diff --git a/phpseclib/File/ASN1.php b/phpseclib/File/ASN1.php index 2774e7e8..4bdf13ea 100644 --- a/phpseclib/File/ASN1.php +++ b/phpseclib/File/ASN1.php @@ -272,7 +272,8 @@ class File_ASN1 } $this->encoded = $encoded; - return $this->_decode_ber($encoded); + // encapsulate in an array for BC with the old decodeBER + return array($this->_decode_ber($encoded)); } /** @@ -287,215 +288,233 @@ class File_ASN1 * @return Array * @access private */ - function _decode_ber(&$encoded, $start = 0) + function _decode_ber($encoded, $start = 0) { - $decoded = array(); + $current = array('start' => $start); - while ( strlen($encoded) ) { - $current = array('start' => $start); + $type = ord($this->_string_shift($encoded)); + $start++; - $type = ord($this->_string_shift($encoded)); - $start++; + $constructed = ($type >> 5) & 1; - $constructed = ($type >> 5) & 1; - - $tag = $type & 0x1F; - if ($tag == 0x1F) { - $tag = 0; - // process septets (since the eighth bit is ignored, it's not an octet) - do { - $loop = ord($encoded[0]) >> 7; - $tag <<= 7; - $tag |= ord($this->_string_shift($encoded)) & 0x7F; - $start++; - } while ( $loop ); - } - - // Length, as discussed in paragraph 8.1.3 of X.690-0207.pdf#page=13 - $length = ord($this->_string_shift($encoded)); - $start++; - if ( $length == 0x80 ) { // indefinite length - // "[A sender shall] use the indefinite form (see 8.1.3.6) if the encoding is constructed and is not all - // immediately available." -- paragraph 8.1.3.2.c - //if ( !$constructed ) { - // return false; - //} - $length = strlen($encoded); - } elseif ( $length & 0x80 ) { // definite length, long form - // technically, the long form of the length can be represented by up to 126 octets (bytes), but we'll only - // support it up to four. - $length&= 0x7F; - $temp = $this->_string_shift($encoded, $length); - // tags of indefinite length don't really have a header length; this length includes the tag - $current+= array('headerlength' => $length + 2); - $start+= $length; - extract(unpack('Nlength', substr(str_pad($temp, 4, chr(0), STR_PAD_LEFT), -4))); - } else { - $current+= array('headerlength' => 2); - } - - // End-of-content, see paragraphs 8.1.1.3, 8.1.3.2, 8.1.3.6, 8.1.5, and (for an example) 8.6.4.2 - if (!$type && !$length) { - return $decoded; - } - $content = $this->_string_shift($encoded, $length); - - /* Class is UNIVERSAL, APPLICATION, PRIVATE, or CONTEXT-SPECIFIC. The UNIVERSAL class is restricted to the ASN.1 - built-in types. It defines an application-independent data type that must be distinguishable from all other - data types. The other three classes are user defined. The APPLICATION class distinguishes data types that - have a wide, scattered use within a particular presentation context. PRIVATE distinguishes data types within - a particular organization or country. CONTEXT-SPECIFIC distinguishes members of a sequence or set, the - alternatives of a CHOICE, or universally tagged set members. Only the class number appears in braces for this - data type; the term CONTEXT-SPECIFIC does not appear. - - -- http://www.obj-sys.com/asn1tutorial/node12.html */ - $class = ($type >> 6) & 3; - switch ($class) { - case FILE_ASN1_CLASS_APPLICATION: - case FILE_ASN1_CLASS_PRIVATE: - case FILE_ASN1_CLASS_CONTEXT_SPECIFIC: - $decoded[] = array( - 'type' => $class, - 'constant' => $tag, - 'content' => $constructed ? $this->_decode_ber($content, $start) : $content, - 'length' => $length + $start - $current['start'] - ) + $current; - $start+= $length; - continue 2; - } - - $current+= array('type' => $tag); - - // decode UNIVERSAL tags - switch ($tag) { - case FILE_ASN1_TYPE_BOOLEAN: - // "The contents octets shall consist of a single octet." -- paragraph 8.2.1 - //if (strlen($content) != 1) { - // return false; - //} - $current['content'] = (bool) ord($content[0]); - break; - case FILE_ASN1_TYPE_INTEGER: - case FILE_ASN1_TYPE_ENUMERATED: - $current['content'] = new Math_BigInteger($content, -256); - break; - case FILE_ASN1_TYPE_REAL: // not currently supported - return false; - case FILE_ASN1_TYPE_BIT_STRING: - // The initial octet shall encode, as an unsigned binary integer with bit 1 as the least significant bit, - // the number of unused bits in the final subsequent octet. The number shall be in the range zero to - // seven. - if (!$constructed) { - $current['content'] = $content; - } else { - $temp = $this->_decode_ber($content, $start); - $length-= strlen($content); - $last = count($temp) - 1; - for ($i = 0; $i < $last; $i++) { - // all subtags should be bit strings - //if ($temp[$i]['type'] != FILE_ASN1_TYPE_BIT_STRING) { - // return false; - //} - $current['content'].= substr($temp[$i]['content'], 1); - } - // all subtags should be bit strings - //if ($temp[$last]['type'] != FILE_ASN1_TYPE_BIT_STRING) { - // return false; - //} - $current['content'] = $temp[$last]['content'][0] . $current['content'] . substr($temp[$i]['content'], 1); - } - break; - case FILE_ASN1_TYPE_OCTET_STRING: - if (!$constructed) { - $current['content'] = $content; - } else { - $temp = $this->_decode_ber($content, $start); - $length-= strlen($content); - for ($i = 0, $size = count($temp); $i < $size; $i++) { - // all subtags should be octet strings - //if ($temp[$i]['type'] != FILE_ASN1_TYPE_OCTET_STRING) { - // return false; - //} - $current['content'].= $temp[$i]['content']; - } - // $length = - } - break; - case FILE_ASN1_TYPE_NULL: - // "The contents octets shall not contain any octets." -- paragraph 8.8.2 - //if (strlen($content)) { - // return false; - //} - break; - case FILE_ASN1_TYPE_SEQUENCE: - case FILE_ASN1_TYPE_SET: - $current['content'] = $this->_decode_ber($content, $start); - break; - case FILE_ASN1_TYPE_OBJECT_IDENTIFIER: - $temp = ord($this->_string_shift($content)); - $current['content'] = sprintf('%d.%d', floor($temp / 40), $temp % 40); - $valuen = 0; - // process septets - while (strlen($content)) { - $temp = ord($this->_string_shift($content)); - $valuen <<= 7; - $valuen |= $temp & 0x7F; - if (~$temp & 0x80) { - $current['content'].= ".$valuen"; - $valuen = 0; - } - } - // the eighth bit of the last byte should not be 1 - //if ($temp >> 7) { - // return false; - //} - break; - /* Each character string type shall be encoded as if it had been declared: - [UNIVERSAL x] IMPLICIT OCTET STRING - - -- X.690-0207.pdf#page=23 (paragraph 8.21.3) - - Per that, we're not going to do any validation. If there are any illegal characters in the string, - we don't really care */ - case FILE_ASN1_TYPE_NUMERIC_STRING: - // 0,1,2,3,4,5,6,7,8,9, and space - case FILE_ASN1_TYPE_PRINTABLE_STRING: - // Upper and lower case letters, digits, space, apostrophe, left/right parenthesis, plus sign, comma, - // hyphen, full stop, solidus, colon, equal sign, question mark - case FILE_ASN1_TYPE_TELETEX_STRING: - // The Teletex character set in CCITT's T61, space, and delete - // see http://en.wikipedia.org/wiki/Teletex#Character_sets - case FILE_ASN1_TYPE_VIDEOTEX_STRING: - // The Videotex character set in CCITT's T.100 and T.101, space, and delete - case FILE_ASN1_TYPE_VISIBLE_STRING: - // Printing character sets of international ASCII, and space - case FILE_ASN1_TYPE_IA5_STRING: - // International Alphabet 5 (International ASCII) - case FILE_ASN1_TYPE_GRAPHIC_STRING: - // All registered G sets, and space - case FILE_ASN1_TYPE_GENERAL_STRING: - // All registered C and G sets, space and delete - case FILE_ASN1_TYPE_UTF8_STRING: - // ???? - case FILE_ASN1_TYPE_BMP_STRING: - $current['content'] = $content; - break; - case FILE_ASN1_TYPE_UTC_TIME: - case FILE_ASN1_TYPE_GENERALIZED_TIME: - $current['content'] = $this->_decodeTime($content, $tag); - default: - - } - - $start+= $length; - $decoded[] = $current + array('length' => $start - $current['start']); + $tag = $type & 0x1F; + if ($tag == 0x1F) { + $tag = 0; + // process septets (since the eighth bit is ignored, it's not an octet) + do { + $loop = ord($encoded[0]) >> 7; + $tag <<= 7; + $tag |= ord($this->_string_shift($encoded)) & 0x7F; + $start++; + } while ( $loop ); } - return $decoded; + // Length, as discussed in paragraph 8.1.3 of X.690-0207.pdf#page=13 + $length = ord($this->_string_shift($encoded)); + $start++; + if ( $length == 0x80 ) { // indefinite length + // "[A sender shall] use the indefinite form (see 8.1.3.6) if the encoding is constructed and is not all + // immediately available." -- paragraph 8.1.3.2.c + $length = strlen($encoded); + } elseif ( $length & 0x80 ) { // definite length, long form + // technically, the long form of the length can be represented by up to 126 octets (bytes), but we'll only + // support it up to four. + $length&= 0x7F; + $temp = $this->_string_shift($encoded, $length); + // tags of indefinte length don't really have a header length; this length includes the tag + $current+= array('headerlength' => $length + 2); + $start+= $length; + extract(unpack('Nlength', substr(str_pad($temp, 4, chr(0), STR_PAD_LEFT), -4))); + } else { + $current+= array('headerlength' => 2); + } + + $content = $this->_string_shift($encoded, $length); + + // at this point $length can be overwritten. it's only accurate for definite length things as is + + /* Class is UNIVERSAL, APPLICATION, PRIVATE, or CONTEXT-SPECIFIC. The UNIVERSAL class is restricted to the ASN.1 + built-in types. It defines an application-independent data type that must be distinguishable from all other + data types. The other three classes are user defined. The APPLICATION class distinguishes data types that + have a wide, scattered use within a particular presentation context. PRIVATE distinguishes data types within + a particular organization or country. CONTEXT-SPECIFIC distinguishes members of a sequence or set, the + alternatives of a CHOICE, or universally tagged set members. Only the class number appears in braces for this + data type; the term CONTEXT-SPECIFIC does not appear. + + -- http://www.obj-sys.com/asn1tutorial/node12.html */ + $class = ($type >> 6) & 3; + switch ($class) { + case FILE_ASN1_CLASS_APPLICATION: + case FILE_ASN1_CLASS_PRIVATE: + case FILE_ASN1_CLASS_CONTEXT_SPECIFIC: + $newcontent = $this->_decode_ber($content, $start); + $length = $newcontent['length']; + if (substr($content, $length, 2) == "\0\0") { + $length+= 2; + } + + $start+= $length; + + return array( + 'type' => $class, + 'constant' => $tag, + // the array encapsulation is for BC with the old format + 'content' => array($newcontent), + // the only time when $content['headerlength'] isn't defined is when the length is indefinite. + // the absence of $content['headerlength'] is how we know if something is indefinite or not. + // technically, it could be defined to be 2 and then another indicator could be used but whatever. + 'length' => $start - $current['start'] + ) + $current; + } + + $current+= array('type' => $tag); + + // decode UNIVERSAL tags + switch ($tag) { + case FILE_ASN1_TYPE_BOOLEAN: + // "The contents octets shall consist of a single octet." -- paragraph 8.2.1 + //if (strlen($content) != 1) { + // return false; + //} + $current['content'] = (bool) ord($content[0]); + break; + case FILE_ASN1_TYPE_INTEGER: + case FILE_ASN1_TYPE_ENUMERATED: + $current['content'] = new Math_BigInteger($content, -256); + break; + case FILE_ASN1_TYPE_REAL: // not currently supported + return false; + case FILE_ASN1_TYPE_BIT_STRING: + // The initial octet shall encode, as an unsigned binary integer with bit 1 as the least significant bit, + // the number of unused bits in the final subsequent octet. The number shall be in the range zero to + // seven. + if (!$constructed) { + $current['content'] = $content; + } else { + $temp = $this->_decode_ber($content, $start); + $length-= strlen($content); + $last = count($temp) - 1; + for ($i = 0; $i < $last; $i++) { + // all subtags should be bit strings + //if ($temp[$i]['type'] != FILE_ASN1_TYPE_BIT_STRING) { + // return false; + //} + $current['content'].= substr($temp[$i]['content'], 1); + } + // all subtags should be bit strings + //if ($temp[$last]['type'] != FILE_ASN1_TYPE_BIT_STRING) { + // return false; + //} + $current['content'] = $temp[$last]['content'][0] . $current['content'] . substr($temp[$i]['content'], 1); + } + break; + case FILE_ASN1_TYPE_OCTET_STRING: + if (!$constructed) { + $current['content'] = $content; + } else { + $current['content'] = ''; + $length = 0; + while (substr($content, 0, 2) != "\0\0") { + $temp = $this->_decode_ber($content, $length + $start); + $this->_string_shift($content, $temp['length']); + // all subtags should be octet strings + //if ($temp['type'] != FILE_ASN1_TYPE_OCTET_STRING) { + // return false; + //} + $current['content'].= $temp['content']; + $length+= $temp['length']; + } + if (substr($content, 0, 2) == "\0\0") { + $length+= 2; // +2 for the EOC + } + } + break; + case FILE_ASN1_TYPE_NULL: + // "The contents octets shall not contain any octets." -- paragraph 8.8.2 + //if (strlen($content)) { + // return false; + //} + break; + case FILE_ASN1_TYPE_SEQUENCE: + case FILE_ASN1_TYPE_SET: + $offset = 0; + $current['content'] = array(); + while (strlen($content)) { + // if indefinite length construction was used and we have an end-of-content string next + // see paragraphs 8.1.1.3, 8.1.3.2, 8.1.3.6, 8.1.5, and (for an example) 8.6.4.2 + if (!isset($current['headerlength']) && substr($content, 0, 2) == "\0\0") { + $length = $offset + 2; // +2 for the EOC + break 2; + } + $temp = $this->_decode_ber($content, $start + $offset); + $this->_string_shift($content, $temp['length']); + $current['content'][] = $temp; + $offset+= $temp['length']; + } + break; + case FILE_ASN1_TYPE_OBJECT_IDENTIFIER: + $temp = ord($this->_string_shift($content)); + $current['content'] = sprintf('%d.%d', floor($temp / 40), $temp % 40); + $valuen = 0; + // process septets + while (strlen($content)) { + $temp = ord($this->_string_shift($content)); + $valuen <<= 7; + $valuen |= $temp & 0x7F; + if (~$temp & 0x80) { + $current['content'].= ".$valuen"; + $valuen = 0; + } + } + // the eighth bit of the last byte should not be 1 + //if ($temp >> 7) { + // return false; + //} + break; + /* Each character string type shall be encoded as if it had been declared: + [UNIVERSAL x] IMPLICIT OCTET STRING + + -- X.690-0207.pdf#page=23 (paragraph 8.21.3) + + Per that, we're not going to do any validation. If there are any illegal characters in the string, + we don't really care */ + case FILE_ASN1_TYPE_NUMERIC_STRING: + // 0,1,2,3,4,5,6,7,8,9, and space + case FILE_ASN1_TYPE_PRINTABLE_STRING: + // Upper and lower case letters, digits, space, apostrophe, left/right parenthesis, plus sign, comma, + // hyphen, full stop, solidus, colon, equal sign, question mark + case FILE_ASN1_TYPE_TELETEX_STRING: + // The Teletex character set in CCITT's T61, space, and delete + // see http://en.wikipedia.org/wiki/Teletex#Character_sets + case FILE_ASN1_TYPE_VIDEOTEX_STRING: + // The Videotex character set in CCITT's T.100 and T.101, space, and delete + case FILE_ASN1_TYPE_VISIBLE_STRING: + // Printing character sets of international ASCII, and space + case FILE_ASN1_TYPE_IA5_STRING: + // International Alphabet 5 (International ASCII) + case FILE_ASN1_TYPE_GRAPHIC_STRING: + // All registered G sets, and space + case FILE_ASN1_TYPE_GENERAL_STRING: + // All registered C and G sets, space and delete + case FILE_ASN1_TYPE_UTF8_STRING: + // ???? + case FILE_ASN1_TYPE_BMP_STRING: + $current['content'] = $content; + break; + case FILE_ASN1_TYPE_UTC_TIME: + case FILE_ASN1_TYPE_GENERALIZED_TIME: + $current['content'] = $this->_decodeTime($content, $tag); + default: + } + + $start+= $length; + + // ie. length is the length of the full TLV encoding - it's not just the length of the value + return $current + array('length' => $start - $current['start']); } /** - * ASN.1 Decode + * ASN.1 Map * * Provides an ASN.1 semantic mapping ($mapping) from a parsed BER-encoding to a human readable format. * @@ -805,16 +824,6 @@ class File_ASN1 return $this->_encode_der($source, $mapping, null, $special); } - /** - * ASN.1 Encode (Helper function) - * - * @param String $source - * @param Array $mapping - * @param Integer $idx - * @param Array $special - * @return String - * @access private - */ /** * ASN.1 Encode (Helper function) * diff --git a/tests/Unit/File/ASN1/FE.pdf.p7m b/tests/Unit/File/ASN1/FE.pdf.p7m new file mode 100644 index 00000000..6b912715 Binary files /dev/null and b/tests/Unit/File/ASN1/FE.pdf.p7m differ diff --git a/tests/Unit/File/ASN1/DevTest.php b/tests/Unit/File/ASN1Test.php similarity index 81% rename from tests/Unit/File/ASN1/DevTest.php rename to tests/Unit/File/ASN1Test.php index d0d51042..fa2b150e 100644 --- a/tests/Unit/File/ASN1/DevTest.php +++ b/tests/Unit/File/ASN1Test.php @@ -7,7 +7,7 @@ require_once 'File/ASN1.php'; -class Unit_File_ASN1_DevTest extends PhpseclibTestCase +class Unit_File_ASN1Test extends PhpseclibTestCase { /** * on older versions of File_ASN1 this would yield a PHP Warning @@ -233,4 +233,39 @@ class Unit_File_ASN1_DevTest extends PhpseclibTestCase $this->assertInternalType('array', $result); } + + /** + * older versions of File_ASN1 didn't handle indefinite length tags very well + */ + public function testIndefiniteLength() + { + $asn1 = new File_ASN1(); + $decoded = $asn1->decodeBER(file_get_contents(dirname(__FILE__) . '/ASN1/FE.pdf.p7m')); + $this->assertEquals(count($decoded[0]['content'][1]['content'][0]['content']), 5); // older versions would have returned 3 + } + + public function testDefiniteLength() + { + // the following base64-encoded string is the X.509 cert from + $str = 'MIIDITCCAoqgAwIBAgIQT52W2WawmStUwpV8tBV9TTANBgkqhkiG9w0BAQUFADBM' . + 'MQswCQYDVQQGEwJaQTElMCMGA1UEChMcVGhhd3RlIENvbnN1bHRpbmcgKFB0eSkg' . + 'THRkLjEWMBQGA1UEAxMNVGhhd3RlIFNHQyBDQTAeFw0xMTEwMjYwMDAwMDBaFw0x' . + 'MzA5MzAyMzU5NTlaMGgxCzAJBgNVBAYTAlVTMRMwEQYDVQQIEwpDYWxpZm9ybmlh' . + 'MRYwFAYDVQQHFA1Nb3VudGFpbiBWaWV3MRMwEQYDVQQKFApHb29nbGUgSW5jMRcw' . + 'FQYDVQQDFA53d3cuZ29vZ2xlLmNvbTCBnzANBgkqhkiG9w0BAQEFAAOBjQAwgYkC' . + 'gYEA3rcmQ6aZhc04pxUJuc8PycNVjIjujI0oJyRLKl6g2Bb6YRhLz21ggNM1QDJy' . + 'wI8S2OVOj7my9tkVXlqGMaO6hqpryNlxjMzNJxMenUJdOPanrO/6YvMYgdQkRn8B' . + 'd3zGKokUmbuYOR2oGfs5AER9G5RqeC1prcB6LPrQ2iASmNMCAwEAAaOB5zCB5DAM' . + 'BgNVHRMBAf8EAjAAMDYGA1UdHwQvMC0wK6ApoCeGJWh0dHA6Ly9jcmwudGhhd3Rl' . + 'LmNvbS9UaGF3dGVTR0NDQS5jcmwwKAYDVR0lBCEwHwYIKwYBBQUHAwEGCCsGAQUF' . + 'BwMCBglghkgBhvhCBAEwcgYIKwYBBQUHAQEEZjBkMCIGCCsGAQUFBzABhhZodHRw' . + 'Oi8vb2NzcC50aGF3dGUuY29tMD4GCCsGAQUFBzAChjJodHRwOi8vd3d3LnRoYXd0' . + 'ZS5jb20vcmVwb3NpdG9yeS9UaGF3dGVfU0dDX0NBLmNydDANBgkqhkiG9w0BAQUF' . + 'AAOBgQAhrNWuyjSJWsKrUtKyNGadeqvu5nzVfsJcKLt0AMkQH0IT/GmKHiSgAgDp' . + 'ulvKGQSy068Bsn5fFNum21K5mvMSf3yinDtvmX3qUA12IxL/92ZzKbeVCq3Yi7Le' . + 'IOkKcGQRCMha8X2e7GmlpdWC1ycenlbN0nbVeSv3JUMcafC4+Q=='; + $asn1 = new File_ASN1(); + $decoded = $asn1->decodeBER(base64_decode($str)); + $this->assertEquals(count($decoded[0]['content']), 3); + } }