½gXb†föÎB#¨ˆêØ)ªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIß4&ðºDÝ.5ú;³AÏ³Þ†:ü}W¹«Û»¨õ¼p%¡©äè”ûÔ7Þ”²ž_”#žÐâ5üLÂ<mÃ‹$ã mèJÑ$Þl“`ÉäHÜ¹¾&ºÏ9Iëc×sàúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]O5&'0…·ÝºÚoaW	I¿<áÁëî7
ÉA—ÆR(ÿ=³Ò£ÃFIÎ§ÅZœª)SÙÊ¯u›N4«ý——0šúw½ÊgÖFœN5¶·Ký]Oš…ÿù‡ü˜~@P_¹eù.~Ì—Õ!¨'©a²1^äˆ·þõ-¯ñbšÅ{cQ_œë=eëÒ‘™¤ô} ò½Ç’ké‹u&T]På`k[Ež½ÃWãÜe´p®AV8úw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]O?sóÚs®\òÀ,óøÿÏuÑÞD'œôrÓ&¤îAt)x†»÷øâ»ÙçÑ"xÒ|W®Ø)°ŽWê.“?|Y:ÍDV‡Ý{@ü¯t8­€Æ‘Ýc°ÇþºT‹k€—´p	ÈTô$9óÈÓU‰ÉÚçò#%w%•6£÷‡Áú¤•þæoÎW¸äW’ãøòæ/R¹¡!)\ª"hâ,N‰ÏYkl3ý$ê3¯³x¾…÷Ó`eAÑ™ÎgUg`ŠÇ¬“%J?KÕ‰XÌY@²TÙ;|½t÷þ’[_·×®æ¢ |Süúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]O”B¦19ï9'¦d©³ii˜É´èýú=h˜8c‡«â1ÒKúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]Oúw½ÊgÖFœN5¶·Ký]OÂÝ×Wá‹–,ƒ¸uR´å	ï\î7âBõTÃÂ“•C•Dv“:¸wumWk¥ÿt<Žc|Y”…]$k‡q´s|`õ{êbž‰QÙ´nëºR”Ycxàsc!^¥C’®ªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIß¸ìHh)+|‡JÈÜÎ&ËOÖÀ©£%,oBüç&°ÐÖ¯QËhÚyêÀõ$kç4òÿóÎÿht]4úw½ÊgÖFœN5¶·Ký]OPXI%gùè¢.ÙËìË±Q	ÌÝ9Õ:¹)63Þ‘qìÀ7bâf€ÂaYË{ä /ÅªÎ\ëS­77üì8pÍú”õ›3`ªh.®Q2Âz/)z¯˜pîÈØmöXL¢”®Šy!èôËÆ5I5)(+ûïSçøúw½ÊgÖFœN5¶·Ký]O7–XÛUÜ—€’EÖ†ßtêªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIßªÛ»bi—s²_‹ØIIß?½õÀ—ËRž«ÏÓc_µ`a¡J™ÁPõ•d7âˆ…u ‚„<©‚õ˜4a Â<’njy,¦cg‚.MK\úXž&¦Ê].|€Ä pÃM£ßg8¤´€ê«€}ÝÊ}4ü~»`	åÑÛ½g8•¶··ÇÃ[’1t0ƒéðÔ9è¢Åœ
wj·â“ñí˜—:ç#ø0PèèðežÏŸ0 5Å¿™ùQ³6Úwkƒù'ç&›3–@w@‹OÛ_Å8"k‘ç‡W¯c¿Ö!™yÙa—1gÿ,+;øâËÆ4+1êR&Ñ‚GEs®M8S$WPö8’”bKùÅp8šØ²jV9P«ÔïîÁg€OaC[ž“2ödKáªŸâµ‹&~Ó›s ª³`VÝbb80í-öÕ;÷øöÓ/	xUüOÌÖžÖ¹Åß€À}¤NEí±ðÊyAM½SmkAã£ÁŽYÉ‹ µŒ	O3ATI/eÅ‡9;3FX¶ö6Æ§ÌÄ| €Ë±ù™@aÍ§“Ø„ˆ>è†“;ÿ·¢­‡t-ñã›ÜzÛÙÅ^K™F»ãƒ^=êp|!vâÀ½Á õ(Øº‚[œ¯}ÓÖnÃ s¨¼çzŒÉï	zó•½½ýµ—ÈTUr¥*°=ô°Q¯å˜ž1µ•Œ%wÌòáGKÁ!þ½6LP„ÜŒ4V{]j¶R•ÉþùóÊ*­½Hl'ÛG{%n.çäö+Ø\ZëÚz¸×5!,ò

VíÐLJ«gìC(ú×Ý»˜¼¶³äQ‚ E˜õ#ÚüŠ&Âû4Ï…uÅÆ·uQ™æëmP9€ŒM„áµàuÎ¾ÞO¯¢¸]ÓÂ5¬«Úã¨Z-D|—DJžpÞJd¸®1¬éGâ™x6¢
¢µÃåÏñþŒÉŽ@?öÀ™¹I>ä8/…46æÂqnÇ(*Ô?­‚{6òw÷Ç%`›ÝayDŽHGòïyë)åbšÌ®Êô›m1! žáðr}Ùqþ|ÞÃÉ^Û,úöM]V	ÄrCbï”¤£èŒ¤ÑˆÐf–>Þ2bf˜-ú¤XµK£{s`Uê<ÛXŒ2xÎÚ=«(ËþÝæÿ›$D³ñ=UšÇó·WN½{ó#Ô@ÿ¶1ð±»Lº3Äweä˜y"CZwMS½FhŽìIñ+sIòçöNHÐ=Po¾
pYÖ‚'‚PõZög„„Ï”/±o‰ 
ÎcÚsíuËH,ï¬,xT¶sQ½¨ú7#­áz”!cr÷‹j¦³Æ‚GøYÚ¡Uî¯wr'Á‘ŠU2— îï•vÄÞÈ¹\¶Ñâ÷ˆÃË~'|—î“â0w‰IJ‹hAî -ÝÑ•ß1\c¿Wß\ˆ®;±•÷º´è‹$5œ £Ð‰í%<"ÚÎ¼;N€:Ž‹¨ÎþjsFe	ÿwËÊc›G~Fàvèr!‰È"‘Œ=H—L\ÙBcÉ>]±KDþ‰ÎV/ÖsŠfê$ýç\‰7‚W`L‹ho¦qu&ÓD†I†ÕÂÅJ2q€öWmA\Êâ‰V)ÏÙÙàKxKš÷2…(p&€DÝÌýù)±©“dâöÕƒ]².›ß÷xµ6žm_ä¤B‹5³=‘®V_!HQO
k	v¶BxG*ÙG­9þDTózÅz‰5-#t¼¶P— xüö„xý“íôÿ@“¼‡¹ IýÅlñ;°!Txyµ iZ~#f†E:Ô8fûŠIýŠ`D ‡GÚ¯¿þçrSß#¶½v–Ø=#v?ï-“g£/µ,ñ&z†ºœ2
Õ½à~8©×N3¨
ÐÔ ÿäþ¯m›ïÃcþ ÉJ}s£/!{³W‹'ÿÝBÚÄÈ³L
yÍ.ŠÖïŸá;›>ÓuñXwÑ°ZÎ-P–"WÂŽ¢c*¯7*·±UÕ-õÑ_<. ½1l#ûo3Ç,I¹¯Ù Œß£™Ð[û$°aSã)~ÕOé©Ü—¿[ªûÈ¾i|W‘6vF6S?‰|íüÔâå˜ž1µ•Œ%wÌòáGKéâ•Ëã8ËCënô•øÒ•¯¼H¥C‹ ¬Þp?±ì
%ër`¼KîP;åJaÉ7jßé?öÈå±pAýc1Ú˜ÙÀäØ€D·¡ÎŸ•r'`'Ò­ÍÆ+ÏvˆSÇ#'Î×Rc_Ô[˜íôÕÜÿm†3z9Ù?³N^áÈèŽ7Â2š£˜~ð¥„4PR€¨üksÈ¢ü‹èŠùð³¨pÝ†´ÆØ6š;?ybÖæ8z”¤£èŒ¤ÑˆÐf–>žh^_œÿR+§íQ¦Näú™g¡«WÆ(—`e_ˆ•uÎ*ý¨}²¯¶)EýEÄ ë…IXK­ã¯Óë›F.g GÛ’s@–5¦kÕÏ^Ù¼¼Xi³_˜‚ëÚ£2TüTŸŽšÚá’Þ(ã$¦ŽÑ¯‹Íú~L¦æ¤àèäç×åMIQ+ñ#–sŠö4/Ž„êbà•aW[%~^ö–Ôü#Îí0wÛõ-ƒSäWžæmL³P¦DÙy8É.#×áÂfCw³lÀ™‘ëª -hÁåC_ìÂ·úLüR_wk1@.cr›hOx¿\uìÂ1¾ÝqÂ³æqú¢ ¶‘–G1-=•¨e)É‚¯ZÌVJ RSÏYúJ=A3ýO%¬V@úS¸äb:]°|h¡˜HòŒ4Òí&9«9ßPKÓ~²tçEÑU+´õâÿƒ3¸E4Jì&vû¡D!Ù±Îò0ç¿,ùz|öHì,t†AÐ/Ü[ZGªÒSÁg. º?âÝ­I|ä®¨Ð¬8Ùïƒ7›¶z‹æš¨ru:ðŠH„ËëøxÝ•êÔàZ»½w?š™4ÒØÀ‘åàä6í»	†™0¹Fl¡·´ÎÍ ‹‹Dßl=è€âc‚5T¿­š”G²Sô)Mà¦œÙ:ilÐ^Çwród?8qùg½”Dû„UQ>ã-SŒ§ïCWÑÌ6ÙgÇü	a|óäg“TqŒ«žLß$KïÑå¿:ìÜL·s>ß°Ÿ%ˆ^d«.lz2Z•Ü§fp¹EÔN9½îôkéUU‚¹%xG-úácÿôˆÅ>ÏnƒˆºÑcKÇœ¢Ó^+lÕ %\]×Pv¦¯ùt«VÓ¨”¤m:–@±6íŽç"·N…Á¡Y|
Ë”Î¬aH í›—a•	ñ0¥”¤£èŒ¤ÑˆÐf–>Žµ§¹SCí*|¤7…(##ÂîJ*d½6×ÕgDó¹á—õ~Ë‹p…¯Ÿ'š¨Jè)x¾9“)S0šM$N@*¦>º0št2¸1U—j¤0{/ùƒ)[Ç¯µ‹Uîz8J!‰*FKtÿCô]àÅ?ÀÌåëq«ûÂeRŸ…ëÁ‹lâ½ž¡Q~{ËÛˆÝg÷¤9?¸Nâ–L·‘²’6¨*;³fy‹(wwÙ½Á©÷OAé«
Øéª(.'öNë’~âAnJ#ÆÃ¯×¹Fch™×e
ë¼Ì4mGÌ½Ëüém<4¨PèÇ´ .\Üèõ$OCI‹µÞãÃÔ ˜éQ‰tÎX’áÑ­Ù9€7YbyÕç'äx‡†Æ,œ#Øã¦ðØš/1PM t (yet)
 *  incorporated all of Wojciech's changes, though I may at some
 *  future time.
 *
 *  *** End of the housecleaning updates. Updated 08/08/07.
 */

/**
 * Converts HTML to formatted plain text
 *
 * @package    Framework
 * @subpackage Utils
 */
class rcube_html2text
{
    const LINKS_NONE = 0;
    const LINKS_END = 1;
    const LINKS_INLINE = 2;
    const LINKS_DEFAULT = self::LINKS_END;

    /**
     * Contains the HTML content to convert.
     *
     * @var string $html
     */
    protected $html;

    /**
     * Contains the converted, formatted text.
     *
     * @var string $text
     */
    protected $text;

    /**
     * Maximum width of the formatted text, in columns.
     *
     * Set this value to 0 (or less) to ignore word wrapping
     * and not constrain text to a fixed-width column.
     *
     * @var int $width
     */
    protected $width = 70;

    /**
     * Target character encoding for output text
     *
     * @var string $charset
     */
    protected $charset = 'UTF-8';

    /**
     * List of preg* regular expression patterns to search for,
     * used in conjunction with $replace.
     *
     * @var array $search
     * @see self::$replace
     */
    protected $search = [
        '/\r/',                                  // Non-legal carriage return
        '/\n*<\/?html>\n*/is',                   // <html>
        '/\n*<head[^>]*>.*?<\/head>\n*/is',      // <head>
        '/\n*<script[^>]*>.*?<\/script>\n*/is',  // <script>
        '/\n*<style[^>]*>.*?<\/style>\n*/is',    // <style>
        '/[\n\t]+/',                             // Newlines and tabs
        '/<p[^>]*>/i',                           // <p>
        '/<\/p>[\s\n\t]*<div[^>]*>/i',           // </p> before <div>
        '/<br[^>]*>[\s\n\t]*<div[^>]*>/i',       // <br> before <div>
        '/<br[^>]*>\s*/i',                       // <br>
        '/<i[^>]*>(.*?)<\/i>/i',                 // <i>
        '/<em[^>]*>(.*?)<\/em>/i',               // <em>
        '/(<ul[^>]*>|<\/ul>)/i',                 // <ul> and </ul>
        '/(<ol[^>]*>|<\/ol>)/i',                 // <ol> and </ol>
        '/<li[^>]*>(.*?)<\/li>/i',               // <li> and </li>
        '/<li[^>]*>/i',                          // <li>
        '/<hr[^>]*>/i',                          // <hr>
        '/<div[^>]*>/i',                         // <div>
        '/(<table[^>]*>|<\/table>)/i',           // <table> and </table>
        '/(<tr[^>]*>|<\/tr>)/i',                 // <tr> and </tr>
        '/<td[^>]*>(.*?)<\/td>/i',               // <td> and </td>
    ];

    /**
     * List of pattern replacements corresponding to patterns searched.
     *
     * @var array $replace
     * @see self::$search
     */
    protected $replace = [
        '',                                     // Non-legal carriage return
        '',                                     // <html>|</html>
        '',                                     // <head>
        '',                                     // <script>
        '',                                     // <style>
        ' ',                                    // Newlines and tabs
        "\n\n",                                 // <p>
        "\n<div>",                              // </p> before <div>
        '<div>',                                // <br> before <div>
        "\n",                                   // <br>
        '_\\1_',                                // <i>
        '_\\1_',                                // <em>
        "\n\n",                                 // <ul> and </ul>
        "\n\n",                                 // <ol> and </ol>
        "\t* \\1\n",                            // <li> and </li>
        "\n\t* ",                               // <li>
        "\n-------------------------\n",        // <hr>
        "<div>\n",                              // <div>
        "\n\n",                                 // <table> and </table>
        "\n",                                   // <tr> and </tr>
        "\t\t\\1\n",                            // <td> and </td>
    ];

    /**
     * List of preg* regular expression patterns to search for,
     * used in conjunction with $ent_replace.
     *
     * @var array $ent_search
     * @see self::$ent_replace
     */
    protected $ent_search = [
        '/&(nbsp|#160);/i',                      // Non-breaking space
        '/&(quot|rdquo|ldquo|#8220|#8221|#147|#148);/i', // Double quotes
        '/&(apos|rsquo|lsquo|#8216|#8217);/i',   // Single quotes
        '/&gt;/i',                               // Greater-than
        '/&lt;/i',                               // Less-than
        '/&(copy|#169);/i',                      // Copyright
        '/&(trade|#8482|#153);/i',               // Trademark
        '/&(reg|#174);/i',                       // Registered
        '/&(mdash|#151|#8212);/i',               // mdash
        '/&(ndash|minus|#8211|#8722);/i',        // ndash
        '/&(bull|#149|#8226);/i',                // Bullet
        '/&(pound|#163);/i',                     // Pound sign
        '/&(euro|#8364);/i',                     // Euro sign
        '/&(amp|#38);/i',                        // Ampersand: see _converter()
        '/[ ]{2,}/',                             // Runs of spaces, post-handling
    ];

    /**
     * List of pattern replacements corresponding to patterns searched.
     *
     * @var array $ent_replace
     * @see self::$ent_search
     */
    protected $ent_replace = [
        "\xC2\xA0",                             // Non-breaking space
        '"',                                    // Double quotes
        "'",                                    // Single quotes
        '>',
        '<',
        '(c)',
        '(tm)',
        '(R)',
        '--',
        '-',
        '*',
        'Â£',
        'EUR',                                  // Euro sign. â‚¬
        '|+|amp|+|',                            // Ampersand: see _converter()
        ' ',                                    // Runs of spaces, post-handling
    ];

    /**
     * List of preg* regular expression patterns to search for
     * and replace using callback function.
     *
     * @var array $callback_search
     */
    protected $callback_search = [
        '/<(a) [^>]*href=("|\')([^"\']+)\2[^>]*>(.*?)<\/a>/i', // <a href="">
        '/<(h)[123456]( [^>]*)?>(.*?)<\/h[123456]>/i',         // h1 - h6
        '/<(th)( [^>]*)?>(.*?)<\/th>/i',                       // <th> and </th>
    ];

    /**
     * List of preg* regular expression patterns to search for in PRE body,
     * used in conjunction with $pre_replace.
     *
     * @var array $pre_search
     * @see self::$pre_replace
     */
    protected $pre_search = [
        "/\n/",
        "/\t/",
        '/ /',
        '/<pre[^>]*>/',
        '/<\/pre>/'
    ];

    /**
     * List of pattern replacements corresponding to patterns searched for PRE body.
     *
     * @var array $pre_replace
     * @see self::$pre_search
     */
    protected $pre_replace = [
        '<br>',
        '&nbsp;&nbsp;&nbsp;&nbsp;',
        '&nbsp;',
        '',
        ''
    ];

    /**
     * Temp. PRE content
     *
     * @var string $pre_content
     */
    protected $pre_content = '';

    /**
     * Contains a list of HTML tags to allow in the resulting text.
     *
     * @var string $allowed_tags
     * @see self::set_allowed_tags()
     */
    protected $allowed_tags = '';

    /**
     * Contains the base URL that relative links should resolve to.
     *
     * @var string $url
     */
    protected $url;

    /**
     * Indicates whether content in the $html variable has been converted yet.
     *
     * @var bool $_converted
     * @see self::$html
     * @see self::$text
     */
    protected $_converted = false;

    /**
     * Contains URL addresses from links to be rendered in plain text.
     *
     * @var array $_link_list
     * @see self::_build_link_list()
     */
    protected $_link_list = [];

    /**
     * Links handling.
     * - 0 if links should be removed
     * - 1 if a table of link URLs should be listed after the text
     * - 2 if the link should be displayed to the original point in the text they appeared
     *
     * @var int $_links_mode
     */
    protected $_links_mode = 1;

    /**
     * Constructor.
     *
     * If the HTML source string (or file) is supplied, the class
     * will instantiate with that source propagated, all that has
     * to be done it to call get_text().
     *
     * @param string   $source     HTML content
     * @param bool     $from_file  Indicates $source is a file to pull content from
     * @param bool|int $links_mode Links handling mode
     * @param int      $width      Maximum width of the formatted text, 0 for no limit
     */
    function __construct($source = '', $from_file = false, $links_mode = self::LINKS_DEFAULT, $width = 75, $charset = 'UTF-8')
    {
        if (!empty($source)) {
            $this->set_html($source, $from_file);
        }

        $this->set_base_url();
        $this->set_links_mode($links_mode);

        $this->width   = $width;
        $this->charset = $charset;
    }

    /**
     * Sets the links behavior mode
     *
     * @param bool|int $mode
     */
    private function set_links_mode($mode)
    {
        $allowed = [
            self::LINKS_NONE,
            self::LINKS_END,
            self::LINKS_INLINE
        ];

        if (!in_array((int) $mode, $allowed)) {
            $this->_links_mode = self::LINKS_DEFAULT;
            return;
        }

        $this->_links_mode = (int) $mode;
    }

    /**
     * Loads source HTML into memory, either from $source string or a file.
     *
     * @param string $source    HTML content
     * @param bool   $from_file Indicates $source is a file to pull content from
     */
    function set_html($source, $from_file = false)
    {
        if ($from_file && file_exists($source)) {
            $this->html = file_get_contents($source);
        }
        else {
            $this->html = $source;
        }

        $this->_converted = false;
    }

    /**
     * Returns the text, converted from HTML.
     *
     * @return string Plain text
     */
    function get_text()
    {
        if (!$this->_converted) {
            $this->_convert();
        }

        return $this->text;
    }

    /**
     * Prints the text, converted from HTML.
     */
    function print_text()
    {
        print $this->get_text();
    }

    /**
     * Sets the allowed HTML tags to pass through to the resulting text.
     *
     * Tags should be in the form "<p>", with no corresponding closing tag.
     */
    function set_allowed_tags($allowed_tags = '')
    {
        if (!empty($allowed_tags)) {
            $this->allowed_tags = $allowed_tags;
        }
    }

    /**
     * Sets a base URL to handle relative links.
     */
    function set_base_url($url = '')
    {
        if (empty($url)) {
            if (!empty($_SERVER['HTTP_HOST'])) {
                $this->url = 'http://' . $_SERVER['HTTP_HOST'];
            }
            else {
                $this->url = '';
            }
        }
        else {
            // Strip any trailing slashes for consistency (relative
            // URLs may already start with a slash like "/file.html")
            if (substr($url, -1) == '/') {
                $url = substr($url, 0, -1);
            }
            $this->url = $url;
        }
    }

    /**
     * Workhorse function that does actual conversion (calls _converter() method).
     */
    protected function _convert()
    {
        // Variables used for building the link list
        $this->_link_list = [];

        $text = $this->html;

        // Convert HTML to TXT
        $this->_converter($text);

        // Add link list
        if (!empty($this->_link_list)) {
            $text .= "\n\nLinks:\n------\n";
            foreach ($this->_link_list as $idx => $url) {
                $text .= '[' . ($idx+1) . '] ' . $url . "\n";
            }
        }

        $this->text       = $text;
        $this->_converted = true;
    }

    /**
     * Workhorse function that does actual conversion.
     *
     * First performs custom tag replacement specified by $search and
     * $replace arrays. Then strips any remaining HTML tags, reduces whitespace
     * and newlines to a readable format, and word wraps the text to
     * $width characters.
     *
     * @param string &$text Reference to HTML content string
     */
    protected function _converter(&$text)
    {
        // Convert <BLOCKQUOTE> (before PRE!)
        $this->_convert_blockquotes($text);

        // Convert <PRE>
        $this->_convert_pre($text);

        // Remove body tag and anything before
        // We used to have '/^.*<body[^>]*>\n*/is' in $this->search, but this requires
        // high pcre.backtrack_limit setting when converting long HTML strings (#8137)
        if (($pos = stripos($text, '<body')) !== false) {
            $pos = strpos($text, '>', $pos);
            $text = substr($text, $pos + 1);
            $text = ltrim($text);
        }

        // Run our defined tags search-and-replace
        $text = preg_replace($this->search, $this->replace, $text);

        // Run our defined tags search-and-replace with callback
        $text = preg_replace_callback($this->callback_search, [$this, 'tags_preg_callback'], $text);

        // Strip any other HTML tags
        $text = strip_tags($text, $this->allowed_tags);

        // Run our defined entities/characters search-and-replace
        $text = preg_replace($this->ent_search, $this->ent_replace, $text);

        // Replace known html entities
        $text = html_entity_decode($text, ENT_QUOTES, $this->charset);

        // Replace unicode nbsp to regular spaces
        $text = preg_replace('/\xC2\xA0/', ' ', $text);

        // Remove unknown/unhandled entities (this cannot be done in search-and-replace block)
        $text = preg_replace('/&([a-zA-Z0-9]{2,6}|#[0-9]{2,4});/', '', $text);

        // Convert "|+|amp|+|" into "&", need to be done after handling of unknown entities
        // This properly handles situation of "&amp;quot;" in input string
        $text = str_replace('|+|amp|+|', '&', $text);

        // Bring down number of empty lines to 2 max
        $text = preg_replace("/\n\s+\n/", "\n\n", $text);
        $text = preg_replace("/[\n]{3,}/", "\n\n", $text);

        // remove leading empty lines (can be produced by e.g. P tag on the beginning)
        $text = ltrim($text, "\n");

        // Wrap the text to a readable format
        // for PHP versions >= 4.0.2. Default width is 75
        // If width is 0 or less, don't wrap the text.
        if ($this->width > 0) {
            $text = wordwrap($text, $this->width);
        }
    }

    /**
     * Helper function called by preg_replace() on link replacement.
     *
     * Maintains an internal list of links to be displayed at the end of the
     * text, with numeric indices or simply the link to the original point in the text they
     * appeared. Also makes an effort at identifying and handling absolute
     * and relative links.
     *
     * @param string $link    URL of the link
     * @param string $display Part of the text to associate number with
     */
    protected function _handle_link($link, $display)
    {
        if (empty($link)) {
            return $display;
        }

        // Ignored link types
        if (preg_match('!^(javascript:|mailto:|#)!i', $link)) {
            return $display;
        }

        // skip links with href == content (#1490434)
        if ($link === $display) {
            return $display;
        }

        if (preg_match('!^([a-z][a-z0-9.+-]+:)!i', $link)) {
            $url = $link;
        }
        else {
            $url = $this->url;
            if (substr($link, 0, 1) != '/') {
                $url .= '/';
            }
            $url .= "$link";
        }

        if (self::LINKS_NONE === $this->_links_mode) {
            // When not using link list use URL if there's no content (#5795)
            // The content here is HTML, convert it to text first
            $h2t     = new rcube_html2text($display, false, false, 1024, $this->charset);
            $display = $h2t->get_text();

            if (empty($display) && preg_match('!^([a-z][a-z0-9.+-]+://)!i', $link)) {
                return $link;
            }

            return $display;
        }

        if (self::LINKS_INLINE === $this->_links_mode) {
            return $this->_build_link_inline($url, $display);
        }

        return $this->_build_link_list($url, $display);
    }

    /**
     * Helper function called by _handle_link() on link replacement.
     *
     * Displays the link next to the original point in the text they
     * appeared.
     *
     * @param string $url     URL of the link
     * @param string $display linktext
     */
    protected function _build_link_inline($url, $display)
    {
        return $display . ' &lt;' . $url . '&gt;';
    }

    /**
     * Helper function called by _handle_link() on link replacement.
     *
     * Maintains an internal list of links to be displayed at the end of the
     * text, with numeric indices to the original point in the text they
     * appeared.
     *
     * @param string $url    URL of the link
     * @param string $display Part of the text to associate number with
     */
    protected function _build_link_list($url, $display)
    {
        if (($index = array_search($url, $this->_link_list)) === false) {
            $index = count($this->_link_list);
            $this->_link_list[] = $url;
        }

        return $display . ' [' . ($index+1) . ']';
    }

    /**
     * Helper function for PRE body conversion.
     *
     * @param string &$text HTML content
     */
    protected function _convert_pre(&$text)
    {
        // get the content of PRE element
        while (preg_match('/<pre[^>]*>(.*)<\/pre>/ismU', $text, $matches)) {
            $this->pre_content = $matches[1];

            // Run our defined tags search-and-replace with callback
            $this->pre_content = preg_replace_callback($this->callback_search,
                [$this, 'tags_preg_callback'], $this->pre_content);

            // convert the content
            $this->pre_content = sprintf('<div><br>%s<br></div>',
                preg_replace($this->pre_search, $this->pre_replace, $this->pre_content));

            // replace the content (use callback because content can contain $0 variable)
            $text = preg_replace_callback('/<pre[^>]*>.*<\/pre>/ismU',
                [$this, 'pre_preg_callback'], $text, 1);

            // free memory
            $this->pre_content = '';
        }
    }

    /**
     * Helper function for BLOCKQUOTE body conversion.
     *
     * @param string &$text HTML content
     */
    protected function _convert_blockquotes(&$text)
    {
        $level  = 0;
        $offset = 0;

        while (($start = stripos($text, '<blockquote', $offset)) !== false) {
            $offset = $start + 12;

            do {
                $end = stripos($text, '</blockquote>', $offset);
                $next = stripos($text, '<blockquote', $offset);

                // nested <blockquote>, skip
                if ($next !== false && $next < $end) {
                    $offset = $next + 12;
                    $level++;
                }
                // nested </blockquote> tag
                if ($end !== false && $level > 0) {
                    $offset = $end + 12;
                    $level--;
                }
                // found matching end tag
                else if ($end !== false && $level == 0) {
                    $taglen = strpos($text, '>', $start) - $start;
                    $startpos = $start + $taglen + 1;

                    // get blockquote content
                    $body = trim(substr($text, $startpos, $end - $startpos));

                    // adjust text wrapping width
                    $p_width = $this->width;
                    if ($this->width > 0) $this->width -= 2;

                    // replace content with inner blockquotes
                    $this->_converter($body);

                    // restore text width
                    $this->width = $p_width;

                    // Add citation markers and create <pre> block
                    $body = preg_replace_callback('/((?:^|\n)>*)([^\n]*)/', [$this, 'blockquote_citation_callback'], trim($body));
                    $body = '<pre>' . htmlspecialchars($body, ENT_COMPAT | ENT_HTML401 | ENT_SUBSTITUTE, $this->charset) . '</pre>';

                    $text = substr_replace($text, $body . "\n", $start, $end + 13 - $start);
                    $offset = 0;

                    break;
                }
                // abort on invalid tag structure (e.g. no closing tag found)
                else {
                    break;
                }
            }
            while ($end || $next);
        }
    }

    /**
     * Callback function to correctly add citation markers for blockquote contents
     */
    public function blockquote_citation_callback($m)
    {
        $line  = ltrim($m[2]);
        $space = isset($line[0]) && $line[0] == '>' ? '' : ' ';

        return $m[1] . '>' . $space . $line;
    }

    /**
     * Callback function for preg_replace_callback use.
     *
     * @param array $matches PREG matches
     *
     * @return string Element content
     */
    public function tags_preg_callback($matches)
    {
        switch (strtolower($matches[1])) {
        case 'th':
            return $this->_toupper("\t\t". $matches[3] ."\n");
        case 'h':
            return $this->_toupper("\n\n". $matches[3] ."\n\n");
        case 'a':
            // Remove spaces in URL (#1487805)
            $url = str_replace(' ', '', $matches[3]);
            return $this->_handle_link($url, $matches[4]);
        }
    }

    /**
     * Callback function for preg_replace_callback use in PRE content handler.
     *
     * @param array $matches PREG matches
     *
     * @return string PRE content
     */
    public function pre_preg_callback($matches)
    {
        return $this->pre_content;
    }

    /**
     * Strtoupper function with HTML tags and entities handling.
     *
     * @param string $str Text to convert
     *
     * @return string Converted text
     */
    private function _toupper($str)
    {
        // string can containing HTML tags
        $chunks = preg_split('/(<[^>]*>)/', $str, -1, PREG_SPLIT_NO_EMPTY | PREG_SPLIT_DELIM_CAPTURE);

        // convert toupper only the text between HTML tags
        foreach ($chunks as $idx => $chunk) {
            if ($chunk[0] != '<') {
                $chunks[$idx] = $this->_strtoupper($chunk);
            }
        }

        return implode($chunks);
    }

    /**
     * Strtoupper multibyte wrapper function with HTML entities handling.
     *
     * @param string $str Text to convert
     *
     * @return string Converted text
     */
    private function _strtoupper($str)
    {
        $str = html_entity_decode($str, ENT_COMPAT, $this->charset);
        $str = mb_strtoupper($str);
        $str = htmlspecialchars($str, ENT_COMPAT, $this->charset);

        return $str;
    }
}