::create('XenForo_BbCode_Formatter_BbCode_AutoLink', false);
$formatter->setAutoEmbed($autoEmbedIfEnabled);
$parser = XenForo_BbCode_Parser::create($formatter);
return $parser->render($string);
}
/**
* Auto-links URLs in plain text. This text should generally already be HTML
* escaped, because it can't be done after the linking.
*
* @param string $string
* @param boolean $htmlEncoded Denotes whether the text is already encoded; if false, the URL will be encoded before being put into the link
*
* @return string Text with links added
*/
public static function autoLinkPlainText($string, $htmlEncoded = true)
{
self::$_alptHtmlEncoded = $htmlEncoded;
$string = self::moveHtmlToPlaceholders($string);
$string = preg_replace_callback(
'#(?<=[^a-z0-9@-]|^)(https?://|www\.)[^\s"<>{}`]+#i',
array('self', '_autoLinkPlainTextCallback'),
$string
);
$string = self::restoreHtmlPlaceholders($string);
return $string;
}
/**
* Internal handler for auto-linking regex.
*
* @param array $match
*
* @return string
*/
protected static function _autoLinkPlainTextCallback(array $match)
{
$url = self::removeHtmlPlaceholders($match[0]);
$link = self::prepareAutoLinkedUrl($url);
if (!self::$_alptHtmlEncoded)
{
$link['url'] = htmlspecialchars($link['url']);
}
list($class, $target) = self::getLinkClassTarget($link['url']);
$class = $class ? " class=\"$class\"" : '';
$target = $target ? " target=\"$target\"" : '';
return '<a href="' . $link['url'] . "\" rel=\"nofollow\"$class$target>" . $link['linkText'] . '</a>' . $link['suffixText'];
}
/**
* Gets the class and target to apply to a specified link URL.
* @param string $url
*
* @return array [class, target, type (internal/external), scheme match]
*/
public static function getLinkClassTarget($url)
{
$target = '_blank';
$class = 'externalLink';
$type = 'external';
$schemeMatch = true;
$urlInfo = @parse_url($url);
if ($urlInfo)
{
if (empty($urlInfo['host']))
{
$isInternal = true;
}
else
{
$host = $urlInfo['host'] . (!empty($urlInfo['port']) ? ":$urlInfo[port]" : '');
$isInternal = ($host == XenForo_Application::$host);
$scheme = (!empty($urlInfo['scheme']) ? strtolower($urlInfo['scheme']) : 'http');
$schemeMatch = $scheme == (XenForo_Application::$secure ? 'https' : 'http');
}
if ($isInternal)
{
$target = '';
$class = 'internalLink';
$type = 'internal';
}
}
return array($class, $target, $type, $schemeMatch);
}
/**
* Given a text that appears to be a URL, extracts the components from it,
* possibly moving characters after the link or adding http://.
*
* @param string $url URL that may have trailing characters or missing scheme
*
* @return array Keys: url, linkText, suffixText
*/
public static function prepareAutoLinkedUrl($url)
{
$suffixText = '';
if (preg_match('/&(?:quot|gt|lt);/i', $url, $match, PREG_OFFSET_CAPTURE))
{
$suffixText = substr($url, $match[0][1]);
$url = substr($url, 0, $match[0][1]);
}
$linkText = $url;
if (strpos($url, '://') === false)
{
$url = 'http://' . $url;
}
do
{
$matchedTrailer = false;
$lastChar = substr($url, -1);
switch ($lastChar)
{
case ')':
case ']':
$closer = $lastChar;
$opener = $lastChar == ']' ? '[' : '(';
if (substr_count($url, $closer) == substr_count($url, $opener))
{
break;
}
// break missing intentionally
case '(':
case '[':
case '.':
case ',':
case '!':
case ':':
case "'":
$suffixText = $lastChar . $suffixText;
$url = substr($url, 0, -1);
$linkText = substr($linkText, 0, -1);
$matchedTrailer = true;
break;
}
}
while ($matchedTrailer);
return array(
'url' => $url,
'linkText' => $linkText,
'suffixText' => $suffixText
);
}
/**
* Strips BB code from a string
*
* @param string $string
* @param boolean $stripQuote If true, contents from within quote tags are stripped
*
* @return string
*/
public static function bbCodeStrip($string, $stripQuote = false)
{
/*$formatter = XenForo_BbCode_Formatter_Base::create('XenForo_BbCode_Formatter_BbCode_Strip', false);
$formatter->setMaxQuoteDepth(0);
$parser = XenForo_BbCode_Parser::create($formatter);
return $parser->render($string);*/
if ($stripQuote)
{
$parts = preg_split('#(\[quote[^\]]*\]|\[/quote\])#i', $string, -1, PREG_SPLIT_DELIM_CAPTURE);
$string = '';
$quoteLevel = 0;
foreach ($parts AS $i => $part)
{
if ($i % 2 == 0)
{
// always text, only include if not inside quotes
if ($quoteLevel == 0)
{
$string .= rtrim($part) . "\n";
}
}
else
{
// quote start/end
if ($part[1] == '/')
{
// close tag, down a level if open
if ($quoteLevel)
{
$quoteLevel--;
}
}
else
{
// up a level
$quoteLevel++;
}
}
}
}
// replaces unviewable tags with a text representation
$string = str_replace('[*]', '', $string);
$string = preg_replace('#\[(attach|media|img|spoiler)(?![a-z0-9_])[^\]]*\].*\[/\\1\]#siU', "[\x1A\\1]", $string);
// split the string into possible delimiters and text; even keys (from 0) are strings, odd are delimiters
$parts = preg_split('#(\[[a-z0-9_]+(?:=[^\]]*)?\]|\[/[a-z0-9_]+\])#si', $string, -1, PREG_SPLIT_DELIM_CAPTURE);
$total = count($parts);
if ($total < 2)
{
$string = str_replace("\x1A", '', $string);
return trim($string);
}
$closes = array();
$skips = array();
$newString = '';
// first pass: find all the closing tags and note their keys
for ($i = 1; $i < $total; $i += 2)
{
if (preg_match("#^\\[/([a-z0-9_]+)]#i", $parts[$i], $match))
{
$closes[strtolower($match[1])][$i] = $i;
}
}
// second pass: look for all the text elements and any opens, then find
// the first corresponding close that comes after it and remove it.
// if we find that, don't display the open or that close
for ($i = 0; $i < $total; $i++)
{
$part = $parts[$i];
if ($i % 2 == 0)
{
// string part
$newString .= $part;
continue;
}
if (!empty($skips[$i]))
{
// known close
continue;
}
if (preg_match('/^\[([a-z0-9_]+)(?:=|\])/i', $part, $match))
{
$tagName = strtolower($match[1]);
if (!empty($closes[$tagName]))
{
do
{
$closeKey = reset($closes[$tagName]);
if ($closeKey)
{
unset($closes[$tagName][$closeKey]);
}
}
while ($closeKey && $closeKey < $i);
if ($closeKey)
{
// found a matching close after this tag
$skips[$closeKey] = true;
continue;
}
}
}
$newString .= $part;
}
$newString = str_replace("\x1A", '', $newString);
return trim($newString);
}
/**
* Creates a snippet of $length maximum characters around a search term.
* If the term is not found, the snippet is taken from the string start.
*
* @param $string
* @param $maxLength
* @param $term
*
* @return string
*/
public static function wholeWordTrimAroundSearchTerm($string, $maxLength, $term)
{
$stringLength = utf8_strlen($string);
if ($stringLength > $maxLength)
{
$term = strval($term);
if ($term !== '')
{
// TODO: slightly more intelligent search term matching, breaking up multiple words etc.
$termPosition = utf8_strpos(utf8_strtolower($string), utf8_strtolower($term));
}
else
{
$termPosition = false;
}
if ($termPosition !== false)
{
// add term length to term start position
$startPos = $termPosition + utf8_strlen($term);
// count back half the max characters
$startPos -= $maxLength / 2;
// don't overflow the beginning
$startPos = max(0, $startPos);
// don't overflow the end
$startPos = min($startPos, $stringLength - $maxLength);
}
else
{
$startPos = 0;
}
$string = self::wholeWordTrim($string, $maxLength, $startPos);
}
return $string;
}
/**
* Returns an HTML string with instances a search term highlighted with <em class="$emClass">...</em>
*
* @param string Haystack
* @param string Needle
* @param string Class with which to style the wrapping <em>
*
* @return string HTML
*/
public static function highlightSearchTerm($string, $term, $emClass = 'highlight')
{
$term = trim(preg_replace('#((^|\s)[+|-]|[/()"~^])#', ' ', strval($term)));
if ($term !== '')
{
return preg_replace('/(' . preg_replace('#\s+#', '|', preg_quote(htmlspecialchars($term), '/')) . ')/siu', '<em class="' . $emClass . '">\1</em>', htmlspecialchars($string));
}
return htmlspecialchars($string);
}
/**
* Strips out quoted text from the specified string, allowing for quoted text
* up to a specified depth.
*
* @param string $string
* @param integer $allowedDepth -1 for unlimited depth
* @param boolean $censorResults
*
* @return string Quotes stripped
*/
public static function stripQuotes($string, $allowedDepth = -1, $censorResults = true)
{
if ($allowedDepth == -1)
{
return $string;
}
else
{
$formatter = XenForo_BbCode_Formatter_Base::create('XenForo_BbCode_Formatter_BbCode_Strip', false);
$formatter->setMaxQuoteDepth($allowedDepth);
$formatter->setCensoring($censorResults);
$parser = XenForo_BbCode_Parser::create($formatter);
return $parser->render($string);
}
}
/**
* Finds simple phrase syntax in a string, e.g. {phrase:phrase_name} and builds an array of phrase names
* e.g. ['{phrase:phrase_name}'] => 'phrase_name'
*
* The resulting array can then be used to query for the phrase_text of the found phrase names.
*
* @param $string
* @return array
*/
public static function findPhraseNamesFromStringSimple($string)
{
preg_match_all(
'#\{phrase:([a-z0-9_]+)\}#iU', $string, $matches, PREG_SET_ORDER
);
$replacements = array();
foreach ($matches AS $match)
{
$replacements[$match[0]] = $match[1];
}
return $replacements;
}
}