Ticket #2665: xanga.2.php

<?php

//Xanga archive importer by Jeremy Jay
//Modified for current WP release by Daniel Kozlowski
//
//Borrows heavily from the LiveJournal import script.

class Xanga_Import {

    var $file;

    function header() {
        echo '<div class="wrap">';
        echo '<h2>'.__('Import Xanga').'</h2>';
    }

    function footer() {
        echo '</div>';
    }

    function unhtmlentities($string) { // From php.net for < 4.3 compatability
        $trans_tbl = get_html_translation_table(HTML_ENTITIES);
        $trans_tbl = array_flip($trans_tbl);
        return strtr($string, $trans_tbl);
    }
    
    function greet() {
        echo '<p>'.__('Howdy! This importer allows you to extract posts and comments from Xanga Premium Archive files into your blog.  If you do not have Premium but have enough posts to be looking at this, just pay $4 for a month to get the archive and you will at least be supporting Xanga for all the hosting they have done for you.  Pick an archive file to upload and click Import.').'</p>';
        wp_import_upload_form("admin.php?import=xanga&amp;step=1");
    }

    function import_posts() {
        global $wpdb, $current_user;
        
        set_magic_quotes_runtime(0);
        $importdata = file($this->file); // Read the file into an array
        $importdata = implode('', $importdata); // squish it
        $importdata = str_replace(array ("\r\n", "\r"), "\n", $importdata);
        
        preg_match_all('|<div class="blogheader">(.*?)<hr size=1 noshade>(<div class="blogheader">)*?|is', $importdata, $posts);
        $posts = $posts[1];
        unset($importdata);
        echo '<ol>';        
        
        foreach ($posts as $post) {
            flush();
            preg_match('|^(.*?)</div>|is', $post, $post_title);
            $post_title = $wpdb->escape(trim($post_title[1]));
            
            
            // <div class="smalltext">Posted 7/31/2004 at 4:56 PM</div>
            preg_match('|<div class="smalltext">Posted (\d{1,2}/\d{1,2}/\d{4}) at|i', $post, $post_date);
            $post_date = strtotime($post_date[1]);

            $com = split(' ', $post_title);

            if( $com[1]=='Comments' ) {
                preg_match_all('|<div class="ctextfooterwrap"><div class="ctext">(.*?)</div></div>|is', $post, $comments);
                $comments = $comments[1];

                $comment_post_ID = $post_id;
                $num_comments = 0;
                foreach ($comments as $comment) {
                    preg_match('|^(.*?)</div><div class="cfooter">|is', $comment, $comment_content);
                    $comment_content = str_replace(array ('<![CDATA[', ']]>'), '', trim($comment_content[1]));
                    $comment_content = $this->unhtmlentities($comment_content);

                    // Clean up content
                    $comment_content = preg_replace('|<(/?[A-Z]+)|e', "'<' . strtolower('$1')", $comment_content);
                    $comment_content = str_replace('<br>', '<br />', $comment_content);
                    $comment_content = str_replace('<hr>', '<hr />', $comment_content);
                    $comment_content = $wpdb->escape($comment_content);

                    preg_match('|<div class="cfooter">Posted (\d{1,2}/\d{1,2}/\d{4}) at |i', $comment, $comment_date);
                    $comment_date = trim($comment_date[1]);
                    $comment_date = date('Y-m-d H:i:s', strtotime($comment_date));

                    preg_match('|<a href="http://www\.xanga\.com/home\.aspx\?user=(.*?)">(.*?)</a>|i', $comment, $comment_author);
                    $comment_author = $wpdb->escape(trim($comment_author[1]));
                    $comment_author_url = 'http://www.xanga.com/home.aspx?user='.$comment_author;

                    $comment_approved = 1;
                    // Check if it's already there
                    if (!comment_exists($comment_author, $comment_date)) {
                        $commentdata = compact('comment_post_ID', 'comment_author', 'comment_author_url', 'comment_date', 'comment_content', 'comment_approved');
                        $commentdata = wp_filter_comment($commentdata);
                        wp_insert_comment($commentdata);
                        $num_comments++;
                    }
                }
                if ( $num_comments ) {
                    echo ' ';
                    printf(__('(%s comments)'), $num_comments);
                }
            } else {
                $post_title.=' at '.date('h:i a', $post_date);
                $post_date = gmdate('Y-m-d H:i:s', $post_date);

                preg_match('|<td style="padding-left:20; padding-bottom:20">(.*?)<div class="smalltext">Posted (\d{1,2}/\d{1,2}/\d{4}) at |i', $post, $post_content);
                $post_content = str_replace(array ('<![CDATA[', ']]>'), '', trim($post_content[1]));
                $post_content = $this->unhtmlentities($post_content);

                // Clean up content
                $post_content = preg_replace('|<(/?[A-Z]+)|e', "'<' . strtolower('$1')", $post_content);
                $post_content = str_replace('<br>', '<br />', $post_content);
                $post_content = str_replace('<hr>', '<hr />', $post_content);
                $post_content = $wpdb->escape($post_content);

                $post_author = $current_user->ID;
                $post_status = 'publish';

                echo '<li>';
                if ($post_id = post_exists($post_title, $post_content, $post_date)) {
                    printf(__('Post <i>%s</i> already exists.'), stripslashes($post_title));
                } else {
                    printf(__('Importing post <i>%s</i>...'), stripslashes($post_title));
                    $postdata = compact('post_author', 'post_date', 'post_content', 'post_title', 'post_status');
                    $post_id = wp_insert_post($postdata);
                    if (!$post_id) {
                        _e("Couldn't get post ID");
                        echo '</li>';
                        break;
                    }
                }
            }

            echo '</li>';
            flush();
            ob_flush();
        }
        echo '</ol>';
    }

    function import() {
        $file = wp_import_handle_upload();
        if ( isset($file['error']) ) {
            echo $file['error'];
            return;
        }

        $this->file = $file['file'];
        $this->import_posts();
        wp_import_cleanup($file['id']);
        
        echo '<h3>';
        printf(__('All done. <a href="%s">Have fun!</a>'), get_option('home'));
        echo '</h3>';
    }

    function dispatch() {
        if (empty ($_GET['step']))
            $step = 0;
        else
            $step = (int) $_GET['step'];

        $this->header();
        
        switch ($step) {
            case 0 :
                $this->greet();
                break;
            case 1 :
                $this->import();
                break;
        }
        
        $this->footer();
    }

    function Xanga_Import() {
        // Nothing.    
    }
}

$xanga_import = new Xanga_Import();

register_importer('xanga', 'Xanga', __('Import posts from Xanga Archives'), array ($xanga_import, 'dispatch'));
?>