Index: .htaccess =================================================================== RCS file: /cvs/drupal/drupal/.htaccess,v retrieving revision 1.99 diff -u -9 -p -r1.99 .htaccess --- .htaccess 9 Jan 2009 02:49:01 -0000 1.99 +++ .htaccess 12 Jan 2009 18:39:13 -0000 @@ -80,13 +80,14 @@ DirectoryIndex index.php # # If your site is running in a VirtualDocumentRoot at http://example.com/, # uncomment the following line: # RewriteBase / # Rewrite URLs of the form 'x' to the form 'index.php?q=x'. RewriteCond %{REQUEST_FILENAME} !-f RewriteCond %{REQUEST_FILENAME} !-d RewriteCond %{REQUEST_URI} !=/favicon.ico - RewriteRule ^(.*)$ index.php?q=$1 [L,QSA] + RewriteCond %{THE_REQUEST} ^[^\/]+\/([^?\ ]*) + RewriteRule ^.*$ /index.php?q=%1 [L,QSA,NE] # $Id: .htaccess,v 1.99 2009/01/09 02:49:01 webchick Exp $ Index: includes/common.inc =================================================================== RCS file: /cvs/drupal/drupal/includes/common.inc,v retrieving revision 1.845 diff -u -9 -p -r1.845 common.inc --- includes/common.inc 11 Jan 2009 08:39:07 -0000 1.845 +++ includes/common.inc 12 Jan 2009 18:39:14 -0000 @@ -2802,26 +2802,19 @@ function drupal_json($var = NULL) { * URLs are used, which are interpreted as delimiters by PHP. These * characters are double escaped so PHP will still see the encoded version. * - With clean URLs, Apache changes '//' to '/', so every second slash is * double escaped. * * @param $text * String to encode */ function drupal_urlencode($text) { - if (variable_get('clean_url', '0')) { - return str_replace(array('%2F', '%26', '%23', '//'), - array('/', '%2526', '%2523', '/%252F'), - rawurlencode($text)); - } - else { - return str_replace('%2F', '/', rawurlencode($text)); - } + return str_replace('%2F', '/', rawurlencode($text)); } /** * Returns a string of highly randomized bytes (over the full 8-bit range). * * This function is better than simply calling mt_rand() or any other built-in * PHP function because it can return a long string of bytes (compared to < 4 * bytes normally from mt_rand()) and uses the best available pseudo-random source. * Index: includes/file.inc =================================================================== RCS file: /cvs/drupal/drupal/includes/file.inc,v retrieving revision 1.152 diff -u -9 -p -r1.152 file.inc --- includes/file.inc 10 Jan 2009 06:09:54 -0000 1.152 +++ includes/file.inc 12 Jan 2009 18:39:14 -0000 @@ -82,21 +82,40 @@ define('FILE_STATUS_PERMANENT', 1); * @param $path A string containing the path of the file to generate URL for. * @return A string containing a URL that can be used to download the file. */ function file_create_url($path) { // Strip file_directory_path from $path. We only include relative paths in // URLs. if (strpos($path, file_directory_path() . '/') === 0) { $path = trim(substr($path, strlen(file_directory_path())), '\\/'); } + if (substr(PHP_OS, 0, 3) == 'WIN') { + // On Windows, both "/" and "\" may be used as directory separator, but + // use "/" for prettier URLs. + $path = strtr($path, '\\', '/'); + } switch (variable_get('file_downloads', FILE_DOWNLOADS_PUBLIC)) { case FILE_DOWNLOADS_PUBLIC: - return $GLOBALS['base_url'] . '/' . file_directory_path() . '/' . str_replace('\\', '/', $path); + if (substr(PHP_OS, 0, 3) == 'WIN') { + // The filesystem functions in PHP 5 on Windows do not support + // UTF-8-encoded filenames but always assume that filenames are encoded + // in Windows-1252. In order to support filenames containing characters + // not in Windows-1252, Drupal passes UTF-8-encoded filenames to the + // filesystem functions, making PHP treat each octet in the + // UTF-8-encoded string as a Windows-1252 character. This will make the + // filenames look mangled (like viewing an UTF-8-encoded file in a + // text editor without support for UTF-8) when viewed with external + // programs, e.g. Windows Explorer. The web server does not know about + // this convention, so we need to make the URL reflecting the actual + // filenames when using public files. + $path = drupal_convert_to_utf8($path, 'Windows-1252'); + } + return $GLOBALS['base_url'] . '/' . file_directory_path() . '/' . str_replace('%2F', '/', rawurlencode($path)); case FILE_DOWNLOADS_PRIVATE: return url('system/files/' . $path, array('absolute' => TRUE)); } } /** * Make sure the destination is a complete path and resides in the file system * directory, if it is not prepend the file system directory. * @@ -696,18 +715,24 @@ function file_unmunge_filename($filename * @param $basename * String filename * @param $directory * String directory * @return * File path consisting of $directory and a unique filename based off * of $basename. */ function file_create_filename($basename, $directory) { + // Strip control characters. + $basename = preg_replace('/[\x00-\x1F]/u', '_', $basename); + if (substr(PHP_OS, 0, 3) == 'WIN') { + // These characters are not allowed in filenames on Windows. + $basename = str_replace(array(':', '*', '?', '"', '<', '>', '|'), '_', $basename); + } $destination = $directory . '/' . $basename; if (file_exists($destination)) { // Destination file already exists, generate an alternative. $pos = strrpos($basename, '.'); if ($pos !== FALSE) { $name = substr($basename, 0, $pos); $ext = substr($basename, $pos); } Index: modules/simpletest/tests/file.test =================================================================== RCS file: /cvs/drupal/drupal/modules/simpletest/tests/file.test,v retrieving revision 1.20 diff -u -9 -p -r1.20 file.test --- modules/simpletest/tests/file.test 10 Jan 2009 06:09:54 -0000 1.20 +++ modules/simpletest/tests/file.test 12 Jan 2009 18:39:14 -0000 @@ -1760,9 +1760,184 @@ class FileNameMungingTest extends FileTe * Ensure that unmunge gets your name back. */ function testUnMunge() { $munged_name = file_munge_filename($this->name, '', FALSE); $unmunged_name = file_unmunge_filename($munged_name); $this->assertIdentical($unmunged_name, $this->name, t('The unmunged (%unmunged) filename matches the original (%original)', array('%unmunged' => $unmunged_name, '%original' => $this->name))); } } +/** + * Test file_create_url() by making round-trips through the web server. + */ +class FileCreateUrlUnitTest extends FileTestCase { + function getInfo() { + return array( + 'name' => t('URL generation'), + 'description' => t('Test URL generation'), + 'group' => t('File'), + ); + } + + function setUp() { + parent::setUp('file_test'); + // Clear out any hook calls. + file_test_reset(); + } + + /** + * Test file_create_url() using FILE_DOWNLOADS_PUBLIC. + */ + function testFileCreateUrlPublic() { + global $base_url; + variable_set('file_downloads', FILE_DOWNLOADS_PUBLIC); + + $file = " -._~!$'\"()*@[]?&+%#,;=:\n\x00" . // "Special" ASCII characters. + "%23%25%26%2B%2F%3F" . // Characters that look like a percent-escaped string. + "éøïвβ中國書۞"; // Characters from various non-ASCII alphabets. + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/' . file_directory_path() . '/' . + '%20-._%7E%21%24%27_%28%29_%40%5B%5D_%26%2B%25%23%2C%3B%3D___' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%83%C2%A9%C3%83%C2%B8%C3%83%C2%AF%C3%90%C2%B2%C3%8E%C2%B2%C3%A4%C2%B8%C2%AD%C3%A5%C5%93%E2%80%B9%C3%A6%E2%80%BA%C2%B8%C3%9B%C5%BE'; + } + else { + $expected_url = $base_url . '/' . file_directory_path() . '/' . + '%20-._%7E%21%24%27%22%28%29%2A%40%5B%5D%3F%26%2B%25%23%2C%3B%3D%3A__' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%A9%C3%B8%C3%AF%D0%B2%CE%B2%E4%B8%AD%E5%9C%8B%E6%9B%B8%DB%9E'; + } + $this->checkUrl($file, $expected_url); + + // On Windows, "\" works as path separator; on other platforms it is + // treated as any other character. + $file = 'foo\bar'; + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/' . file_directory_path() . '/foo/bar'; + } + else { + $expected_url = $base_url . '/' . file_directory_path() . '/foo%5Cbar'; + } + $this->checkUrl($file, $expected_url); + } + + /** + * Test file_create_url() using FILE_DOWNLOADS_PRIVATE and clean URLs enabled. + */ + function testFileCreateUrlPrivateCleanUrlEnabled() { + global $base_url; + variable_set('clean_url', '1'); + variable_set('file_downloads', FILE_DOWNLOADS_PRIVATE); + + $file = " -._~!$'\"()*@[]?&+%#,;=:\n\x00" . // "Special" ASCII characters. + "%23%25%26%2B%2F%3F" . // Characters that look like a percent-escaped string. + "éøïвβ中國書۞"; // Characters from various non-ASCII alphabets. + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/system/files/' . + '%20-._%7E%21%24%27_%28%29_%40%5B%5D_%26%2B%25%23%2C%3B%3D___' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%A9%C3%B8%C3%AF%D0%B2%CE%B2%E4%B8%AD%E5%9C%8B%E6%9B%B8%DB%9E'; + } + else { + $expected_url = $base_url . '/system/files/' . + '%20-._%7E%21%24%27%22%28%29%2A%40%5B%5D%3F%26%2B%25%23%2C%3B%3D%3A__' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%A9%C3%B8%C3%AF%D0%B2%CE%B2%E4%B8%AD%E5%9C%8B%E6%9B%B8%DB%9E'; + } + $this->checkUrl($file, $expected_url); + + // "0" is tricky because "0" == FALSE. + $file = '0'; + $expected_url = $base_url . '/system/files/0'; + $this->checkUrl($file, $expected_url); + + // On Windows, "\" works as path separator; on other platforms it is + // treated as any other character. + $file = 'foo\bar'; + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/system/files/foo/bar'; + } + else { + $expected_url = $base_url . '/system/files/foo%5Cbar'; + } + $this->checkUrl($file, $expected_url); + } + + /** + * Test file_create_url() using FILE_DOWNLOADS_PRIVATE and clean URLs + * disabled. + */ + function testFileCreateUrlPrivateCleanUrlDisabled() { + global $base_url; + variable_set('clean_url', '0'); + variable_set('file_downloads', FILE_DOWNLOADS_PRIVATE); + + $file = " -._~!$'\"()*@[]?&+%#,;=:\n\x00" . // "Special" ASCII characters. + "%23%25%26%2B%2F%3F" . // Characters that look like a percent-escaped string. + "éøïвβ中國書۞"; // Characters from various non-ASCII alphabets. + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/?q=system/files/' . + '%20-._%7E%21%24%27_%28%29_%40%5B%5D_%26%2B%25%23%2C%3B%3D___' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%A9%C3%B8%C3%AF%D0%B2%CE%B2%E4%B8%AD%E5%9C%8B%E6%9B%B8%DB%9E'; + } + else { + $expected_url = $base_url . '/?q=system/files/' . + '%20-._%7E%21%24%27%22%28%29%2A%40%5B%5D%3F%26%2B%25%23%2C%3B%3D%3A__' . + '%2523%2525%2526%252B%252F%253F' . + '%C3%A9%C3%B8%C3%AF%D0%B2%CE%B2%E4%B8%AD%E5%9C%8B%E6%9B%B8%DB%9E'; + } + $this->checkUrl($file, $expected_url); + + // "0" is tricky because "0" == FALSE. + $file = '0'; + $expected_url = $base_url . '/?q=system/files/0'; + $this->checkUrl($file, $expected_url); + + // On Windows, "\" works as path separator; on other platforms it is + // treated as any other character. + $file = 'foo\bar'; + if (substr(PHP_OS, 0, 3) == 'WIN') { + $expected_url = $base_url . '/?q=system/files/foo/bar'; + } + else { + $expected_url = $base_url . '/?q=system/files/foo%5Cbar'; + } + $file = 'foo\bar'; + $this->checkUrl($file, $expected_url); + } + + /** + * Check that the URL generated by file_create_url() for the specified file + * equals the specified URL, then fetch the URL and compare the contents to + * the file. + * + * @param $path + * A filepath. + * @param $expected_url + * The expected URL. + */ + private function checkUrl($path, $expected_url) { + // Convert $path to a valid filename, i.e. strip characters not supported + // by the filesystem, and create the file. + $filepath = file_create_filename($path, file_directory_path()); + $directory = dirname($filepath); + file_check_directory($directory, FILE_CREATE_DIRECTORY); + $file = $this->createFile($filepath); + + $url = file_create_url($file->filepath); + $this->assertEqual($url, $expected_url, t('Generated URL matches expected URL')); + + if (variable_get('file_downloads', FALSE) == FILE_DOWNLOADS_PRIVATE) { + // Tell the implementation of hook_file_download() in file_test.module + // that this file may be downloaded. + file_test_set_return('download', array('X-Foo: Bar')); + } + + $this->drupalGet($url); + if ($this->assertResponse(200) == 'pass') { + $this->assertRaw(file_get_contents($file->filepath), t('Contents of the file are correct.')); + } + + file_delete($file); + } +}