From 53604a0550f9940584b7e4d4260b96714ae0edbf Mon Sep 17 00:00:00 2001
From: alecpl <alec@alec.pl>
Date: Wed, 01 Dec 2010 05:49:20 -0500
Subject: [PATCH] - Fix setting charset of attachment filenames (#1487122)
---
program/lib/Mail/mimePart.php | 506 ++++++++++++++++++++++++++++++++++++++++++++++++-------
1 files changed, 437 insertions(+), 69 deletions(-)
diff --git a/program/lib/Mail/mimePart.php b/program/lib/Mail/mimePart.php
index 7863028..60b3601 100644
--- a/program/lib/Mail/mimePart.php
+++ b/program/lib/Mail/mimePart.php
@@ -141,17 +141,19 @@
* content_type - The content type for this part eg multipart/mixed
* encoding - The encoding to use, 7bit, 8bit,
* base64, or quoted-printable
+ * charset - Content character set
* cid - Content ID to apply
* disposition - Content disposition, inline or attachment
* dfilename - Filename parameter for content disposition
* description - Content description
- * charset - Character set to use
- * name_encoding - Encoding for attachment name (Content-Type)
+ * name_encoding - Encoding of the attachment name (Content-Type)
* By default filenames are encoded using RFC2231
* Here you can set RFC2047 encoding (quoted-printable
* or base64) instead
- * filename_encoding - Encoding for attachment filename (Content-Disposition)
+ * filename_encoding - Encoding of the attachment filename (Content-Disposition)
* See 'name_encoding'
+ * headers_charset - Charset of the headers e.g. filename, description.
+ * If not set, 'charset' will be used
* eol - End of line sequence. Default: "\r\n"
* body_file - Location of file with part's body (instead of $body)
*
@@ -165,14 +167,8 @@
$this->_eol = MAIL_MIMEPART_CRLF;
}
- $c_type = array();
- $c_disp = array();
foreach ($params as $key => $value) {
switch ($key) {
- case 'content_type':
- $c_type['type'] = $value;
- break;
-
case 'encoding':
$this->_encoding = $value;
$headers['Content-Transfer-Encoding'] = $value;
@@ -180,29 +176,6 @@
case 'cid':
$headers['Content-ID'] = '<' . $value . '>';
- break;
-
- case 'disposition':
- $c_disp['disp'] = $value;
- break;
-
- case 'dfilename':
- $c_disp['filename'] = $value;
- $c_type['name'] = $value;
- break;
-
- case 'description':
- $headers['Content-Description'] = $value;
- break;
-
- case 'charset':
- $c_type['charset'] = $value;
- $c_disp['charset'] = $value;
- break;
-
- case 'language':
- $c_type['language'] = $value;
- $c_disp['language'] = $value;
break;
case 'location':
@@ -215,41 +188,59 @@
}
}
+ // Default content-type
+ if (empty($params['content_type'])) {
+ $params['content_type'] = 'text/plain';
+ }
+
// Content-Type
- if (isset($c_type['type'])) {
- $headers['Content-Type'] = $c_type['type'];
- if (isset($c_type['name'])) {
- $headers['Content-Type'] .= ';' . $this->_eol;
- $headers['Content-Type'] .= $this->_buildHeaderParam(
- 'name', $c_type['name'],
- isset($c_type['charset']) ? $c_type['charset'] : 'US-ASCII',
- isset($c_type['language']) ? $c_type['language'] : null,
- isset($params['name_encoding']) ? $params['name_encoding'] : null
- );
+ $headers['Content-Type'] = $params['content_type'];
+ if (!empty($params['charset'])) {
+ $charset = "charset={$params['charset']}";
+ // place charset parameter in the same line, if possible
+ if ((strlen($headers['Content-Type']) + strlen($charset) + 16) <= 76) {
+ $headers['Content-Type'] .= '; ';
+ } else {
+ $headers['Content-Type'] .= ';' . $this->_eol . ' ';
}
- if (isset($c_type['charset'])) {
- $headers['Content-Type']
- .= ';' . $this->_eol . " charset={$c_type['charset']}";
+ $headers['Content-Type'] .= $charset;
+
+ // Default headers charset
+ if (!isset($params['headers_charset'])) {
+ $params['headers_charset'] = $params['charset'];
}
+ }
+ if (!empty($params['filename'])) {
+ $headers['Content-Type'] .= ';' . $this->_eol;
+ $headers['Content-Type'] .= $this->_buildHeaderParam(
+ 'name', $params['filename'],
+ !empty($params['headers_charset']) ? $params['headers_charset'] : 'US-ASCII',
+ !empty($params['language']) ? $params['language'] : null,
+ !empty($params['name_encoding']) ? $params['name_encoding'] : null
+ );
}
// Content-Disposition
- if (isset($c_disp['disp'])) {
- $headers['Content-Disposition'] = $c_disp['disp'];
- if (isset($c_disp['filename'])) {
+ if (!empty($params['disposition'])) {
+ $headers['Content-Disposition'] = $params['disposition'];
+ if (!empty($params['filename'])) {
$headers['Content-Disposition'] .= ';' . $this->_eol;
$headers['Content-Disposition'] .= $this->_buildHeaderParam(
- 'filename', $c_disp['filename'],
- isset($c_disp['charset']) ? $c_disp['charset'] : 'US-ASCII',
- isset($c_disp['language']) ? $c_disp['language'] : null,
- isset($params['filename_encoding']) ? $params['filename_encoding'] : null
+ 'filename', $params['filename'],
+ !empty($params['headers_charset']) ? $params['headers_charset'] : 'US-ASCII',
+ !empty($params['language']) ? $params['language'] : null,
+ !empty($params['filename_encoding']) ? $params['filename_encoding'] : null
);
}
}
- // Default content-type
- if (!isset($headers['Content-Type'])) {
- $headers['Content-Type'] = 'text/plain';
+ if (!empty($params['description'])) {
+ $headers['Content-Description'] = $this->encodeHeader(
+ 'Content-Description', $params['description'],
+ !empty($params['headers_charset']) ? $params['headers_charset'] : 'US-ASCII',
+ !empty($params['name_encoding']) ? $params['name_encoding'] : 'quoted-printable',
+ $this->_eol
+ );
}
// Default encoding
@@ -769,17 +760,275 @@
}
/**
- * Callback function to replace extended characters (\x80-xFF) with their
- * ASCII values (RFC2231)
+ * Encodes a header as per RFC2047
*
- * @param array $matches Preg_replace's matches array
+ * @param string $name The header name
+ * @param string $value The header data to encode
+ * @param string $charset Character set name
+ * @param string $encoding Encoding name (base64 or quoted-printable)
+ * @param string $eol End-of-line sequence. Default: "\r\n"
*
- * @return string Encoded character string
+ * @return string Encoded header data (without a name)
+ * @access public
+ * @since 1.6.1
+ */
+ function encodeHeader($name, $value, $charset='ISO-8859-1',
+ $encoding='quoted-printable', $eol="\r\n"
+ ) {
+ // Structured headers
+ $comma_headers = array(
+ 'from', 'to', 'cc', 'bcc', 'sender', 'reply-to',
+ 'resent-from', 'resent-to', 'resent-cc', 'resent-bcc',
+ 'resent-sender', 'resent-reply-to',
+ 'return-receipt-to', 'disposition-notification-to',
+ );
+ $other_headers = array(
+ 'references', 'in-reply-to', 'message-id', 'resent-message-id',
+ );
+
+ $name = strtolower($name);
+
+ if (in_array($name, $comma_headers)) {
+ $separator = ',';
+ } else if (in_array($name, $other_headers)) {
+ $separator = ' ';
+ }
+
+ if (!$charset) {
+ $charset = 'ISO-8859-1';
+ }
+
+ // Structured header (make sure addr-spec inside is not encoded)
+ if (!empty($separator)) {
+ $parts = Mail_mimePart::_explodeQuotedString($separator, $value);
+ $value = '';
+
+ foreach ($parts as $part) {
+ $part = preg_replace('/\r?\n[\s\t]*/', $eol . ' ', $part);
+ $part = trim($part);
+
+ if (!$part) {
+ continue;
+ }
+ if ($value) {
+ $value .= $separator==',' ? $separator.' ' : ' ';
+ } else {
+ $value = $name . ': ';
+ }
+
+ // let's find phrase (name) and/or addr-spec
+ if (preg_match('/^<\S+@\S+>$/', $part)) {
+ $value .= $part;
+ } else if (preg_match('/^\S+@\S+$/', $part)) {
+ // address without brackets and without name
+ $value .= $part;
+ } else if (preg_match('/<*\S+@\S+>*$/', $part, $matches)) {
+ // address with name (handle name)
+ $address = $matches[0];
+ $word = str_replace($address, '', $part);
+ $word = trim($word);
+ // check if phrase requires quoting
+ if ($word) {
+ // non-ASCII: require encoding
+ if (preg_match('#([\x80-\xFF]){1}#', $word)) {
+ if ($word[0] == '"' && $word[strlen($word)-1] == '"') {
+ // de-quote quoted-string, encoding changes
+ // string to atom
+ $search = array("\\\"", "\\\\");
+ $replace = array("\"", "\\");
+ $word = str_replace($search, $replace, $word);
+ $word = substr($word, 1, -1);
+ }
+ // find length of last line
+ if (($pos = strrpos($value, $eol)) !== false) {
+ $last_len = strlen($value) - $pos;
+ } else {
+ $last_len = strlen($value);
+ }
+ $word = Mail_mimePart::encodeHeaderValue(
+ $word, $charset, $encoding, $last_len, $eol
+ );
+ } else if (($word[0] != '"' || $word[strlen($word)-1] != '"')
+ && preg_match('/[\(\)\<\>\\\.\[\]@,;:"]/', $word)
+ ) {
+ // ASCII: quote string if needed
+ $word = '"'.addcslashes($word, '\\"').'"';
+ }
+ }
+ $value .= $word.' '.$address;
+ } else {
+ // addr-spec not found, don't encode (?)
+ $value .= $part;
+ }
+
+ // RFC2822 recommends 78 characters limit, use 76 from RFC2047
+ $value = wordwrap($value, 76, $eol . ' ');
+ }
+
+ // remove header name prefix (there could be EOL too)
+ $value = preg_replace(
+ '/^'.$name.':('.preg_quote($eol, '/').')* /', '', $value
+ );
+
+ } else {
+ // Unstructured header
+ // non-ASCII: require encoding
+ if (preg_match('#([\x80-\xFF]){1}#', $value)) {
+ if ($value[0] == '"' && $value[strlen($value)-1] == '"') {
+ // de-quote quoted-string, encoding changes
+ // string to atom
+ $search = array("\\\"", "\\\\");
+ $replace = array("\"", "\\");
+ $value = str_replace($search, $replace, $value);
+ $value = substr($value, 1, -1);
+ }
+ $value = Mail_mimePart::encodeHeaderValue(
+ $value, $charset, $encoding, strlen($name) + 2, $eol
+ );
+ } else if (strlen($name.': '.$value) > 78) {
+ // ASCII: check if header line isn't too long and use folding
+ $value = preg_replace('/\r?\n[\s\t]*/', $eol . ' ', $value);
+ $tmp = wordwrap($name.': '.$value, 78, $eol . ' ');
+ $value = preg_replace('/^'.$name.':\s*/', '', $tmp);
+ // hard limit 998 (RFC2822)
+ $value = wordwrap($value, 998, $eol . ' ', true);
+ }
+ }
+
+ return $value;
+ }
+
+ /**
+ * Explode quoted string
+ *
+ * @param string $delimiter Delimiter expression string for preg_match()
+ * @param string $string Input string
+ *
+ * @return array String tokens array
* @access private
*/
- function _encodeReplaceCallback($matches)
+ function _explodeQuotedString($delimiter, $string)
{
- return sprintf('%%%02X', ord($matches[1]));
+ $result = array();
+ $strlen = strlen($string);
+
+ for ($q=$p=$i=0; $i < $strlen; $i++) {
+ if ($string[$i] == "\""
+ && (empty($string[$i-1]) || $string[$i-1] != "\\")
+ ) {
+ $q = $q ? false : true;
+ } else if (!$q && preg_match("/$delimiter/", $string[$i])) {
+ $result[] = substr($string, $p, $i - $p);
+ $p = $i + 1;
+ }
+ }
+
+ $result[] = substr($string, $p);
+ return $result;
+ }
+
+ /**
+ * Encodes a header value as per RFC2047
+ *
+ * @param string $value The header data to encode
+ * @param string $charset Character set name
+ * @param string $encoding Encoding name (base64 or quoted-printable)
+ * @param int $prefix_len Prefix length. Default: 0
+ * @param string $eol End-of-line sequence. Default: "\r\n"
+ *
+ * @return string Encoded header data
+ * @access public
+ * @since 1.6.1
+ */
+ function encodeHeaderValue($value, $charset, $encoding, $prefix_len=0, $eol="\r\n")
+ {
+ // #17311: Use multibyte aware method (requires mbstring extension)
+ if ($result = Mail_mimePart::encodeMB($value, $charset, $encoding, $prefix_len, $eol)) {
+ return $result;
+ }
+
+ // Generate the header using the specified params and dynamicly
+ // determine the maximum length of such strings.
+ // 75 is the value specified in the RFC.
+ $encoding = $encoding == 'base64' ? 'B' : 'Q';
+ $prefix = '=?' . $charset . '?' . $encoding .'?';
+ $suffix = '?=';
+ $maxLength = 75 - strlen($prefix . $suffix);
+ $maxLength1stLine = $maxLength - $prefix_len;
+
+ if ($encoding == 'B') {
+ // Base64 encode the entire string
+ $value = base64_encode($value);
+
+ // We can cut base64 every 4 characters, so the real max
+ // we can get must be rounded down.
+ $maxLength = $maxLength - ($maxLength % 4);
+ $maxLength1stLine = $maxLength1stLine - ($maxLength1stLine % 4);
+
+ $cutpoint = $maxLength1stLine;
+ $output = '';
+
+ while ($value) {
+ // Split translated string at every $maxLength
+ $part = substr($value, 0, $cutpoint);
+ $value = substr($value, $cutpoint);
+ $cutpoint = $maxLength;
+ // RFC 2047 specifies that any split header should
+ // be seperated by a CRLF SPACE.
+ if ($output) {
+ $output .= $eol . ' ';
+ }
+ $output .= $prefix . $part . $suffix;
+ }
+ $value = $output;
+ } else {
+ // quoted-printable encoding has been selected
+ $value = Mail_mimePart::encodeQP($value);
+
+ // This regexp will break QP-encoded text at every $maxLength
+ // but will not break any encoded letters.
+ $reg1st = "|(.{0,$maxLength1stLine}[^\=][^\=])|";
+ $reg2nd = "|(.{0,$maxLength}[^\=][^\=])|";
+
+ if (strlen($value) > $maxLength1stLine) {
+ // Begin with the regexp for the first line.
+ $reg = $reg1st;
+ $output = '';
+ while ($value) {
+ // Split translated string at every $maxLength
+ // But make sure not to break any translated chars.
+ $found = preg_match($reg, $value, $matches);
+
+ // After this first line, we need to use a different
+ // regexp for the first line.
+ $reg = $reg2nd;
+
+ // Save the found part and encapsulate it in the
+ // prefix & suffix. Then remove the part from the
+ // $value_out variable.
+ if ($found) {
+ $part = $matches[0];
+ $len = strlen($matches[0]);
+ $value = substr($value, $len);
+ } else {
+ $part = $value;
+ $value = '';
+ }
+
+ // RFC 2047 specifies that any split header should
+ // be seperated by a CRLF SPACE
+ if ($output) {
+ $output .= $eol . ' ';
+ }
+ $output .= $prefix . $part . $suffix;
+ }
+ $value = $output;
+ } else {
+ $value = $prefix . $value . $suffix;
+ }
+ }
+
+ return $value;
}
/**
@@ -793,16 +1042,121 @@
*/
function encodeQP($str)
{
- // Replace all special characters used by the encoder
- $search = array('=', '_', '?', ' ');
- $replace = array('=3D', '=5F', '=3F', '_');
- $str = str_replace($search, $replace, $str);
+ // Bug #17226 RFC 2047 restricts some characters
+ // if the word is inside a phrase, permitted chars are only:
+ // ASCII letters, decimal digits, "!", "*", "+", "-", "/", "=", and "_"
- // Replace all extended characters (\x80-xFF) with their
- // ASCII values.
- return preg_replace_callback(
- '/([\x80-\xFF])/', array('Mail_mimePart', '_qpReplaceCallback'), $str
+ // "=", "_", "?" must be encoded
+ $regexp = '/([\x22-\x29\x2C\x2E\x3A-\x40\x5B-\x60\x7B-\x7E\x80-\xFF])/';
+ $str = preg_replace_callback(
+ $regexp, array('Mail_mimePart', '_qpReplaceCallback'), $str
);
+
+ return str_replace(' ', '_', $str);
+ }
+
+ /**
+ * Encodes the given string using base64 or quoted-printable.
+ * This method makes sure that encoded-word represents an integral
+ * number of characters as per RFC2047.
+ *
+ * @param string $str String to encode
+ * @param string $charset Character set name
+ * @param string $encoding Encoding name (base64 or quoted-printable)
+ * @param int $prefix_len Prefix length. Default: 0
+ * @param string $eol End-of-line sequence. Default: "\r\n"
+ *
+ * @return string Encoded string
+ * @access public
+ * @since 1.8.0
+ */
+ function encodeMB($str, $charset, $encoding, $prefix_len=0, $eol="\r\n")
+ {
+ if (!function_exists('mb_substr') || !function_exists('mb_strlen')) {
+ return;
+ }
+
+ $encoding = $encoding == 'base64' ? 'B' : 'Q';
+ // 75 is the value specified in the RFC
+ $prefix = '=?' . $charset . '?'.$encoding.'?';
+ $suffix = '?=';
+ $maxLength = 75 - strlen($prefix . $suffix);
+
+ // A multi-octet character may not be split across adjacent encoded-words
+ // So, we'll loop over each character
+ // mb_stlen() with wrong charset will generate a warning here and return null
+ $length = mb_strlen($str, $charset);
+ $result = '';
+ $line_length = $prefix_len;
+
+ if ($encoding == 'B') {
+ // base64
+ $start = 0;
+ $prev = '';
+
+ for ($i=1; $i<=$length; $i++) {
+ // See #17311
+ $chunk = mb_substr($str, $start, $i-$start, $charset);
+ $chunk = base64_encode($chunk);
+ $chunk_len = strlen($chunk);
+
+ if ($line_length + $chunk_len == $maxLength || $i == $length) {
+ if ($result) {
+ $result .= "\n";
+ }
+ $result .= $chunk;
+ $line_length = 0;
+ $start = $i;
+ } else if ($line_length + $chunk_len > $maxLength) {
+ if ($result) {
+ $result .= "\n";
+ }
+ if ($prev) {
+ $result .= $prev;
+ }
+ $line_length = 0;
+ $start = $i - 1;
+ } else {
+ $prev = $chunk;
+ }
+ }
+ } else {
+ // quoted-printable
+ // see encodeQP()
+ $regexp = '/([\x22-\x29\x2C\x2E\x3A-\x40\x5B-\x60\x7B-\x7E\x80-\xFF])/';
+
+ for ($i=0; $i<=$length; $i++) {
+ $char = mb_substr($str, $i, 1, $charset);
+ // RFC recommends underline (instead of =20) in place of the space
+ // that's one of the reasons why we're not using iconv_mime_encode()
+ if ($char == ' ') {
+ $char = '_';
+ $char_len = 1;
+ } else {
+ $char = preg_replace_callback(
+ $regexp, array('Mail_mimePart', '_qpReplaceCallback'), $char
+ );
+ $char_len = strlen($char);
+ }
+
+ if ($line_length + $char_len > $maxLength) {
+ if ($result) {
+ $result .= "\n";
+ }
+ $line_length = 0;
+ }
+
+ $result .= $char;
+ $line_length += $char_len;
+ }
+ }
+
+ if ($result) {
+ $result = $prefix
+ .str_replace("\n", $suffix.$eol.' '.$prefix, $result).$suffix;
+ }
+
+ return $result;
}
/**
@@ -819,4 +1173,18 @@
return sprintf('=%02X', ord($matches[1]));
}
+ /**
+ * Callback function to replace extended characters (\x80-xFF) with their
+ * ASCII values (RFC2231)
+ *
+ * @param array $matches Preg_replace's matches array
+ *
+ * @return string Encoded character string
+ * @access private
+ */
+ function _encodeReplaceCallback($matches)
+ {
+ return sprintf('%%%02X', ord($matches[1]));
+ }
+
} // End of class
--
Gitblit v1.9.1