util.php

<?php
/*
 * StatusNet - the distributed open-source microblogging tool
 * Copyright (C) 2008-2011, StatusNet, Inc.
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU Affero General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU Affero General Public License for more details.
 *
 * You should have received a copy of the GNU Affero General Public License
 * along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

/* XXX: break up into separate modules (HTTP, user, files) */

/**
 * Show a server error.
 */
function common_server_error($msg, $code=500)
{
    $err = new ServerErrorAction($msg, $code);
    $err->showPage();
}

/**
 * Show a user error.
 */
function common_user_error($msg, $code=400)
{
    $err = new ClientErrorAction($msg, $code);
    $err->showPage();
}

/**
 * This should only be used at setup; processes switching languages
 * to send text to other users should use common_switch_locale().
 *
 * @param string $language Locale language code (optional; empty uses
 *                         current user's preference or site default)
 * @return mixed success
 */
function common_init_locale($language=null)
{
    if(!$language) {
        $language = common_language();
    }
    putenv('LANGUAGE='.$language);
    putenv('LANG='.$language);
    $ok =  setlocale(LC_ALL, $language . ".utf8",
                     $language . ".UTF8",
                     $language . ".utf-8",
                     $language . ".UTF-8",
                     $language);

    return $ok;
}

/**
 * Initialize locale and charset settings and gettext with our message catalog,
 * using the current user's language preference or the site default.
 *
 * This should generally only be run at framework initialization; code switching
 * languages at runtime should call common_switch_language().
 *
 * @access private
 */
function common_init_language()
{
    mb_internal_encoding('UTF-8');

    // Note that this setlocale() call may "fail" but this is harmless;
    // gettext will still select the right language.
    $language = common_language();
    $locale_set = common_init_locale($language);

    if (!$locale_set) {
        // The requested locale doesn't exist on the system.
        //
        // gettext seems very picky... We first need to setlocale()
        // to a locale which _does_ exist on the system, and _then_
        // we can set in another locale that may not be set up
        // (say, ga_ES for Galego/Galician) it seems to take it.
        //
        // For some reason C and POSIX which are guaranteed to work
        // don't do the job. en_US.UTF-8 should be there most of the
        // time, but not guaranteed.
        $ok = common_init_locale("en_US");
        if (!$ok && strtolower(substr(PHP_OS, 0, 3)) != 'win') {
            // Try to find a complete, working locale on Unix/Linux...
            // @fixme shelling out feels awfully inefficient
            // but I don't think there's a more standard way.
            $all = `locale -a`;
            foreach (explode("\n", $all) as $locale) {
                if (preg_match('/\.utf[-_]?8$/i', $locale)) {
                    $ok = setlocale(LC_ALL, $locale);
                    if ($ok) {
                        break;
                    }
                }
            }
        }
        if (!$ok) {
            common_log(LOG_ERR, "Unable to find a UTF-8 locale on this system; UI translations may not work.");
        }
        $locale_set = common_init_locale($language);
    }

    common_init_gettext();
}

/**
 * @access private
 */
function common_init_gettext()
{
    setlocale(LC_CTYPE, 'C');
    // So we do not have to make people install the gettext locales
    $path = common_config('site','locale_path');
    bindtextdomain("statusnet", $path);
    bind_textdomain_codeset("statusnet", "UTF-8");
    textdomain("statusnet");
}

/**
 * Switch locale during runtime, and poke gettext until it cries uncle.
 * Otherwise, sometimes it doesn't actually switch away from the old language.
 *
 * @param string $language code for locale ('en', 'fr', 'pt_BR' etc)
 */
function common_switch_locale($language=null)
{
    common_init_locale($language);

    setlocale(LC_CTYPE, 'C');
    // So we do not have to make people install the gettext locales
    $path = common_config('site','locale_path');
    bindtextdomain("statusnet", $path);
    bind_textdomain_codeset("statusnet", "UTF-8");
    textdomain("statusnet");
}

function common_timezone()
{
    if (common_logged_in()) {
        $user = common_current_user();
        if ($user->timezone) {
            return $user->timezone;
        }
    }

    return common_config('site', 'timezone');
}

function common_valid_language($lang)
{
    if ($lang) {
        // Validate -- we don't want to end up with a bogus code
        // left over from some old junk.
        foreach (common_config('site', 'languages') as $code => $info) {
            if ($info['lang'] == $lang) {
                return true;
            }
        }
    }
    return false;
}

function common_language()
{
    // Allow ?uselang=xx override, very useful for debugging
    // and helping translators check usage and context.
    if (isset($_GET['uselang'])) {
        $uselang = strval($_GET['uselang']);
        if (common_valid_language($uselang)) {
            return $uselang;
        }
    }

    // If there is a user logged in and they've set a language preference
    // then return that one...
    if (_have_config() && common_logged_in()) {
        $user = common_current_user();

        if (common_valid_language($user->language)) {
            return $user->language;
        }
    }

    // Otherwise, find the best match for the languages requested by the
    // user's browser...
    if (common_config('site', 'langdetect')) {
        $httplang = isset($_SERVER['HTTP_ACCEPT_LANGUAGE']) ? $_SERVER['HTTP_ACCEPT_LANGUAGE'] : null;
        if (!empty($httplang)) {
            $language = client_prefered_language($httplang);
            if ($language)
              return $language;
        }
    }

    // Finally, if none of the above worked, use the site's default...
    return common_config('site', 'language');
}

/**
 * Salted, hashed passwords are stored in the DB.
 */
function common_munge_password($password, Profile $profile=null)
{
    $hashed = null;

    if (Event::handle('StartHashPassword', array(&$hashed, $password, $profile))) {
        Event::handle('EndHashPassword', array(&$hashed, $password, $profile));
    }
    if (empty($hashed)) {
        throw new PasswordHashException();
    }

    return $hashed;
}

/**
 * Check if a username exists and has matching password.
 */
function common_check_user($nickname, $password)
{
    // empty nickname always unacceptable
    if (empty($nickname)) {
        return false;
    }

    $authenticatedUser = false;

    if (Event::handle('StartCheckPassword', array($nickname, $password, &$authenticatedUser))) {

        if (common_is_email($nickname)) {
            $user = User::getKV('email', common_canonical_email($nickname));
        } else {
            $user = User::getKV('nickname', Nickname::normalize($nickname));
        }

        if ($user instanceof User && !empty($password)) {
            if (0 == strcmp(common_munge_password($password, $user->getProfile()), $user->password)) {
                //internal checking passed
                $authenticatedUser = $user;
            }
        }
    }
    Event::handle('EndCheckPassword', array($nickname, $password, $authenticatedUser));

    return $authenticatedUser;
}

/**
 * Is the current user logged in?
 */
function common_logged_in()
{
    return (!is_null(common_current_user()));
}

function common_have_session()
{
    return (0 != strcmp(session_id(), ''));
}

function common_ensure_session()
{
    $c = null;
    if (array_key_exists(session_name(), $_COOKIE)) {
        $c = $_COOKIE[session_name()];
    }
    if (!common_have_session()) {
        if (common_config('sessions', 'handle')) {
            Session::setSaveHandler();
        }
	if (array_key_exists(session_name(), $_GET)) {
	    $id = $_GET[session_name()];
	} else if (array_key_exists(session_name(), $_COOKIE)) {
	    $id = $_COOKIE[session_name()];
	}
	if (isset($id)) {
	    session_id($id);
	}
        @session_start();
        if (!isset($_SESSION['started'])) {
            $_SESSION['started'] = time();
            if (!empty($id)) {
                common_log(LOG_WARNING, 'Session cookie "' . $_COOKIE[session_name()] . '" ' .
                           ' is set but started value is null');
            }
        }
    }
}

// Three kinds of arguments:
// 1) a user object
// 2) a nickname
// 3) null to clear

// Initialize to false; set to null if none found
$_cur = false;

function common_set_user($user)
{
    global $_cur;

    if (is_null($user) && common_have_session()) {
        $_cur = null;
        unset($_SESSION['userid']);
        return true;
    } else if (is_string($user)) {
        $nickname = $user;
        $user = User::getKV('nickname', $nickname);
    } else if (!$user instanceof User) {
        return false;
    }

    if ($user) {
        if (Event::handle('StartSetUser', array(&$user))) {
            if (!empty($user)) {
                if (!$user->hasRight(Right::WEBLOGIN)) {
                    // TRANS: Authorisation exception thrown when a user a not allowed to login.
                    throw new AuthorizationException(_('Not allowed to log in.'));
                }
                common_ensure_session();
                $_SESSION['userid'] = $user->id;
                $_cur = $user;
                Event::handle('EndSetUser', array($user));
                return $_cur;
            }
        }
    }
    return false;
}

function common_set_cookie($key, $value, $expiration=0)
{
    $path = common_config('site', 'path');
    $server = common_config('site', 'server');

    if ($path && ($path != '/')) {
        $cookiepath = '/' . $path . '/';
    } else {
        $cookiepath = '/';
    }
    return setcookie($key,
                     $value,
                     $expiration,
                     $cookiepath,
                     $server,
                     common_config('site', 'ssl')=='always');
}

define('REMEMBERME', 'rememberme');
define('REMEMBERME_EXPIRY', 30 * 24 * 60 * 60); // 30 days

function common_rememberme($user=null)
{
    if (!$user) {
        $user = common_current_user();
        if (!$user) {
            return false;
        }
    }

    $rm = new Remember_me();

    $rm->code = common_random_hexstr(16);
    $rm->user_id = $user->id;

    // Wrap the insert in some good ol' fashioned transaction code

    $rm->query('BEGIN');

    $result = $rm->insert();

    if (!$result) {
        common_log_db_error($rm, 'INSERT', __FILE__);
        $rm->query('ROLLBACK');
        return false;
    }

    $rm->query('COMMIT');

    $cookieval = $rm->user_id . ':' . $rm->code;

    common_log(LOG_INFO, 'adding rememberme cookie "' . $cookieval . '" for ' . $user->nickname);

    common_set_cookie(REMEMBERME, $cookieval, time() + REMEMBERME_EXPIRY);

    return true;
}

function common_remembered_user()
{
    $user = null;

    $packed = isset($_COOKIE[REMEMBERME]) ? $_COOKIE[REMEMBERME] : null;

    if (!$packed) {
        return null;
    }

    list($id, $code) = explode(':', $packed);

    if (!$id || !$code) {
        common_log(LOG_WARNING, 'Malformed rememberme cookie: ' . $packed);
        common_forgetme();
        return null;
    }

    $rm = Remember_me::getKV('code', $code);

    if (!$rm) {
        common_log(LOG_WARNING, 'No such remember code: ' . $code);
        common_forgetme();
        return null;
    }

    if ($rm->user_id != $id) {
        common_log(LOG_WARNING, 'Rememberme code for wrong user: ' . $rm->user_id . ' != ' . $id);
        common_forgetme();
        return null;
    }

    $user = User::getKV('id', $rm->user_id);

    if (!$user instanceof User) {
        common_log(LOG_WARNING, 'No such user for rememberme: ' . $rm->user_id);
        common_forgetme();
        return null;
    }

    // successful!
    $result = $rm->delete();

    if (!$result) {
        common_log_db_error($rm, 'DELETE', __FILE__);
        common_log(LOG_WARNING, 'Could not delete rememberme: ' . $code);
        common_forgetme();
        return null;
    }

    common_log(LOG_INFO, 'logging in ' . $user->nickname . ' using rememberme code ' . $rm->code);

    common_set_user($user);
    common_real_login(false);

    // We issue a new cookie, so they can log in
    // automatically again after this session

    common_rememberme($user);

    return $user;
}

/**
 * must be called with a valid user!
 */
function common_forgetme()
{
    common_set_cookie(REMEMBERME, '', 0);
}

/**
 * Who is the current user?
 */
function common_current_user()
{
    global $_cur;

    if (!_have_config()) {
        return null;
    }

    if ($_cur === false) {

        if (isset($_COOKIE[session_name()]) || isset($_GET[session_name()])
            || (isset($_SESSION['userid']) && $_SESSION['userid'])) {
            common_ensure_session();
            $id = isset($_SESSION['userid']) ? $_SESSION['userid'] : false;
            if ($id) {
                $user = User::getKV('id', $id);
                if ($user instanceof User) {
                	$_cur = $user;
                	return $_cur;
                }
            }
        }

        // that didn't work; try to remember; will init $_cur to null on failure
        $_cur = common_remembered_user();

        if ($_cur) {
            // XXX: Is this necessary?
            $_SESSION['userid'] = $_cur->id;
        }
    }

    return $_cur;
}

/**
 * Logins that are 'remembered' aren't 'real' -- they're subject to
 * cookie-stealing. So, we don't let them do certain things. New reg,
 * OpenID, and password logins _are_ real.
 */
function common_real_login($real=true)
{
    common_ensure_session();
    $_SESSION['real_login'] = $real;
}

function common_is_real_login()
{
    return common_logged_in() && $_SESSION['real_login'];
}

/**
 * Get a hash portion for HTTP caching Etags and such including
 * info on the current user's session. If login/logout state changes,
 * or we've changed accounts, or we've renamed the current user,
 * we'll get a new hash value.
 *
 * This should not be considered secure information.
 *
 * @param User $user (optional; uses common_current_user() if left out)
 * @return string
 */
function common_user_cache_hash($user=false)
{
    if ($user === false) {
        $user = common_current_user();
    }
    if ($user) {
        return crc32($user->id . ':' . $user->nickname);
    } else {
        return '0';
    }
}

/**
 * get canonical version of nickname for comparison
 *
 * @param string $nickname
 * @return string
 *
 * @throws NicknameException on invalid input
 * @deprecated call Nickname::normalize() directly.
 */
function common_canonical_nickname($nickname)
{
    return Nickname::normalize($nickname);
}

/**
 * get canonical version of email for comparison
 *
 * @fixme actually normalize
 * @fixme reject invalid input
 *
 * @param string $email
 * @return string
 */
function common_canonical_email($email)
{
    // XXX: canonicalize UTF-8
    // XXX: lcase the domain part
    return $email;
}

function common_purify($html)
{
    require_once INSTALLDIR.'/extlib/htmLawed/htmLawed.php';

    $config = array('safe' => 1,    // means that elements=* means elements=*-applet-embed-iframe-object-script or so
                    'elements' => '*',
                    'deny_attribute' => 'id,style,on*',
                    'cdata' => 1);

    // Remove more elements than what the 'safe' filter gives (elements must be '*' before this)
    // http://www.bioinformatics.org/phplabware/internal_utilities/htmLawed/htmLawed_README.htm#s3.6
    foreach (common_config('htmlfilter') as $tag=>$filter) {
        if ($filter === true) {
            $config['elements'] .= "-{$tag}";
        }
    }

    $html = common_remove_unicode_formatting($html);

    $purified = htmLawed($html, $config);
    Event::handle('EndCommonPurify', array(&$purified, $html));
    
    return $purified;
}

function common_remove_unicode_formatting($text)
{
    // Strip Unicode text formatting/direction codes
    // this is pretty dangerous for visualisation of text and can be used for mischief
    return preg_replace('/[\\x{200b}-\\x{200f}\\x{202a}-\\x{202e}]/u', '', $text);
}

/**
 * Partial notice markup rendering step: build links to !group references.
 *
 * @param string    $text partially rendered HTML
 * @param Profile   $author the Profile that is composing the current notice
 * @param Notice    $parent the Notice this is sent in reply to, if any
 * @return string partially rendered HTML
 */
function common_render_content($text, Profile $author, Notice $parent=null)
{
    $text = common_render_text($text);
    $text = common_linkify_mentions($text, $author, $parent);
    return $text;
}

/**
 * Finds @-mentions within the partially-rendered text section and
 * turns them into live links.
 *
 * Should generally not be called except from common_render_content().
 *
 * @param string    $text   partially-rendered HTML
 * @param Profile   $author the Profile that is composing the current notice
 * @param Notice    $parent the Notice this is sent in reply to, if any
 * @return string partially-rendered HTML
 */
function common_linkify_mentions($text, Profile $author, Notice $parent=null)
{
    $mentions = common_find_mentions($text, $author, $parent);

    // We need to go through in reverse order by position,
    // so our positions stay valid despite our fudging with the
    // string!

    $points = array();

    foreach ($mentions as $mention)
    {
        $points[$mention['position']] = $mention;
    }

    krsort($points);

    foreach ($points as $position => $mention) {

        $linkText = common_linkify_mention($mention);

        $text = substr_replace($text, $linkText, $position, $mention['length']);
    }

    return $text;
}

function common_linkify_mention(array $mention)
{
    $output = null;

    if (Event::handle('StartLinkifyMention', array($mention, &$output))) {

        $xs = new XMLStringer(false);

        $attrs = array('href' => $mention['url'],
                       'class' => 'h-card '.$mention['type']);

        if (!empty($mention['title'])) {
            $attrs['title'] = $mention['title'];
        }

        $xs->element('a', $attrs, $mention['text']);

        $output = $xs->getString();

        Event::handle('EndLinkifyMention', array($mention, &$output));
    }

    return $output;
}

function common_get_attentions($text, Profile $sender, Notice $parent=null)
{
    $mentions = common_find_mentions($text, $sender, $parent);
    $atts = array();
    foreach ($mentions as $mention) {
        foreach ($mention['mentioned'] as $mentioned) {
            $atts[$mentioned->getUri()] = $mentioned->getObjectType();
        }
    }
    if ($parent instanceof Notice) {
        $parentAuthor = $parent->getProfile();
        // afaik groups can't be authors
        $atts[$parentAuthor->getUri()] = ActivityObject::PERSON;
    }
    return $atts;
}

/**
 * Find @-mentions in the given text, using the given notice object as context.
 * References will be resolved with common_relative_profile() against the user
 * who posted the notice.
 *
 * Note the return data format is internal, to be used for building links and
 * such. Should not be used directly; rather, call common_linkify_mentions().
 *
 * @param string    $text
 * @param Profile   $sender the Profile that is sending the current text
 * @param Notice    $parent the Notice this text is in reply to, if any
 *
 * @return array
 *
 * @access private
 */
function common_find_mentions($text, Profile $sender, Notice $parent=null)
{
    $mentions = array();

    if (Event::handle('StartFindMentions', array($sender, $text, &$mentions))) {
        // Get the context of the original notice, if any
        $origMentions = array();

        // Does it have a parent notice for context?
        if ($parent instanceof Notice) {
            foreach ($parent->getAttentionProfiles() as $repliedTo) {
                if (!$repliedTo->isPerson()) {
                    continue;
                }
                $origMentions[$repliedTo->getNickname()] = $repliedTo;
            }
        }

        $matches = common_find_mentions_raw($text);

        foreach ($matches as $match) {
            try {
                $nickname = Nickname::normalize($match[0]);
            } catch (NicknameException $e) {
                // Bogus match? Drop it.
                continue;
            }

            // Try to get a profile for this nickname.
            // Start with conversation context, then go to
            // sender context.

            if ($parent instanceof Notice && $parent->getProfile()->getNickname() === $nickname) {
                $mentioned = $parent->getProfile();
            } else if (!empty($origMentions) &&
                       array_key_exists($nickname, $origMentions)) {
                $mentioned = $origMentions[$nickname];
            } else {
                // sets to null if no match
                $mentioned = common_relative_profile($sender, $nickname);
            }

            if ($mentioned instanceof Profile) {
                try {
                    $url = $mentioned->getUrl();
                } catch (InvalidUrlException $e) {
                    $url = common_local_url('userbyid', array('id' => $mentioned->getID()));
                }

                $mention = array('mentioned' => array($mentioned),
                                 'type' => 'mention',
                                 'text' => $match[0],
                                 'position' => $match[1],
                                 'length' => mb_strlen($match[0]),
                                 'title' => $mentioned->getFullname(),
                                 'url' => $url);

                $mentions[] = $mention;
            }
        }

        // @#tag => mention of all subscriptions tagged 'tag'

        preg_match_all('/(?:^|[\s\.\,\:\;]+)@#([\pL\pN_\-\.]{1,64})/',
                       $text, $hmatches, PREG_OFFSET_CAPTURE);
        foreach ($hmatches[1] as $hmatch) {
            $tag = common_canonical_tag($hmatch[0]);
            $plist = Profile_list::getByTaggerAndTag($sender->getID(), $tag);
            if (!$plist instanceof Profile_list || $plist->private) {
                continue;
            }
            $tagged = $sender->getTaggedSubscribers($tag);

            $url = common_local_url('showprofiletag',
                                    array('nickname' => $sender->getNickname(),
                                          'tag' => $tag));

            $mentions[] = array('mentioned' => $tagged,
                                'type'      => 'list',
                                'text' => $hmatch[0],
                                'position' => $hmatch[1],
                                'length' => mb_strlen($hmatch[0]),
                                'url' => $url);
        }

        preg_match_all('/(?:^|[\s\.\,\:\;]+)!(' . Nickname::DISPLAY_FMT . ')/',
                       $text, $hmatches, PREG_OFFSET_CAPTURE);
        foreach ($hmatches[1] as $hmatch) {
            $nickname = Nickname::normalize($hmatch[0]);
            $group = User_group::getForNickname($nickname, $sender);

            if (!$group instanceof User_group || !$sender->isMember($group)) {
                continue;
            }

            $profile = $group->getProfile();

            $mentions[] = array('mentioned' => array($profile),
                                'type'      => 'group',
                                'text'      => $hmatch[0],
                                'position'  => $hmatch[1],
                                'length'    => mb_strlen($hmatch[0]),
                                'url'       => $group->permalink(),
                                'title'     => $group->getFancyName());
        }

        Event::handle('EndFindMentions', array($sender, $text, &$mentions));
    }

    return $mentions;
}

/**
 * Does the actual regex pulls to find @-mentions in text.
 * Should generally not be called directly; for use in common_find_mentions.
 *
 * @param string $text
 * @return array of PCRE match arrays
 */
function common_find_mentions_raw($text)
{
    $tmatches = array();
    preg_match_all('/^T (' . Nickname::DISPLAY_FMT . ') /',
                   $text,
                   $tmatches,
                   PREG_OFFSET_CAPTURE);

    $atmatches = array();
    preg_match_all('/(?:^|\s+)@(' . Nickname::DISPLAY_FMT . ')\b/',
                   $text,
                   $atmatches,
                   PREG_OFFSET_CAPTURE);

    $matches = array_merge($tmatches[1], $atmatches[1]);
    return $matches;
}

function common_render_text($text)
{
    $text = common_remove_unicode_formatting($text);
    $text = nl2br(htmlspecialchars($text));

    $text = preg_replace('/[\x{0}-\x{8}\x{b}-\x{c}\x{e}-\x{19}]/', '', $text);
    $text = common_replace_urls_callback($text, 'common_linkify');
    $text = preg_replace_callback('/(^|\&quot\;|\'|\(|\[|\{|\s+)#([\pL\pN_\-\.]{1,64})/u',
                function ($m) { return "{$m[1]}#".common_tag_link($m[2]); }, $text);
    // XXX: machine tags
    return $text;
}

/**
 * Find links in the given text and pass them to the given callback function.
 *
 * @param string $text
 * @param function($text, $arg) $callback: return replacement text
 * @param mixed $arg: optional argument will be passed on to the callback
 */
function common_replace_urls_callback($text, $callback, $arg = null) {
    // Start off with a regex
    $regex = '#'.
    '(?:^|[\s\<\>\(\)\[\]\{\}\\\'\\\";]+)(?![\@\!\#])'.
    '('.
        '(?:'.
            '(?:'. //Known protocols
                '(?:'.
                    '(?:(?:https?|ftps?|mms|rtsp|gopher|news|nntp|telnet|wais|file|prospero|webcal|ircs?)://)'.
                    '|'.
                    '(?:(?:bitcoin|mailto|aim|tel|xmpp):)'.
                ')'.
                '(?:[\pN\pL\-\_\+\%\~]+(?::[\pN\pL\-\_\+\%\~]+)?\@)?'. //user:pass@
                '(?:'.
                    '(?:'.
                        '\[[\pN\pL\-\_\:\.]+(?<![\.\:])\]'. //[dns]
                    ')|(?:'.
                        '[\pN\pL\-\_\:\.]+(?<![\.\:])'. //dns
                    ')'.
                ')'.
            ')'.
            '|(?:(?:magnet):)'. // URLs without domain name
            '|(?:(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.){3}(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)'. //IPv4
            '|(?:'. //IPv6
                '\[?(?:(?:(?:[0-9A-Fa-f]{1,4}:){7}(?:(?:[0-9A-Fa-f]{1,4})|:))|(?:(?:[0-9A-Fa-f]{1,4}:){6}(?::|(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})|(?::[0-9A-Fa-f]{1,4})))|(?:(?:[0-9A-Fa-f]{1,4}:){5}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?:(?:[0-9A-Fa-f]{1,4}:){4}(?::[0-9A-Fa-f]{1,4}){0,1}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?:(?:[0-9A-Fa-f]{1,4}:){3}(?::[0-9A-Fa-f]{1,4}){0,2}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?:(?:[0-9A-Fa-f]{1,4}:){2}(?::[0-9A-Fa-f]{1,4}){0,3}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?:(?:[0-9A-Fa-f]{1,4}:)(?::[0-9A-Fa-f]{1,4}){0,4}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?::(?::[0-9A-Fa-f]{1,4}){0,5}(?:(?::(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})?)|(?:(?::[0-9A-Fa-f]{1,4}){1,2})))|(?:(?:(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})(?:\.(?:25[0-5]|2[0-4]\d|[01]?\d{1,2})){3})))\]?(?<!:)'.
            ')'.
            (common_config('linkify', 'bare_domains')
                ? '|(?:'. //DNS
                    '(?:[\pN\pL\-\_\+\%\~]+(?:\:[\pN\pL\-\_\+\%\~]+)?\@)?'. //user:pass@
                    '[\pN\pL\-\_]+(?:\.[\pN\pL\-\_]+)*\.'.
                    //tld list from http://data.iana.org/TLD/tlds-alpha-by-domain.txt, also added local, loc, and onion
                    '(?:AC|AD|AE|AERO|AF|AG|AI|AL|AM|AN|AO|AQ|AR|ARPA|AS|ASIA|AT|AU|AW|AX|AZ|BA|BB|BD|BE|BF|BG|BH|BI|BIZ|BJ|BM|BN|BO|BR|BS|BT|BV|BW|BY|BZ|CA|CAT|CC|CD|CF|CG|CH|CI|CK|CL|CM|CN|CO|COM|COOP|CR|CU|CV|CX|CY|CZ|DE|DJ|DK|DM|DO|DZ|EC|EDU|EE|EG|ER|ES|ET|EU|FI|FJ|FK|FM|FO|FR|GA|GB|GD|GE|GF|GG|GH|GI|GL|GM|GN|GOV|GP|GQ|GR|GS|GT|GU|GW|GY|HK|HM|HN|HR|HT|HU|ID|IE|IL|IM|IN|INFO|INT|IO|IQ|IR|IS|IT|JE|JM|JO|JOBS|JP|KE|KG|KH|KI|KM|KN|KP|KR|KW|KY|KZ|LA|LB|LC|LI|LK|LR|LS|LT|LU|LV|LY|MA|MC|MD|ME|MG|MH|MIL|MK|ML|MM|MN|MO|MOBI|MP|MQ|MR|MS|MT|MU|MUSEUM|MV|MW|MX|MY|MZ|NA|NAME|NC|NE|NET|NF|NG|NI|NL|NO|NP|NR|NU|NZ|OM|ORG|PA|PE|PF|PG|PH|PK|PL|PM|PN|PR|PRO|PS|PT|PW|PY|QA|RE|RO|RS|RU|RW|SA|SB|SC|SD|SE|SG|SH|SI|SJ|SK|SL|SM|SN|SO|SR|ST|SU|SV|SY|SZ|TC|TD|TEL|TF|TG|TH|TJ|TK|TL|TM|TN|TO|TP|TR|TRAVEL|TT|TV|TW|TZ|UA|UG|UK|US|UY|UZ|VA|VC|VE|VG|VI|VN|VU|WF|WS|XN--0ZWM56D|测试|XN--11B5BS3A9AJ6G|परीक्षा|XN--80AKHBYKNJ4F|испытание|XN--9T4B11YI5A|테스트|XN--DEBA0AD|טעסט|XN--G6W251D|測試|XN--HGBK6AJ7F53BBA|آزمایشی|XN--HLCJ6AYA9ESC7A|பரிட்சை|XN--JXALPDLP|δοκιμή|XN--KGBECHTV|إختبار|XN--ZCKZAH|テスト|YE|YT|YU|ZA|ZM|ZONE|ZW|local|loc|onion)'.
            ')(?![\pN\pL\-\_])'
                : '') . // if common_config('linkify', 'bare_domains') is false, don't add anything here
        ')'.
        '(?:'.
            '(?:\:\d+)?'. //:port
            '(?:/[\pN\pL$\,\!\(\)\.\:\-\_\+\/\=\&\;\%\~\*\$\+\'@]*)?'. // /path
            '(?:\?[\pN\pL\$\,\!\(\)\.\:\-\_\+\/\=\&\;\%\~\*\$\+\'@\/]*)?'. // ?query string
            '(?:\#[\pN\pL$\,\!\(\)\.\:\-\_\+\/\=\&\;\%\~\*\$\+\'\@/\?\#]*)?'. // #fragment
        ')(?<![\?\.\,\#\,])'.
    ')'.
    '#ixu';
    //preg_match_all($regex,$text,$matches);
    //print_r($matches);
    return preg_replace_callback($regex, curry('callback_helper',$callback,$arg) ,$text);
}

/**
 * Intermediate callback for common_replace_links(), helps resolve some
 * ambiguous link forms before passing on to the final callback.
 *
 * @param array $matches
 * @param callable $callback
 * @param mixed $arg optional argument to pass on as second param to callback
 * @return string
 *
 * @access private
 */
function callback_helper($matches, $callback, $arg=null) {
    $url=$matches[1];
    $left = strpos($matches[0],$url);
    $right = $left+strlen($url);

    $groupSymbolSets=array(
        array(
            'left'=>'(',
            'right'=>')'
        ),
        array(
            'left'=>'[',
            'right'=>']'
        ),
        array(
            'left'=>'{',
            'right'=>'}'
        ),
        array(
            'left'=>'<',
            'right'=>'>'
        )
    );
    $cannotEndWith=array('.','?',',','#');
    $original_url=$url;
    do{
        $original_url=$url;
        foreach($groupSymbolSets as $groupSymbolSet){
            if(substr($url,-1)==$groupSymbolSet['right']){
                $group_left_count = substr_count($url,$groupSymbolSet['left']);
                $group_right_count = substr_count($url,$groupSymbolSet['right']);
                if($group_left_count<$group_right_count){
                    $right-=1;
                    $url=substr($url,0,-1);
                }
            }
        }
        if(in_array(substr($url,-1),$cannotEndWith)){
            $right-=1;
            $url=substr($url,0,-1);
        }
    }while($original_url!=$url);

    $result = call_user_func_array($callback, array($url, $arg));
    return substr($matches[0],0,$left) . $result . substr($matches[0],$right);
}

require_once INSTALLDIR . "/lib/curry.php";

function common_linkify($url) {
    // It comes in special'd, so we unspecial it before passing to the stringifying
    // functions
    $url = htmlspecialchars_decode($url);

    if (strpos($url, '@') !== false && strpos($url, ':') === false && Validate::email($url)) {
        //url is an email address without the mailto: protocol
        $canon = "mailto:$url";
        $longurl = "mailto:$url";
    } else {
        $canon = File_redirection::_canonUrl($url);
        $longurl_data = File_redirection::where($canon, common_config('attachments', 'process_links'));
        
        if(isset($longurl_data->redir_url)) {
			$longurl = $longurl_data->redir_url;
        } else {
            // e.g. local files
	        $longurl = $longurl_data->url;
        }
    }