X-Git-Url: https://vcs.fsf.org/?p=squirrelmail.git;a=blobdiff_plain;f=functions%2Fstrings.php;h=9383d83d09120c06c0b0f8bb67d59eb2171ea6f8;hp=9a5de4226fa616da2ae9a7c449f2674aa7c10cbd;hb=70a4976073135f5c221a121f45f5a551dcd39cde;hpb=c7aff9387859d88801fd2c6738ff32d6bbcc0114 diff --git a/functions/strings.php b/functions/strings.php index 9a5de422..9383d83d 100644 --- a/functions/strings.php +++ b/functions/strings.php @@ -3,45 +3,24 @@ /** * strings.php * - * Copyright (c) 1999-2005 The SquirrelMail Project Team - * Licensed under the GNU GPL. For full terms see the file COPYING. - * * This code provides various string manipulation functions that are * used by the rest of the SquirrelMail code. * + * @copyright © 1999-2007 The SquirrelMail Project Team + * @license http://opensource.org/licenses/gpl-license.php GNU Public License * @version $Id$ * @package squirrelmail */ -/** - * SquirrelMail version number -- DO NOT CHANGE - */ -global $version; -$version = '1.5.1 [CVS]'; - -/** - * SquirrelMail internal version number -- DO NOT CHANGE - * $sm_internal_version = array (release, major, minor) - */ -global $SQM_INTERNAL_VERSION; -$SQM_INTERNAL_VERSION = array(1,5,1); - -/** - * There can be a circular issue with includes, where the $version string is - * referenced by the include of global.php, etc. before it's defined. - * For that reason, bring in global.php AFTER we define the version strings. - */ -require_once(SM_PATH . 'functions/global.php'); - /** * Appends citation markers to the string. * Also appends a trailing space. * * @author Justus Pendleton - * - * @param string str The string to append to - * @param int citeLevel the number of markers to append + * @param string $str The string to append to + * @param int $citeLevel the number of markers to append * @return null + * @since 1.5.1 */ function sqMakeCite (&$str, $citeLevel) { for ($i = 0; $i < $citeLevel; $i++) { @@ -57,11 +36,11 @@ function sqMakeCite (&$str, $citeLevel) { * markers to the newline as necessary. * * @author Justus Pendleton - * - * @param string str the string to make a newline in - * @param int citeLevel the citation level the newline is at - * @param int column starting column of the newline + * @param string $str the string to make a newline in + * @param int $citeLevel the citation level the newline is at + * @param int $column starting column of the newline * @return null + * @since 1.5.1 */ function sqMakeNewLine (&$str, $citeLevel, &$column) { $str .= "\n"; @@ -77,8 +56,6 @@ function sqMakeNewLine (&$str, $citeLevel, &$column) { /** * Checks for spaces in strings - only used if PHP doesn't have native ctype support * - * @author Tomas Kuliavas - * * You might be able to rewrite the function by adding short evaluation form. * * possible problems: @@ -94,6 +71,7 @@ function sqMakeNewLine (&$str, $citeLevel, &$column) { * * @param string $string tested string * @return bool true when only whitespace symbols are present in test string + * @since 1.5.1 */ function sm_ctype_space($string) { if ( preg_match('/^[\x09-\x0D]|^\x20/', $string) || $string=='') { @@ -110,10 +88,10 @@ function sm_ctype_space($string) { * bit smarter and when and how to wrap. * * @author Justus Pendleton - * - * @param string body the entire body of text - * @param int wrap the maximum line length + * @param string $body the entire body of text + * @param int $wrap the maximum line length * @return string the wrapped text + * @since 1.5.1 */ function &sqBodyWrap (&$body, $wrap) { //check for ctype support, and fake it if it doesn't exist @@ -127,7 +105,7 @@ function &sqBodyWrap (&$body, $wrap) { $outString = ''; // current column since the last newline in the outstring $outStringCol = 0; - $length = strlen($body); + $length = sq_strlen($body); // where we are in the original string $pos = 0; // the number of >>> citation markers we are currently at @@ -139,12 +117,12 @@ function &sqBodyWrap (&$body, $wrap) { // we're at the beginning of a line, get the new cite level $newCiteLevel = 0; - while (($pos < $length) && ($body{$pos} == '>')) { + while (($pos < $length) && (sq_substr($body,$pos,1) == '>')) { $newCiteLevel++; $pos++; // skip over any spaces interleaved among the cite markers - while (($pos < $length) && ($body{$pos} == ' ')) { + while (($pos < $length) && (sq_substr($body,$pos,1) == ' ')) { $pos++; @@ -157,8 +135,8 @@ function &sqBodyWrap (&$body, $wrap) { // special case: if this is a blank line then maintain it // (i.e. try to preserve original paragraph breaks) // unless they occur at the very beginning of the text - if (($body{$pos} == "\n" ) && (strlen($outString) != 0)) { - $outStringLast = $outString{strlen($outString) - 1}; + if ((sq_substr($body,$pos,1) == "\n" ) && (sq_strlen($outString) != 0)) { + $outStringLast = $outString{sq_strlen($outString) - 1}; if ($outStringLast != "\n") { $outString .= "\n"; } @@ -192,7 +170,7 @@ function &sqBodyWrap (&$body, $wrap) { } // find the next newline -- we don't want to go further than that - $nextNewline = strpos ($body, "\n", $pos); + $nextNewline = sq_strpos ($body, "\n", $pos); if ($nextNewline === FALSE) { $nextNewline = $length; } @@ -201,7 +179,7 @@ function &sqBodyWrap (&$body, $wrap) { // will work fine for this. Maybe revisit this later though // (for completeness more than anything else, I think) if ($citeLevel == 0) { - $outString .= substr ($body, $pos, ($nextNewline - $pos)); + $outString .= sq_substr ($body, $pos, ($nextNewline - $pos)); $outStringCol = $nextNewline - $pos; if ($nextNewline != $length) { sqMakeNewLine ($outString, 0, $outStringCol); @@ -217,7 +195,7 @@ function &sqBodyWrap (&$body, $wrap) { // the next newline while ($pos < $nextNewline) { // skip over initial spaces - while (($pos < $nextNewline) && (ctype_space ($body{$pos}))) { + while (($pos < $nextNewline) && (ctype_space (sq_substr($body,$pos,1)))) { $pos++; } // if this is a short line then just append it and continue outer loop @@ -225,24 +203,24 @@ function &sqBodyWrap (&$body, $wrap) { // if this is the final line in the input string then include // any trailing newlines // echo substr($body,$pos,$wrap). "
"; - if (($nextNewline + 1 == $length) && ($body{$nextNewline} == "\n")) { + if (($nextNewline + 1 == $length) && (sq_substr($body,$nextNewline,1) == "\n")) { $nextNewline++; } // trim trailing spaces $lastRealChar = $nextNewline; - while (($lastRealChar > $pos && $lastRealChar < $length) && (ctype_space ($body{$lastRealChar}))) { + while (($lastRealChar > $pos && $lastRealChar < $length) && (ctype_space (sq_substr($body,$lastRealChar,1)))) { $lastRealChar--; } // decide if appending the short string is what we want - if (($nextNewline < $length && $body{$nextNewline} == "\n") && + if (($nextNewline < $length && sq_substr($body,$nextNewline,1) == "\n") && isset($lastRealChar)) { $mypos = $pos; //check the first word: - while (($mypos < $length) && ($body{$mypos} == '>')) { + while (($mypos < $length) && (sq_substr($body,$mypos,1) == '>')) { $mypos++; // skip over any spaces interleaved among the cite markers - while (($mypos < $length) && ($body{$mypos} == ' ')) { + while (($mypos < $length) && (sq_substr($body,$mypos,1) == ' ')) { $mypos++; } } @@ -255,15 +233,15 @@ function &sqBodyWrap (&$body, $wrap) { } */ - $firstword = substr($body,$mypos,strpos($body,' ',$mypos) - $mypos); + $firstword = sq_substr($body,$mypos,sq_strpos($body,' ',$mypos) - $mypos); //if ($dowrap || $ldnspacecnt > 1 || ($firstword && ( if (!$smartwrap || $firstword && ( $firstword{0} == '-' || $firstword{0} == '+' || $firstword{0} == '*' || - $firstword{0} == strtoupper($firstword{0}) || + sq_substr($firstword,0,1) == sq_strtoupper(sq_substr($firstword,0,1)) || strpos($firstword,':'))) { - $outString .= substr($body,$pos,($lastRealChar - $pos+1)); + $outString .= sq_substr($body,$pos,($lastRealChar - $pos+1)); $outStringCol += ($lastRealChar - $pos); sqMakeNewLine($outString,$citeLevel,$outStringCol); $nextNewline++; @@ -274,7 +252,7 @@ function &sqBodyWrap (&$body, $wrap) { } - $outString .= substr ($body, $pos, ($lastRealChar - $pos + 1)); + $outString .= sq_substr ($body, $pos, ($lastRealChar - $pos + 1)); $outStringCol += ($lastRealChar - $pos); $pos = $nextNewline + 1; continue; @@ -293,7 +271,7 @@ function &sqBodyWrap (&$body, $wrap) { // start looking backwards for whitespace to break at. $breakPoint = $eol; - while (($breakPoint > $pos) && (! ctype_space ($body{$breakPoint}))) { + while (($breakPoint > $pos) && (! ctype_space (sq_substr($body,$breakPoint,1)))) { $breakPoint--; } @@ -326,13 +304,13 @@ function &sqBodyWrap (&$body, $wrap) { } // skip newlines or whitespace at the beginning of the string - $substring = substr ($body, $pos, ($breakPoint - $pos)); + $substring = sq_substr ($body, $pos, ($breakPoint - $pos)); $substring = rtrim ($substring); // do rtrim and ctype_space have the same ideas about whitespace? $outString .= $substring; - $outStringCol += strlen ($substring); + $outStringCol += sq_strlen ($substring); // advance past the whitespace which caused the wrap $pos = $breakPoint; - while (($pos < $length) && (ctype_space ($body{$pos}))) { + while (($pos < $length) && (ctype_space (sq_substr($body,$pos,1)))) { $pos++; } if ($pos < $length) { @@ -350,17 +328,19 @@ function &sqBodyWrap (&$body, $wrap) { * Has a problem with special HTML characters, so call this before * you do character translation. * - * Specifically, ' comes up as 5 characters instead of 1. + * Specifically, &#039; comes up as 5 characters instead of 1. * This should not add newlines to the end of lines. * - * @param string line the line of text to wrap, by ref - * @param int wrap the maximum line lenth - * @param string charset name of charset used in $line string. Available since v.1.5.1. + * @param string $line the line of text to wrap, by ref + * @param int $wrap the maximum line lenth + * @param string $charset name of charset used in $line string. Available since v.1.5.1. * @return void + * @since 1.0 */ function sqWordWrap(&$line, $wrap, $charset='') { global $languages, $squirrelmail_language; + // Use custom wrapping function, if translation provides it if (isset($languages[$squirrelmail_language]['XTRA_CODE']) && function_exists($languages[$squirrelmail_language]['XTRA_CODE'] . '_wordwrap')) { if (mb_detect_encoding($line) != 'ASCII') { @@ -412,8 +392,9 @@ function sqWordWrap(&$line, $wrap, $charset='') { /** * Does the opposite of sqWordWrap() - * @param string body the text to un-wordwrap + * @param string $body the text to un-wordwrap * @return void + * @since 1.0 */ function sqUnWordWrap(&$body) { global $squirrelmail_language; @@ -457,6 +438,7 @@ function sqUnWordWrap(&$body) { * @param string haystack full mailbox name to search * @param string needle the mailbox separator character * @return string the last part of the mailbox name + * @since 1.0 */ function readShortMailboxName($haystack, $needle) { @@ -472,50 +454,23 @@ function readShortMailboxName($haystack, $needle) { return( $elem ); } -/** - * php_self - * - * Creates an URL for the page calling this function, using either the PHP global - * REQUEST_URI, or the PHP global PHP_SELF with QUERY_STRING added. - * - * @return string the complete url for this page - */ -function php_self () { - if ( sqgetGlobalVar('REQUEST_URI', $req_uri, SQ_SERVER) && !empty($req_uri) ) { - return $req_uri; - } - - if ( sqgetGlobalVar('PHP_SELF', $php_self, SQ_SERVER) && !empty($php_self) ) { - - // need to add query string to end of PHP_SELF to match REQUEST_URI - // - if ( sqgetGlobalVar('QUERY_STRING', $query_string, SQ_SERVER) && !empty($query_string) ) { - $php_self .= '?' . $query_string; - } - - return $php_self; - } - - return ''; -} - /** * get_location * * Determines the location to forward to, relative to your server. * This is used in HTTP Location: redirects. - * If this doesnt work correctly for you (although it should), you can - * remove all this code except the last two lines, and have it return - * the right URL for your site, something like: * - * http://www.example.com/squirrelmail/ + * If set, it uses $config_location_base as the first part of the URL, + * specifically, the protocol, hostname and port parts. The path is + * always autodetected. * * @return string the base url for this SquirrelMail installation + * @since 1.0 */ function get_location () { - global $imap_server_type; + global $imap_server_type, $config_location_base; /* Get the path, handle virtual directories */ if(strpos(php_self(), '?')) { @@ -524,9 +479,16 @@ function get_location () { $path = php_self(); } $path = substr($path, 0, strrpos($path, '/')); + + // proto+host+port are already set in config: + if ( !empty($config_location_base) ) { + return $config_location_base . $path ; + } + // we computed it before, get it from the session: if ( sqgetGlobalVar('sq_base_url', $full_url, SQ_SESSION) ) { return $full_url . $path; } + // else: autodetect /* Check if this is a HTTPS or regular HTTP request. */ $proto = 'http://'; @@ -534,45 +496,78 @@ function get_location () { /* * If you have 'SSLOptions +StdEnvVars' in your apache config * OR if you have HTTPS=on in your HTTP_SERVER_VARS + * OR if you have HTTP_X_FORWARDED_PROTO=https in your HTTP_SERVER_VARS * OR if you are on port 443 */ $getEnvVar = getenv('HTTPS'); - if ((isset($getEnvVar) && !strcasecmp($getEnvVar, 'on')) || - (sqgetGlobalVar('HTTPS', $https_on, SQ_SERVER) && !strcasecmp($https_on, 'on')) || + if (!sqgetGlobalVar('HTTP_X_FORWARDED_PROTO', $forwarded_proto, SQ_SERVER)) + $forwarded_proto = ''; + if ((isset($getEnvVar) && strcasecmp($getEnvVar, 'on') === 0) || + (sqgetGlobalVar('HTTPS', $https_on, SQ_SERVER) && strcasecmp($https_on, 'on') === 0) || + (strcasecmp($forwarded_proto, 'https') === 0) || (sqgetGlobalVar('SERVER_PORT', $server_port, SQ_SERVER) && $server_port == 443)) { $proto = 'https://'; } /* Get the hostname from the Host header or server config. */ - if ( !sqgetGlobalVar('HTTP_HOST', $host, SQ_SERVER) || empty($host) ) { - if ( !sqgetGlobalVar('SERVER_NAME', $host, SQ_SERVER) || empty($host) ) { - $host = ''; - } + if ( !sqgetGlobalVar('HTTP_X_FORWARDED_HOST', $host, SQ_SERVER) || empty($host) ) { + if ( !sqgetGlobalVar('HTTP_HOST', $host, SQ_SERVER) || empty($host) ) { + if ( !sqgetGlobalVar('SERVER_NAME', $host, SQ_SERVER) || empty($host) ) { + $host = ''; + } + } } $port = ''; if (! strstr($host, ':')) { if (sqgetGlobalVar('SERVER_PORT', $server_port, SQ_SERVER)) { if (($server_port != 80 && $proto == 'http://') || - ($server_port != 443 && $proto == 'https://')) { + ($server_port != 443 && $proto == 'https://' && + strcasecmp($forwarded_proto, 'https') !== 0)) { $port = sprintf(':%d', $server_port); } } } - /* this is a workaround for the weird macosx caching that - causes Apache to return 16080 as the port number, which causes - SM to bail */ + /* this is a workaround for the weird macosx caching that + * causes Apache to return 16080 as the port number, which causes + * SM to bail */ - if ($imap_server_type == 'macosx' && $port == ':16080') { + if ($imap_server_type == 'macosx' && $port == ':16080') { $port = ''; - } + } + + /* Fallback is to omit the server name and use a relative */ + /* URI, although this is not RFC 2616 compliant. */ + $full_url = ($host ? $proto . $host . $port : ''); + sqsession_register($full_url, 'sq_base_url'); + return $full_url . $path; +} + + +/** + * Get Message List URI + * + * @param string $mailbox Current mailbox name (unencoded/raw) + * @param string $startMessage The mailbox page offset + * @param string $what Any current search parameters (OPTIONAL; + * default empty string) + * + * @return string The message list URI + * + * @since 1.5.2 + * + */ +function get_message_list_uri($mailbox, $startMessage, $what='') { - /* Fallback is to omit the server name and use a relative */ - /* URI, although this is not RFC 2616 compliant. */ - $full_url = ($host ? $proto . $host . $port : ''); - sqsession_register($full_url, 'sq_base_url'); - return $full_url . $path; + global $base_uri; + + $urlMailbox = urlencode($mailbox); + + $list_xtra = "?where=read_body.php&what=$what&mailbox=" . $urlMailbox. + "&startMessage=$startMessage"; + + return $base_uri .'src/right_main.php'. $list_xtra; } @@ -583,12 +578,23 @@ function get_location () { * stored in a cookie. The encryption key is generated by * OneTimePadCreate(); * - * @param string string the (password)string to encrypt - * @param string epad the encryption key + * @param string $string the (password)string to encrypt + * @param string $epad the encryption key * @return string the base64-encoded encrypted password + * @since 1.0 */ function OneTimePadEncrypt ($string, $epad) { $pad = base64_decode($epad); + + if (strlen($pad)>0) { + // make sure that pad is longer than string + while (strlen($string)>strlen($pad)) { + $pad.=$pad; + } + } else { + // FIXME: what should we do when $epad is not base64 encoded or empty. + } + $encrypted = ''; for ($i = 0; $i < strlen ($string); $i++) { $encrypted .= chr (ord($string[$i]) ^ ord($pad[$i])); @@ -603,12 +609,23 @@ function OneTimePadEncrypt ($string, $epad) { * Decrypts a password from the cookie, encrypted by OneTimePadEncrypt. * This uses the encryption key that is stored in the session. * - * @param string string the string to decrypt - * @param string epad the encryption key from the session + * @param string $string the string to decrypt + * @param string $epad the encryption key from the session * @return string the decrypted password + * @since 1.0 */ function OneTimePadDecrypt ($string, $epad) { $pad = base64_decode($epad); + + if (strlen($pad)>0) { + // make sure that pad is longer than string + while (strlen($string)>strlen($pad)) { + $pad.=$pad; + } + } else { + // FIXME: what should we do when $epad is not base64 encoded or empty. + } + $encrypted = base64_decode ($string); $decrypted = ''; for ($i = 0; $i < strlen ($encrypted); $i++) { @@ -626,8 +643,9 @@ function OneTimePadDecrypt ($string, $epad) { * appropriately. With strings, it is better to get them long. * Use md5() to lengthen smaller strings. * - * @param mixed val a value to seed the random number generator + * @param mixed $val a value to seed the random number generator. mixed = integer or string. * @return void + * @since 1.0 */ function sq_mt_seed($Val) { /* if mt_getrandmax() does not return a 2^n - 1 number, @@ -658,6 +676,7 @@ function sq_mt_seed($Val) { * the same 'random' numbers twice in one session. * * @return void + * @since 1.0 */ function sq_mt_randomize() { static $randomized; @@ -699,8 +718,10 @@ function sq_mt_randomize() { * Creates an encryption key for encrypting the password stored in the cookie. * The encryption key itself is stored in the session. * - * @param int length optional, length of the string to generate + * Pad must be longer or equal to encoded string length in 1.4.4/1.5.0 and older. + * @param int $length optional, length of the string to generate * @return string the encryption key + * @since 1.0 */ function OneTimePadCreate ($length=100) { sq_mt_randomize(); @@ -716,16 +737,17 @@ function OneTimePadCreate ($length=100) { /** * Returns a string showing the size of the message/attachment. * - * @param int bytes the filesize in bytes + * @param int $bytes the filesize in bytes * @return string the filesize in human readable format + * @since 1.0 */ function show_readable_size($bytes) { $bytes /= 1024; - $type = 'k'; + $type = _("KiB"); if ($bytes / 1024 > 1) { $bytes /= 1024; - $type = 'M'; + $type = _("MiB"); } if ($bytes < 10) { @@ -736,20 +758,21 @@ function show_readable_size($bytes) { settype($bytes, 'integer'); } - return $bytes . ' ' . $type . ''; + return $bytes . ' ' . $type; } /** * Generates a random string from the character set you pass in * - * @param int size the size of the string to generate - * @param string chars a string containing the characters to use - * @param int flags a flag to add a specific set to the characters to use: + * @param int $size the length of the string to generate + * @param string $chars a string containing the characters to use + * @param int $flags a flag to add a specific set to the characters to use: * Flags: * 1 = add lowercase a-z to $chars * 2 = add uppercase A-Z to $chars * 4 = add numbers 0-9 to $chars * @return string the random string + * @since 1.0 */ function GenerateRandomString($size, $chars, $flags = 0) { if ($flags & 0x1) { @@ -780,102 +803,60 @@ function GenerateRandomString($size, $chars, $flags = 0) { /** * Escapes special characters for use in IMAP commands. * - * @param string the string to escape + * @param string $str the string to escape * @return string the escaped string + * @since 1.0.3 */ function quoteimap($str) { return preg_replace("/([\"\\\\])/", "\\\\$1", $str); } -/** - * Trims array - * - * Trims every element in the array, ie. remove the first char of each element - * @param array array the array to trim - */ -function TrimArray(&$array) { - foreach ($array as $k => $v) { - global $$k; - if (is_array($$k)) { - foreach ($$k as $k2 => $v2) { - $$k[$k2] = substr($v2, 1); - } - } else { - $$k = substr($v, 1); - } - - /* Re-assign back to array. */ - $array[$k] = $$k; - } -} - /** * Create compose link * * Returns a link to the compose-page, taking in consideration * the compose_in_new and javascript settings. - * @param string url the URL to the compose page - * @param string text the link text, default "Compose" + * @param string $url the URL to the compose page + * @param string $text the link text, default "Compose" + * @param string $target (since 1.4.3) url target * @return string a link to the compose page + * @since 1.4.2 */ -function makeComposeLink($url, $text = null, $target='') -{ - global $compose_new_win,$javascript_on; +function makeComposeLink($url, $text = null, $target='') { + global $compose_new_win, $compose_width, + $compose_height, $oTemplate; if(!$text) { $text = _("Compose"); } - // if not using "compose in new window", make // regular link and be done with it if($compose_new_win != '1') { return makeInternalLink($url, $text, $target); } - // build the compose in new window link... // if javascript is on, use onclick event to handle it - if($javascript_on) { + if(checkForJavascript()) { sqgetGlobalVar('base_uri', $base_uri, SQ_SESSION); - return ''. $text.''; - } + $compuri = SM_BASE_URI.$url; + return create_hyperlink('javascript:void(0)', $text, '', "comp_in_new('$compuri','$compose_width','$compose_height')"); + } // otherwise, just open new window using regular HTML return makeInternalLink($url, $text, '_blank'); - -} - -/** - * Print variable - * - * sm_print_r($some_variable, [$some_other_variable [, ...]]); - * - * Debugging function - does the same as print_r, but makes sure special - * characters are converted to htmlentities first. This will allow - * values like to be displayed. - * The output is wrapped in <
> and <
> tags. - * - * @return void - */ -function sm_print_r() { - ob_start(); // Buffer output - foreach(func_get_args() as $var) { - print_r($var); - echo "\n"; - } - $buffer = ob_get_contents(); // Grab the print_r output - ob_end_clean(); // Silently discard the output & stop buffering - print '
';
-    print htmlentities($buffer);
-    print '
'; } /** * version of fwrite which checks for failure + * @param resource $fp + * @param string $string + * @return number of written bytes. false on failure + * @since 1.4.3 */ function sq_fwrite($fp, $string) { // write to file @@ -906,6 +887,7 @@ function sq_fwrite($fp, $string) { * * @param string $charset charset used for encoding. default to us-ascii, 'auto' uses $default_charset global value. * @return array html translation array + * @since 1.5.1 */ function sq_get_html_translation_table($table,$quote_style=ENT_COMPAT,$charset='us-ascii') { global $default_charset; @@ -955,7 +937,10 @@ function sq_get_html_translation_table($table,$quote_style=ENT_COMPAT,$charset=' * sq_htmlentities * * Convert all applicable characters to HTML entities. - * Minimal php requirement - v.4.0.5 + * Minimal php requirement - v.4.0.5. + * + * Function is designed for people that want to use full power of htmlentities() in + * i18n environment. * * @param string $string string that has to be sanitized * @param integer $quote_style quote encoding style. Possible values (without quotes): @@ -966,6 +951,7 @@ function sq_get_html_translation_table($table,$quote_style=ENT_COMPAT,$charset=' * * @param string $charset charset used for encoding. defaults to 'us-ascii', 'auto' uses $default_charset global value. * @return string sanitized string + * @since 1.5.1 */ function sq_htmlentities($string,$quote_style=ENT_COMPAT,$charset='us-ascii') { // get translation table @@ -983,7 +969,7 @@ function sq_htmlentities($string,$quote_style=ENT_COMPAT,$charset='us-ascii') { * @param string $string tested string * @param string $charset charset used in a string * @return bool true if 8bit symbols are detected - * @since 1.5.1 + * @since 1.5.1 and 1.4.4 */ function sq_is8bit($string,$charset='') { global $default_charset; @@ -992,7 +978,7 @@ function sq_is8bit($string,$charset='') { /** * Don't use \240 in ranges. Sometimes RH 7.2 doesn't like it. - * Don't use \200-\237 for iso-8859-x charsets. This ranges + * Don't use \200-\237 for iso-8859-x charsets. This range * stores control symbols in those charsets. * Use preg_match instead of ereg in order to avoid problems * with mbstring overloading @@ -1012,18 +998,25 @@ function sq_is8bit($string,$charset='') { * in php 5.x. Function does not test all mbstring encodings. Only the ones * that might be used in SM translations. * - * Supported arrays are stored in session in order to reduce number of + * Supported strings are stored in session in order to reduce number of * mb_internal_encoding function calls. * * If you want to test all mbstring encodings - fill $list_of_encodings * array. - * @return array list of encodings supported by mbstring - * @since 1.5.1 + * @return array list of encodings supported by php mbstring extension + * @since 1.5.1 and 1.4.6 */ function sq_mb_list_encodings() { if (! function_exists('mb_internal_encoding')) return array(); + // php 5+ function + if (function_exists('mb_list_encodings')) { + $ret = mb_list_encodings(); + array_walk($ret,'sq_lowercase_array_vals'); + return $ret; + } + // don't try to test encodings, if they are already stored in session if (sqgetGlobalVar('mb_supported_encodings',$mb_supported_encodings,SQ_SESSION)) return $mb_supported_encodings; @@ -1048,12 +1041,16 @@ function sq_mb_list_encodings() { 'koi8-u', 'big5', 'gb2312', + 'gb18030', 'windows-1251', 'windows-1255', 'windows-1256', 'tis-620', 'iso-2022-jp', + 'euc-cn', 'euc-kr', + 'euc-tw', + 'uhc', 'utf7-imap'); $supported_encodings=array(); @@ -1073,112 +1070,251 @@ function sq_mb_list_encodings() { return $supported_encodings; } +/** + * Callback function used to lowercase array values. + * @param string $val array value + * @param mixed $key array key + * @since 1.5.1 and 1.4.6 + */ +function sq_lowercase_array_vals(&$val,$key) { + $val = strtolower($val); +} + + /** * Function returns number of characters in string. * * Returned number might be different from number of bytes in string, - * if $charset is multibyte charset. Currently only utf-8 charset is - * supported. + * if $charset is multibyte charset. Detection depends on mbstring + * functions. If mbstring does not support tested multibyte charset, + * vanilla string length function is used. * @param string $str string * @param string $charset charset - * @since 1.5.1 - * @return integer number of characters in string + * @since 1.5.1 and 1.4.6 + * @return integer number of characters in string */ -function sq_strlen($str, $charset=''){ +function sq_strlen($str, $charset=null){ // default option - if ($charset=='') return strlen($str); + if (is_null($charset)) return strlen($str); + + // lowercase charset name + $charset=strtolower($charset); // use automatic charset detection, if function call asks for it if ($charset=='auto') { - global $default_charset; + global $default_charset, $squirrelmail_language; set_my_charset(); $charset=$default_charset; + if ($squirrelmail_language=='ja_JP') $charset='euc-jp'; } - // lowercase charset name - $charset=strtolower($charset); - - // set initial returned length number - $real_length=0; + // Use mbstring only with listed charsets + $aList_of_mb_charsets=array('utf-8','big5','gb2312','gb18030','euc-jp','euc-cn','euc-tw','euc-kr'); // calculate string length according to charset - // function can be modulized same way we modulize decode/encode/htmlentities - if ($charset=='utf-8') { - if (function_exists('mb_strlen')) { - $real_length = mb_strlen($str,'utf-8'); - } else { - // function needs length of string in bytes. - // mbstring overloading might break it - $str_length=strlen($str); - $str_index=0; - while ($str_index < $str_length) { - // start of utf-8 multibyte character detection - if (preg_match("/[\xC0-\xDF]/",$str[$str_index]) && - isset($str[$str_index+1]) && - preg_match("/[\x80-\xBF]/",$str[$str_index+1])) { - // two byte utf-8 - $str_index=$str_index+2; - $real_length++; - } elseif (preg_match("/[\xE0-\xEF]/",$str[$str_index]) && - isset($str[$str_index+2]) && - preg_match("/[\x80-\xBF][\x80-\xBF]/",$str[$str_index+1].$str[$str_index+2])) { - // three byte utf-8 - $str_index=$str_index+3; - $real_length++; - } elseif (preg_match("/[\xF0-\xF7]/",$str[$str_index]) && - isset($str[$str_index+3]) && - preg_match("/[\x80-\xBF][\x80-\xBF][\x80-\xBF]/",$str[$str_index+1].$str[$str_index+2].$str[$str_index+3])) { - // four byte utf-8 - $str_index=$str_index+4; - $real_length++; - } elseif (preg_match("/[\xF8-\xFB]/",$str[$str_index]) && - isset($str[$str_index+4]) && - preg_match("/[\x80-\xBF][\x80-\xBF][\x80-\xBF][\x80-\xBF]/", - $str[$str_index+1].$str[$str_index+2].$str[$str_index+3].$str[$str_index+4])) { - // five byte utf-8 - $str_index=$str_index+5; - $real_length++; - } elseif (preg_match("/[\xFC-\xFD]/",$str[$str_index]) && - isset($str[$str_index+5]) && - preg_match("/[\x80-\xBF][\x80-\xBF][\x80-\xBF][\x80-\xBF]/", - $str[$str_index+1].$str[$str_index+2].$str[$str_index+3].$str[$str_index+4].$str[$str_index+5])) { - // six byte utf-8 - $str_index=$str_index+6; - $real_length++; - } else { - $str_index++; - $real_length++; - } - // end of utf-8 multibyte character detection - } - } - // end of utf-8 length detection - } elseif ($charset=='big5') { - // TODO: add big5 string length detection - $real_length=strlen($str); - } elseif ($charset=='gb2312') { - // TODO: add gb2312 string length detection - $real_length=strlen($str); - } elseif ($charset=='gb18030') { - // TODO: add gb18030 string length detection - $real_length=strlen($str); - } elseif ($charset=='euc-jp') { - // TODO: add euc-jp string length detection - $real_length=strlen($str); - } elseif ($charset=='euc-cn') { - // TODO: add euc-cn string length detection - $real_length=strlen($str); - } elseif ($charset=='euc-tw') { - // TODO: add euc-tw string length detection - $real_length=strlen($str); - } elseif ($charset=='euc-kr') { - // TODO: add euc-kr string length detection - $real_length=strlen($str); + if (in_array($charset,$aList_of_mb_charsets) && in_array($charset,sq_mb_list_encodings())) { + $real_length = mb_strlen($str,$charset); } else { + // own strlen detection code is removed because missing strpos, + // strtoupper and substr implementations break string wrapping. $real_length=strlen($str); } return $real_length; } -$PHP_SELF = php_self(); -?> \ No newline at end of file +/** + * string padding with multibyte support + * + * @link http://www.php.net/str_pad + * @param string $string original string + * @param integer $width padded string width + * @param string $pad padding symbols + * @param integer $padtype padding type + * (internal php defines, see str_pad() description) + * @param string $charset charset used in original string + * @return string padded string + */ +function sq_str_pad($string, $width, $pad, $padtype, $charset='') { + + $charset = strtolower($charset); + $padded_string = ''; + + switch ($charset) { + case 'utf-8': + case 'big5': + case 'gb2312': + case 'euc-kr': + /* + * all multibyte charsets try to increase width value by + * adding difference between number of bytes and real length + */ + $width = $width - sq_strlen($string,$charset) + strlen($string); + default: + $padded_string=str_pad($string,$width,$pad,$padtype); + } + return $padded_string; +} + +/** + * Wrapper that is used to switch between vanilla and multibyte substr + * functions. + * @param string $string + * @param integer $start + * @param integer $length + * @param string $charset + * @return string + * @since 1.5.1 + * @link http://www.php.net/substr + * @link http://www.php.net/mb_substr + */ +function sq_substr($string,$start,$length,$charset='auto') { + // use automatic charset detection, if function call asks for it + static $charset_auto, $bUse_mb; + + if ($charset=='auto') { + if (!isset($charset_auto)) { + global $default_charset, $squirrelmail_language; + set_my_charset(); + $charset=$default_charset; + if ($squirrelmail_language=='ja_JP') $charset='euc-jp'; + $charset_auto = $charset; + } else { + $charset = $charset_auto; + } + } + $charset = strtolower($charset); + + // in_array call is expensive => do it once and use a static var for + // storing the results + if (!isset($bUse_mb)) { + if (in_array($charset,sq_mb_list_encodings())) { + $bUse_mb = true; + } else { + $bUse_mb = false; + } + } + + if ($bUse_mb) { + return mb_substr($string,$start,$length,$charset); + } + // TODO: add mbstring independent code + + // use vanilla string functions as last option + return substr($string,$start,$length); +} + +/** + * Wrapper that is used to switch between vanilla and multibyte strpos + * functions. + * @param string $haystack + * @param mixed $needle + * @param integer $offset + * @param string $charset + * @return string + * @since 1.5.1 + * @link http://www.php.net/strpos + * @link http://www.php.net/mb_strpos + */ +function sq_strpos($haystack,$needle,$offset,$charset='auto') { + // use automatic charset detection, if function call asks for it + static $charset_auto, $bUse_mb; + + if ($charset=='auto') { + if (!isset($charset_auto)) { + global $default_charset, $squirrelmail_language; + set_my_charset(); + $charset=$default_charset; + if ($squirrelmail_language=='ja_JP') $charset='euc-jp'; + $charset_auto = $charset; + } else { + $charset = $charset_auto; + } + } + $charset = strtolower($charset); + + // in_array call is expensive => do it once and use a static var for + // storing the results + if (!isset($bUse_mb)) { + if (in_array($charset,sq_mb_list_encodings())) { + $bUse_mb = true; + } else { + $bUse_mb = false; + } + } + if ($bUse_mb) { + return mb_strpos($haystack,$needle,$offset,$charset); + } + // TODO: add mbstring independent code + + // use vanilla string functions as last option + return strpos($haystack,$needle,$offset); +} + +/** + * Wrapper that is used to switch between vanilla and multibyte strtoupper + * functions. + * @param string $string + * @param string $charset + * @return string + * @since 1.5.1 + * @link http://www.php.net/strtoupper + * @link http://www.php.net/mb_strtoupper + */ +function sq_strtoupper($string,$charset='auto') { + // use automatic charset detection, if function call asks for it + static $charset_auto, $bUse_mb; + + if ($charset=='auto') { + if (!isset($charset_auto)) { + global $default_charset, $squirrelmail_language; + set_my_charset(); + $charset=$default_charset; + if ($squirrelmail_language=='ja_JP') $charset='euc-jp'; + $charset_auto = $charset; + } else { + $charset = $charset_auto; + } + } + $charset = strtolower($charset); + + // in_array call is expensive => do it once and use a static var for + // storing the results + if (!isset($bUse_mb)) { + if (function_exists('mb_strtoupper') && + in_array($charset,sq_mb_list_encodings())) { + $bUse_mb = true; + } else { + $bUse_mb = false; + } + } + + if ($bUse_mb) { + return mb_strtoupper($string,$charset); + } + // TODO: add mbstring independent code + + // use vanilla string functions as last option + return strtoupper($string); +} + +/** + * Counts 8bit bytes in string + * @param string $string tested string + * @return integer number of 8bit bytes + */ +function sq_count8bit($string) { + $count=0; + for ($i=0; $i 127) $count++; + } + return $count; +} + +/** + * Callback function to trim whitespace from a value, to be used in array_walk + * @param string $value value to trim + * @since 1.5.2 and 1.4.7 + */ +function sq_trim_value ( &$value ) { + $value = trim($value); +}