Ticket #22157: readymadeweb-filetype-HEAD.patch

wordpress-importer.php

@@ -838 +838 @@
                 wp_update_attachment_metadata( $post_id, wp_generate_attachment_metadata( $post_id, $upload['file'] ) );
 
                 // remap resized image URLs, works by stripping the extension and remapping the URL stub.
-                if ( preg_match( '!^image/!', $info['type'] ) ) {
+                // Note that we don't do this in the case that the original upload didn't
+                // have a valid extension, but we were able to determine one by looking at
+                // the relevant HTTP headers (as denoted by the
+                // 'force_extension_in_replacement' flag)
+                if ( preg_match( '!^image/!', $info['type'] ) && empty($upload['force_extension_in_replacement'])) {
                         $parts = pathinfo( $url );
                         $name = basename( $parts['basename'], ".{$parts['extension']}" ); // PATHINFO_FILENAME in PHP 5.2
 
@@ -859 +863 @@
          * @return array|WP_Error Local file location details on success, WP_Error otherwise
          */
         function fetch_remote_file( $url, $post ) {
+
                 // extract the file name and extension from the url
                 $file_name = basename( $url );
 
+                // Some systems, like moveable-type / typepad, advertise files w/o
+                // extension information attached.      We can get around this limitation
+                // by grabbing information about the file from the webserver
+                $file_type = wp_check_filetype($file_name);
+
+                // If there is a situation where we're adding a file extension
+                // onto an upload that didn't originally have one, we
+                // can tell the remapper to not strip off the extension/
+                $force_extension_in_replacement = false;
+
+                if ( empty( $file_type['ext'] ) ) {
+
+                        $retured_file_info = $this->fetch_attachment_info_from_server( $url );
+
+                        if ( $retured_file_info ) {
+                                $file_name = $retured_file_info['filename'];
+                                $force_extension_in_replacement = true;
+                        }
+                }
+
                 // get placeholder file in the upload dir with a unique, sanitized filename
                 $upload = wp_upload_bits( $file_name, 0, '', $post['upload_date'] );
                 if ( $upload['error'] )
@@ -907 +932 @@
                 if ( isset($headers['x-final-location']) && $headers['x-final-location'] != $url )
                         $this->url_remap[$headers['x-final-location']] = $upload['url'];
 
+                if ($force_extension_in_replacement)
+                        $upload['force_extension_in_replacement'] = true;
                 return $upload;
         }
 
@@ -1081 +1108 @@
         function cmpr_strlen( $a, $b ) {
                 return strlen($b) - strlen($a);
         }
+
+        /**
+         * Attempts to grab information about a file from a URL.  This
+         * call just requests the header information from the server (the
+         * body of the resource isn't fetched) so the call is relativly lite.
+         *
+         * @param string $url
+         *   A valid url to request mime info from
+         *
+         * @return array|bool
+         *   If the server didn't return a valid header, false is returned.  Otherwise,
+         *   an array with the following keys is returned:
+         *     'ext' (string):  The extension of the file, if avaiable
+         *     'type' (string): The advertised mime type of the resource, if available
+         *     'filename' (string):
+         *                      The suggested name of the file, if available
+         */
+        function fetch_attachment_info_from_server($url) {
+
+                $results = array(
+                        'ext' => '',
+                        'type' => '',
+                        'filename' => '',
+                );
+
+                $header_info = wp_remote_head( $url );
+
+                if ( is_wp_error( $header_info ) ) {
+                        return false;
+                }
+
+                $content_type = wp_remote_retrieve_header( $header_info, 'content-type' );
+                if ( $content_type ) {
+                        $results['type'] = trim( $content_type );
+                }
+
+                $display_disposition = wp_remote_retrieve_header( $header_info, 'content-disposition' );
+                $file_name_pattern = '/filename=([^\s]+)/i';
+                $file_name_matches = array();
+                if ( preg_match( $file_name_pattern, $display_disposition, $file_name_matches ) ) {
+                        $results['filename'] = trim( $file_name_matches[1] );
+
+                        if ( ( $index = strripos( $results['filename'], '.' ) ) !== false ) {
+                                $results['ext'] = substr( $results['filename'], $index + 1 );
+                        }
+                }
+
+                return $results;
+        }
 }
 
 } // class_exists( 'WP_Importer' )