<?php

/**
 * Joomla! Content Management System
 *
 * @copyright  (C) 2007 Open Source Matters, Inc. <https://www.joomla.org>
 * @license    GNU General Public License version 2 or later; see LICENSE.txt
 */

namespace Joomla\CMS\Mail;

use Joomla\CMS\Router\Route;
use Joomla\CMS\String\PunycodeHelper;
use Joomla\CMS\Uri\Uri;

// phpcs:disable PSR1.Files.SideEffects
\defined('JPATH_PLATFORM') or die;
// phpcs:enable PSR1.Files.SideEffects

/**
 * Email helper class, provides static methods to perform various tasks relevant
 * to the Joomla email routines.
 *
 * @todo: Test these methods as the regex work is first run and not tested thoroughly
 *
 * @since  1.7.0
 */
abstract class MailHelper
{
    /**
     * Cleans single line inputs.
     *
     * @param   string  $value  String to be cleaned.
     *
     * @return  string  Cleaned string.
     *
     * @since   1.7.0
     */
    public static function cleanLine($value)
    {
        $value = PunycodeHelper::emailToPunycode($value);

        return trim(preg_replace('/(%0A|%0D|\n+|\r+)/i', '', $value));
    }

    /**
     * Cleans multi-line inputs.
     *
     * @param   string  $value  Multi-line string to be cleaned.
     *
     * @return  string  Cleaned multi-line string.
     *
     * @since   1.7.0
     */
    public static function cleanText($value)
    {
        return trim(preg_replace('/(%0A|%0D|\n+|\r+)(content-type:|to:|cc:|bcc:)/i', '', $value));
    }

    /**
     * Cleans any injected headers from the email body.
     *
     * @param   string  $body  email body string.
     *
     * @return  string  Cleaned email body string.
     *
     * @since   1.7.0
     */
    public static function cleanBody($body)
    {
        // Strip all email headers from a string
        return preg_replace("/((From:|To:|Cc:|Bcc:|Subject:|Content-type:) ([\S]+))/", '', $body);
    }

    /**
     * Cleans any injected headers from the subject string.
     *
     * @param   string  $subject  email subject string.
     *
     * @return  string  Cleaned email subject string.
     *
     * @since   1.7.0
     */
    public static function cleanSubject($subject)
    {
        return preg_replace("/((From:|To:|Cc:|Bcc:|Content-type:) ([\S]+))/", '', $subject);
    }

    /**
     * Verifies that an email address does not have any extra headers injected into it.
     *
     * @param   string  $address  email address.
     *
     * @return  mixed   email address string or boolean false if injected headers are present.
     *
     * @since   1.7.0
     */
    public static function cleanAddress($address)
    {
        if (preg_match("[\s;,]", $address)) {
            return false;
        }

        return $address;
    }

    /**
     * Verifies that the string is in a proper email address format.
     *
     * @param   string  $email  String to be verified.
     *
     * @return  boolean  True if string has the correct format; false otherwise.
     *
     * @since   1.7.0
     */
    public static function isEmailAddress($email)
    {
        // Split the email into a local and domain
        $atIndex = strrpos($email, '@');
        $domain = substr($email, $atIndex + 1);
        $local = substr($email, 0, $atIndex);

        // Check Length of domain
        $domainLen = \strlen($domain);

        if ($domainLen < 1 || $domainLen > 255) {
            return false;
        }

        /*
         * Check the local address
         * We're a bit more conservative about what constitutes a "legal" address, that is, a-zA-Z0-9.!#$%&'*+/=?^_`{|}~-
         * The first and last character in local cannot be a period ('.')
         * Also, period should not appear 2 or more times consecutively
         */
        $allowed = "a-zA-Z0-9.!#$%&'*+\/=?^_`{|}~-";
        $regex = "/^[$allowed][\.$allowed]{0,63}$/";

        if (!preg_match($regex, $local) || substr($local, -1) === '.' || $local[0] === '.' || preg_match('/\.\./', $local)) {
            return false;
        }

        // No problem if the domain looks like an IP address, ish
        $regex = '/^[0-9\.]+$/';

        if (preg_match($regex, $domain)) {
            return true;
        }

        // Check Lengths
        $localLen = \strlen($local);

        if ($localLen < 1 || $localLen > 64) {
            return false;
        }

        // Check the domain
        $domain_array = explode('.', $domain);
        $regex = '/^[a-zA-Z0-9-]+(?:\.[a-zA-Z0-9-]+)*$/';

        foreach ($domain_array as $domain) {
            // Convert domain to punycode
            $domain = PunycodeHelper::toPunycode($domain);

            // Must be something
            if (!$domain) {
                return false;
            }

            // Check for invalid characters
            if (!preg_match($regex, $domain)) {
                return false;
            }

            // Check for a dash at the beginning of the domain
            if (strpos($domain, '-') === 0) {
                return false;
            }

            // Check for a dash at the end of the domain
            $length = \strlen($domain) - 1;

            if (strpos($domain, '-', $length) === $length) {
                return false;
            }
        }

        return true;
    }

    /**
     * Convert relative (links, images sources) to absolute urls so that content is accessible in email
     *
     * @param   string  $content  The content need to convert
     *
     * @return  string  The converted content which the relative urls are converted to absolute urls
     *
     * @since  4.1.0
     */
    public static function convertRelativeToAbsoluteUrls($content)
    {
        $siteUrl = Uri::root();

        // Replace none SEF URLs by absolute SEF URLs
        if (strpos($content, 'href="index.php?') !== false) {
            preg_match_all('#href="index.php\?([^"]+)"#m', $content, $matches);

            foreach ($matches[1] as $urlQueryString) {
                $content = str_replace(
                    'href="index.php?' . $urlQueryString . '"',
                    'href="' . Route::link('site', 'index.php?' . $urlQueryString, Route::TLS_IGNORE, true) . '"',
                    $content
                );
            }

            self::checkContent($content);
        }

        // Replace relative links, image sources with absolute Urls
        $protocols  = '[a-zA-Z0-9\-]+:';
        $attributes = array('href=', 'src=', 'poster=');

        foreach ($attributes as $attribute) {
            if (strpos($content, $attribute) !== false) {
                $regex = '#\s' . $attribute . '"(?!/|' . $protocols . '|\#|\')([^"]*)"#m';

                $content = preg_replace($regex, ' ' . $attribute . '"' . $siteUrl . '$1"', $content);

                self::checkContent($content);
            }
        }

        return $content;
    }

    /**
     * Check the content after regular expression function call.
     *
     * @param   string  $content  Content to be checked.
     *
     * @return  void
     *
     * @throws  \RuntimeException  If there is an error in previous regular expression function call.
     * @since  4.1.0
     */
    private static function checkContent($content)
    {
        if ($content !== null) {
            return;
        }

        switch (preg_last_error()) {
            case PREG_BACKTRACK_LIMIT_ERROR:
                $message = 'PHP regular expression limit reached (pcre.backtrack_limit)';
                break;
            case PREG_RECURSION_LIMIT_ERROR:
                $message = 'PHP regular expression limit reached (pcre.recursion_limit)';
                break;
            case PREG_BAD_UTF8_ERROR:
                $message = 'Bad UTF8 passed to PCRE function';
                break;
            default:
                $message = 'Unknown PCRE error calling PCRE function';
        }

        throw new \RuntimeException($message);
    }
}