L2/00-035 Internet Draft James Seng Martin Duerst 28th Jan 2000 Tin Wee Tan Expires End of July 2000 UTF-5, a transformation format of Unicode and ISO 10646 Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet- Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. Distribution of this document is unlimited. Please send comments to the authors at jseng@pobox.org.sg, mduerst@w3.org and tinwee@post1.com. Abstract A new transformation format, called UTF-5 for Unicode is proposed. The resulting string of this UTF is within a [A-V][0-9] alphanumeric range. This enables legacy systems or protocols designed for alpha- numerical character set only to be multilingual enabled and inter- nationalized immediately. Example of such systems are the domain name system and email addresses. 1. Introduction ISO/IEC 10646-1:1993 [ISO-10646] defines a 16 bit character set, UCS-2 and a 31-bit character set, UCS-4. UCS-2 and UCS-4 are coded representation forms of the UCS and UCS-4 has no assignments outside the region correspoding to UCS-2 (the Basic Multilingual Plane, BMP) at this moment. The UCS-2 and UCS-4 encodings, however, are hard to use in many current applications and protocols that assume 8 or even 7 bit characters. Even newer systems able to deal with 16 bit char- acters cannot process UCS-4 data. This situation has led to the development of so-called UCS transformation formats (UTF), each with different characteristics. At this moment, there are 3 standard UTF, namely UTF-7 [UTF7], UTF-8 Expires End of July 2000 [Page 1] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 [UTF8] and UTF-16 [UTF16], each is a variable length transformation which gives 7 bit, 8 bit and 16 bit strings respectively. While these are sufficient for most application uses, there are however some legacy systems which are, unfortunately, unable to handle even 7 bit strings either due to technical restriction or common uses. The object of this memo is to propose a UTF-5 which gives a trans- formed string that is within [A-V][0-9] alphanumerical character set. This enables legacy systems designed for alphanumerical character set only to be multilingual enabled and internationalized immediately. UTF-8 is the preferred transformation format for all new IETF standards [IETFPC]. UTF-5 is not here to change this. It is proposed to support legacy applications or protocols that cannot be modified in a simple way to handle 8 bits using UTF-8 encoding. See Section 4 on the discussion on how UTF-5 can be used for Domain Name System [DNS] and Simple Mail Transfer Protocol [SMTP] Address. 2. UTF-5 definition In UTF-5, each character is encoded using a sequence of 1 to 8 octets. Two transformations are needed for UTF-5, namely 1. Determine the quintet ("5-bit") binary sequence. 2. From a table, translate the quintet to the resulting string. Take note that UTF-5 is not a sequence of quintets but a sequence of octets where each octets are in the alphanumeric range. Alpha- numeric is defined as A to V (uppercase only) and 0 to 9 in this context. This memo does not specify the binary pattern of the alphanumeric characters as the purpose of the transformation is to get a alpha- numeric string which represents a multilingual string. However, it is presumed that US-ASCII [US-ASCII] is used for most purposes. 2.1 Determine the quintet binary sequence The first quintet of a binary sequence will have the highest-order bit set to 1 and the remaining quintet will have the highest-order bit set to 0. The remaining 4 bits of every quintet contain bits from the value of the character to be encoded. The table below summarizes the format of these different quintet types. The letter x indictes bits available for encoding bits of the UCS-4 character value. Expires End of July 2000 [Page 2] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 UCS-4 range (hex.) UTF-5 quintet sequence (binary) 0000 0000-0000 000F 1xxxx 0000 0010-0000 00FF 1xxxx 0xxxx 0000 0100-0000 0FFF 1xxxx 0xxxx 0xxxx 0000 1000-0000 FFFF 1xxxx 0xxxx 0xxxx 0xxxx ... 1000 0000-7FFF FFFF 1xxxx 0xxxx 0xxxx ..... 0xxxx 2.2 Translation table for quintet and alphanumeric character The translation table for quintet binary pattern and alphanumeric character is as follows. This is effectively a duoettrigesimal (base 32) string representation of the quintets. quintet quintet quintet quintet 00000 0 01000 8 10000 G 11000 O 00001 1 01001 9 10001 H 11001 P 00010 2 01010 A 10010 I 11010 Q 00011 3 01011 B 10011 J 11011 R 00100 4 01100 C 10100 K 11100 S 00101 5 01101 D 10101 L 11101 T 00110 6 01110 E 10110 M 11110 U 00111 7 01111 F 10111 N 11111 V 2.3 Encoding from UCS-4 to UTF-5 1) Determine the required number of octets from the character value. Let U be the UCS-4 value, then the required number of octets is log16(U+1) rounded up. 2) Prepare the quintet binary sequence. Put the highest order bit of the first quintet as 1 and highest order bit of the rest of the quintet as 0. 3) Fill in the bits marked x from the bits of the character value, starting from the lower-order bits of the character value and putting them first in the last quintet of the sequence, then the next to last, etc until all x bits are filled in. 4) For each quintet, apply the lookup table in Section 2.2 to get the corresponding alphanumeric character. 2.4 Decoding UTF-5 to UCS-4 1) Determine the length of the octet sequence. As according to the UTF-5 encoding, every character will have the inital octet within the range 'G' to 'V'. Thus, the length of the octet sequence can be determined by looking for 'G' to 'V' in the UTF-5 string. 2) Apply the reverse lookup according to the table in Section 2.2 to get the quintet binary sequence. 3) Initialize the 4 octets of the UCS-4 character with all bits set to 0. Expires End of July 2000 [Page 3] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 4) Distribute the bits from the sequence to the UCS-4 character, first the lower-order bits from the last octet of the sequence and proceeding to the left until no x bits are left. If the UTF-5 sequence is no more than four octets long, the low order bits of the result can be interpreated directly as UTF-16 value or equivalently Unicode. 2.5 Detecting a UTF-5 string As the UTF-5 string is a alphanumeric string, it is difficult to differentiate between a normal ASCII document or a UTF-5 document. Nevertheless, if the string is sufficiently long, it is possible to do some detection of UTF-5 string based on the fact that 1. UTF-5 strings only have characters within '0'-'9' and 'A'-'V'. 2. UTF-5 strings have a well-defined inital octet of 'G' to 'V'. 3. The 'G' character always occurs as the inital and only octet. In other word, the shortest UTF-5 sequence is "G". For example, "GF" is not a valid UTF-5 sequence. 3. Examples of UTF-5 The Unicode sequence "A." (0041, 2262, 0391, 002E) may be encoded as follows: "K1I262J91IE" The Unicode sequence "Hi Mom !" (0048, 0069, 0020, 004D, 006F, 006D, 0020, 263A, 0021) may be encoded as follows: "K8M9I0KDMFMDI0I63AI1" The Unicode sequence representing the Han characters for the Japanese word "nihongo" (65E5, 672C, 8A9E) may be encoded as follows: "M5E5M72COA9E" Note that from the examples, it is obvious that there is a short-cut to the UTF-5 transformation which goes like this: If the hexadecimal notation is 0x00000000, convert it to 'G'; otherwise skip over all leading zeros in the hexadecimal notation and convert the first non-zero hexadecimal digital as follows: '1' to 'H', '2' to 'I', ... 'F' to 'V'. Retain all trailing hexadecimal digits. 4. Applications There are many applications where UTF-5 would be useful for Internationalization ("i18n"). Here are some of the possible uses. Expires End of July 2000 [Page 4] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 a. Internationalised Domain Names In the Domain Name System, although the technical standard does not prevent 8-bits character to be use as domain names, general use of the system restrict it to only A-Z (upper and lower), 0-9 and "-" as a valid domain name. This poses great difficulty when doing i18n of domain names as the current UTF-7, UTF-8 and UTF-16 are not compatible with the existing software system already in used. Please join idn@ops.ietf.org to join the discussion on Internationalised Domain Names "idn". Send an email to idn-request@ops.ietf.org with the word "subscribe" in the body. More information on IDN can be found at the following website: http://www.idns.org/ http://www.imc.org/idn/ b. Internationalization of Simple Mail Transfer Protocol Address While it is possible for a person to send SMTP Mail in different languages using different character set to each another using Multi- purpose Internet Mail Extensions [MIME], the SMTP Mail Address remains a challenge to be Internationalized. Internationalization of SMTP Address has two barriers, 1. the Internationalization of Domain Name System and 2. the Internationalization of the mailbox or username. SMTP mailbox has a very strict check [RFC822] due to many potential security risks when using symbols or special char- acters in mailbox. UTF-5 will allow Unicode to be used as mailbox with minimal change in system and without additional security risks. For example, an SMTP Email address for "yamaguchi@asahi.ninhon" (5C71 53J3 '@' 671D 65E5 '.' 65E5 672C) can be represented in UTF-5 "LC71L3E3@M71DM5E5.M5E5M72C". This is a valid [RFC822] Email address which will not be rejected. It will then be the responsiblity of the user interface to render "LC71L3E3@M71DM5E5.M5E5M72C" properly as "yamaguchi@asahi.ninhon". Internationalization of URIs is not discussed in this memo. Please refer to http://www.w3.org/International/0-URL-and-ident.html. However, uses for UTF-5 extend beyond Internet back to old legacy systems such as Telegram system or even Morse code allowing Multilingual characters to be transmitted. 5. Security Considerations This memo does not address any security consideration at the moment. 6. Acknowledgements UTF-5 was first defined by Martin Duerst at the University of Zurich in draft-duerst-dns-i18n-00.txt. Contributors (not in any order): Expires End of July 2000 [Page 5] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 Marc Blanchet Paul Gampe Ken Whistler Graham Klyne 7. Bibliography [ISO-10646] ISO/IEC 10646-1:1993. International Standard -- Information technology -- Universal Multiple-Octet Coded Character Set (UCS) -- Part 1: Architecture and Basic Multilingual Plane. [UNICODE] The Unicode Standard, Version 2.0 (ISBN 0-201-48345- [UTF-16] 9). The minor reference is Unicode Technical Report #8, The Unicode Standard, Version 2.1. Refer to URL http://www.unicode.org/unicode/standard/versions/ [UTF7] Goldsmith, D., and M. Davis, "UTF-7: A Mail-safe Transformation Format of Unicode", RFC 1642, Taligent, Inc., July 1994. [UTF8] F. Yergeau "UTF-8: a transformation format of Unicode and ISO 10646", RFC2044, Alis Technologies, October 1996. [US-ASCII] Coded Character Set--7-bit American Standard Code for Information Interchange, ANSI X3.4-1986. [DNS] P. Mockapetris "Domain Names - Concepts and Facilities", RFC1034, ISI, November 1987, "Domain Names - Implementation and Specification", RFC1035, ISI, November 1987. [SMTP] Jonathan B. Postel "Simple Mail Transfer Protocol", [RFC822] RFC821, ISI, August 1982. David H. Crocker "Standard for ARPA Internet Text Messages", RFC822, Dept of Electrical Engineering, Univeristy of Delaware, August 1982. [MIME] "Multipurpose Internet Mail Extensions", RFC1341, N. Borensten, Bellcore, N. Freed, Innosoft, June 1992. [IETFPC] "IETF Policy on Character Sets and Languages", RFC2277 BCP18, H. Alvestrand, Jan 1998. Expires End of July 2000 [Page 6] Internet Draft UTF-5, a transformation format of Unicode Jan 2000 8. Author Address James C.H Seng i-DNS.net International Inc. 102 Elm Street Menlo Park CA 94025 Tel: (650) 322-6505 E-mail: jseng@pobox.org.sg Martin J. Duerst World Wide Web Consortium (W3C) Keio Research Institute at SFC Keio University Fujisawa 252-8520 Japan Tel: +81 446 49 11 70 E-mail: mduerst@w3.org NOTE -- Please write the author's name with u-Umlaut wherever possible, e.g. in HTML as Dürst. Tin Wee Tan, Dr National University of Singapore (NUS) c/o BioInformatic Center National University Hospital Lower Kent Ridge Road Singapore 119074 Tel: +65 774 7149 E-mail: tinwee@post1.com This memo is also archived at http://www.idns.org/technical.html Expires End of July 2000 [Page 7]