Linguistic Guidelines for the Use of

the Arabic Language in Internet Domains

Abstract

This document constitutes technical specifications for the use of Arabic in Internet domain names and provides linguistic guidelines for Arabic domain names. It addresses Arabic-specific linguistic issues pertaining to the use of Arabic language in domain names.

Status of This Memo

This document is not an Internet Standards Track specification; it is published for informational purposes.

This is a contribution to the RFC Series, independently of any other RFC stream. The RFC Editor has chosen to publish this document at its discretion and makes no statement about its value for implementation or deployment. Documents approved for publication by the RFC Editor are not a candidate for any level of Internet Standard; see Section 2 of RFC 5741.

Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at http://www.rfc-editor.org/info/rfc5564.

Copyright Notice

Copyright © 2010 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document.

This document may not be modified, and derivative works of it may not be created, except to format it for publication as an RFC or to translate it into languages other than English.

Table of Contents

   1. Introduction ....................................................2
   2. Arabic Language-Specific Issues .................................3
      2.1. Linguistic Issues ..........................................4
           2.1.1. Diacritics (Tashkeel) and Shadda ....................4
           2.1.2. Kasheeda or Tatweel (Horizontal Character
                  Size Extension) .....................................5
           2.1.3. Character Folding ...................................5
      2.2. Supported Character Set ....................................6
      2.3. Arabic Linguistic Issues Affected by Technical
           Constraints ................................................8
           2.3.1. Numerals ............................................8
           2.3.2. The Space Character .................................8
   3. Summary and Conclusion ..........................................8
   4. Security Considerations .........................................9
   5. Acknowledgments .................................................9
   6. References ......................................................9
      6.1. Normative References .......................................9
      6.2. Informative References .....................................9

1. Introduction

The Internet Engineering Task Force (IETF) issued in March 2003 a set of RFCs for Internationalized Domain Names (IDN) ([1], [2], and [3]), which were planned to become the de facto standard for all languages. In 2007 and 2008, the following working drafts were released that propose revisions to the IDNA protocol:

• Internationalized Domain Names for Applications (IDNA): Background, Explanation, and Rationale [5]

• Internationalized Domain Names in Applications (IDNA): Protocol [6]

   o  An updated IDNA criterion for right-to-left scripts [7]

• The Unicode code points and IDNA [8]

These documents are known collectively as "IDNA2008".

This document constitutes a technical specification for the implementation of the IDN standards in the case of the Arabic language. It will allow the use of standard language tables to write domain names in Arabic characters. Therefore, it should be considered as a logical extension to the IDN standards. It thus presents guidelines for the proper use of Arabic characters with the IDN standards in an Arabic language context.

This document reflects the recommendations of the Arab Working Group on Arabic Domain Names (AWG-ADN), established by the League of Arab States (LAS), based on standardisation efforts of the United Nations Economic and Social Commission for Western Asia (UN-ESCWA) and on that group's document, "Guidelines for an Arabic Internet Domain Name" [9]. This document is also in full harmony with recent rigorous discussions that took place within the major language communities that use the Arabic script in their languages.

This document provides guidelines for the ways Arabic characters may be used for registering Internet domain names and how linguistic- specific issues should be handled. A few rules are recommended for application at the protocol level.

The key words "MUST", "REQUIRED", "SHOULD", "RECOMMENDED", and "MAY" in this document are to be interpreted as described in RFC 2119 [4].

Comments on this document are solicited and should be addressed to the working group's mailing list at ESCWA-ICTD@un.org and/or the author(s).

2. Arabic Language-Specific Issues

The main objective of the creation of Arabic domain names is to have a vehicle to increase Internet use amongst all strata of the Arabic- speaking communities.

Furthermore, a non-user-friendly domain name would further add to the ambiguity and the eccentricity of the Internet to the Arabic-speaking communities, thus contributing negatively to the spread of the

Internet and leading to further isolation of these communities at the global level.

Hence, there have been intensive efforts (especially those spearheaded by Dr. Al-Zoman and contributed to by UN-ESCWA and its Arabic Domain Names Task Force (ADN-TF)) to reach consensus on a multitude of linguistic issues with the following goals:

• To define the accepted Arabic character set to be used for writing domain names in Arabic, which is the subject of this document.

• To define the top-level domains of the Arabic domain name tree structure (i.e., Arabic gTLDs and ccTLDs). This goal will be handled in a separate document.

The first meeting of the AWG-ADN, held in Damascus from January- February 2005, gave special attention to the following:

• Simplification of the domain names, whenever possible, to facilitate the interaction of the Arabic user with the Internet.

• Adoption of solutions that do not lead to confusion either in reading or in writing, provided that this does not compromise the linguistic correctness of used words.

• Mixing Arabic and non-Arabic letters in the domain name label is not acceptable.

2.1. Linguistic Issues

There are a number of linguistic issues that have been proposed with respect to the use of the Arabic language in domain names. This section will highlight some of them. This section is based on the papers of Dr. Al-Zoman ([10] and [11]) and on the report of the first meeting of AWG-ADN [12]. For details, the reader is encouraged to review these references.

2.1.1. Diacritics (Tashkeel) and Shadda

Tashkeel and Shadda are accent marks placed above or below Arabic letters to produce proper pronunciation. They are thus used to differentiate different meanings for different words with the same base characters.

Neither Tashkeel nor Shadda are permitted in zone files when registering domain names in the Arabic language, although they are permitted in the current edition of IDNA2008. They can be supported or ignored, if necessary, in the user interface with local mappings and can be stripped before IDNA processing.

The following are their Unicode presentations:

U+064B ARABIC FATHATAN
U+064C ARABIC DAMMATAN
U+064D ARABIC KASRATAN
U+064E ARABIC FATHA
U+064F ARABIC DAMMA
U+0650 ARABIC KASRA
U+0651 ARABIC SHADDA
U+0652 ARABIC SUKUN

2.1.2. Kasheeda or Tatweel (Horizontal Character Size Extension)

Kasheeda (U+0640 ARABIC TATWEEL) must not be used in Arabic domain names and should be disallowed for Arabic language domain names. The Kasheeda is not a letter and does not have an effect on pronunciation. It is used to extend the horizontal length or change the shape of the preceding letter for graphical representation purposes in Arabic writing. Accordingly, it has no value for the writing of domain names. The same applies to all languages using the Arabic script. The authors recommend that it should be disallowed at the protocol level.

2.1.3. Character Folding

Character folding is the process where multiple letters (that may have some similarity with respect to their shapes) are folded into one shape. Examples of such Arabic characters include:

• Folding Teh Marbuta (U+0629) and Heh (U+0647) at the end of a word

   o  Folding different forms of Hamzah (U+0622, U+0623, U+0625, U+0627)

• Folding Alef Maksura (U+0649) and Yeh (U+064A) at the end of a word

• Folding Waw with Hamzah Above (U+0624) and Waw (U+0648)

With respect to the Arabic language, character folding is not acceptable because it changes the meaning of words and is against the principle of spelling rules. Replacing a character valid for use in domain names with another character also valid for use in domain names, which may have a similar shape, will give a different meaning. This will lead to only one word representing several words consisting of all the combinations of folded characters. Hence, the other words will be masked by a single word [10].

Mis-spelling or handwriting errors do occur, leading to mixing different characters despite the fact that this is not the case in published and printed materials. One of the motivations of this effort is to preserve the language, particularly with the spread of the globalization movement. Within this context, character folding is working against this motivation since it is going to have a negative effect on the principle and ethics of the language. Technology should work to preserve the language and not to destroy it. Thus, character folding should not be allowed. The case of digits is treated in a separate section below.

2.2. Supported Character Set

A domain name to be written in Arabic must be composed of a sequence of the following UNICODE characters and the FULL STOP (u+002E) to separate the labels. These are based on UNICODE version 5.0. The tables below are constructed using an inclusion-based approach. Thus, characters that are not part of these tables are prohibited.

             +---------+-------------------------------------+
             | Unicode | Character Name                      |
             +---------+-------------------------------------+
             | 0621    | ARABIC LETTER HAMZA                 |
             | 0622    | ARABIC LETTER ALEF WITH MADDA ABOVE |
             | 0623    | ARABIC LETTER ALEF WITH HAMZA ABOVE |
             | 0624    | ARABIC LETTER WAW WITH HAMZA ABOVE  |
             | 0625    | ARABIC LETTER ALEF WITH HAMZA BELOW |
             | 0626    | ARABIC LETTER YEH WITH HAMZA ABOVE  |
             | 0627    | ARABIC LETTER ALEF                  |
             | 0628    | ARABIC LETTER BEH                   |
             | 0629    | ARABIC LETTER TEH MARBUTA           |
             | 062A    | ARABIC LETTER TEH                   |
             | 062B    | ARABIC LETTER THEH                  |
             | 062C    | ARABIC LETTER JEEM                  |
             | 062D    | ARABIC LETTER HAH                   |
             | 062E    | ARABIC LETTER KHAH                  |
             | 062F    | ARABIC LETTER DAL                   |
             | 0630    | ARABIC LETTER THAL                  |
             | 0631    | ARABIC LETTER REH                   |
             | 0632    | ARABIC LETTER ZAIN                  |
             | 0633    | ARABIC LETTER SEEN                  |
             | 0634    | ARABIC LETTER SHEEN                 |
             | 0635    | ARABIC LETTER SAD                   |
             | 0636    | ARABIC LETTER DAD                   |
             | 0637    | ARABIC LETTER TAH                   |
             
             | 0638    | ARABIC LETTER ZAH                   |
             | 0639    | ARABIC LETTER AIN                   |
             | 063A    | ARABIC LETTER GHAIN                 |
             | 0641    | ARABIC LETTER FEH                   |
             | 0642    | ARABIC LETTER QAF                   |
             | 0643    | ARABIC LETTER KAF                   |
             | 0644    | ARABIC LETTER LAM                   |
             | 0645    | ARABIC LETTER MEEM                  |
             | 0646    | ARABIC LETTER NOON                  |
             | 0647    | ARABIC LETTER HEH                   |
             | 0648    | ARABIC LETTER WAW                   |
             | 0649    | ARABIC LETTER ALEF MAKSURA          |
             | 064A    | ARABIC LETTER YEH                   |
             | 0660    | ARABIC-INDIC DIGIT ZERO             |
             | 0661    | ARABIC-INDIC DIGIT ONE              |
             | 0662    | ARABIC-INDIC DIGIT TWO              |
             | 0663    | ARABIC-INDIC DIGIT THREE            |
             | 0664    | ARABIC-INDIC DIGIT FOUR             |
             | 0665    | ARABIC-INDIC DIGIT FIVE             |
             | 0666    | ARABIC-INDIC DIGIT SIX              |
             | 0667    | ARABIC-INDIC DIGIT SEVEN            |
             | 0668    | ARABIC-INDIC DIGIT EIGHT            |
             | 0669    | ARABIC-INDIC DIGIT NINE             |
             +---------+-------------------------------------+

Source: Supporting the Arabic Language in Domain Names [10]

Table 1: CHARACTERS FROM UNICODE ARABIC TABLE (0600-06FF)

                       +---------+-----------------+
                       | Unicode | Digit Name      |
                       +---------+-----------------+
                       | 0030    | DIGIT ZERO      |
                       | 0031    | DIGIT ONE       |
                       | 0032    | DIGIT TWO       |
                       | 0033    | DIGIT THREE     |
                       | 0034    | DIGIT FOUR      |
                       | 0035    | DIGIT FIVE      |
                       | 0036    | DIGIT SIX       |
                       | 0037    | DIGIT SEVEN     |
                       | 0038    | DIGIT EIGHT     |
                       | 0039    | DIGIT NINE      |
                       | 002D    | HYPHEN-MINUS    |
                       +---------+-----------------+
      
        Source: Supporting the Arabic Language in Domain Names [10]
      Table 2: CHARACTERS FROM UNICODE BASIC LATIN TABLE (0000-007F)

2.3. Arabic Linguistic Issues Affected by Technical Constraints

In this section, technical aspects of some linguistic issues are discussed.

2.3.1. Numerals

In the Arab countries, there are two sets of numerical digits used:

   o  Set I: (0, 1, 2, 3, 4, 5, 6, 7, 8, 9) mostly used in the western
      part of the Arab world.
   
   o  Set II: (u+0660, u+0661, u+0662, u+0663, u+0664, u+0665, u+0666,
      u+0667, u+0668, u+0669) mostly used in the eastern part of the
      Arab world.

Both sets may be supported in the user interface; however, the rule of numeral homogeneity must be observed. The rule specifies that digits from the Arabic-Indic set of numerals (u+0660 to u+0669) should not be allowed to mix with ASCII digits (u+0030 to u+0039) within the same Arabic domain name label. Thus, the appearance of a digit from one set prevents the use of any other digit from the other set.

2.3.2. The Space Character

The space character is strictly disallowed in domain names, as it is a control character. Instead, the hyphen (Al-sharta, i.e., u+02D) is proposed as a separator between Arabic words to avoid confusion that can take place if the words are typed without a separator.

It is acceptable to use the hyphen to separate between words within the same domain name label.

3. Summary and Conclusion

The proposed guidelines are in full accordance with the IETF IDN standards and take into account Arabic-language-specific issues within a compromise between grammatical rules of the Arabic language and ease of use of that language on the Internet.

In summary, the guidelines specify that, in Arabic domain names:

• Accent marks (Tashkeel and Shadda) are not permitted.

• Character folding is not permitted.

• If a numeral from the Arabic-Indic or ASCII digit sets appears in a label, numeral homogeneity is required.

• The hyphen must be used as a word separator instead of space.

4. Security Considerations

No particular security considerations could be identified regarding the use of Arabic characters in writing domain names. In particular, any potential visual confusion between different character strings is avoided using the guidelines proposed in this document.

5. Acknowledgments

ESCWA ICT Division provided support and funding for the development of this document with the objective of reaching a standard for comprehensive Arabic domain names. Thanks are due to SaudiNIC for its continuous efforts in supporting the development of Arabic domain names.

John Klensin and Harald Alvestrand reviewed the document and provided useful editorial and substantive support to enrich it.

6. References

6.1. Normative References

   [1]   Faltstrom, P., Hoffman, P., and A. Costello,
         "Internationalizing Domain Names in Applications (IDNA)", RFC
         3490, March 2003.
   
   [2]   Hoffman, P. and M. Blanchet, "Nameprep: A Stringprep Profile
         for Internationalized Domain Names (IDN)", RFC 3491, March
         2003.
   
   [3]   Costello, A., "Punycode: A Bootstring encoding of Unicode for
         Internationalized Domain Names in Applications (IDNA)", RFC
         3492, March 2003.
   
   [4]   Bradner, S., "Key words for use in RFCs to Indicate Requirement
         Levels", BCP 14, RFC 2119, March 1997.

6.2. Informative References

   [5]   Klensin, J., "Internationalized Domain Names for Applications
         (IDNA): Definitions, Background and Rationale", Work in
         Progress, September 2008.
   
   [6]   Klensin, J., "Internationalized Domain Names in Applications
         (IDNA): Protocol", Work in Progress, September 2008.
   
   [7]   Alvestrand, H. and C. Karp, "An updated IDNA criterion for
         right-to-left scripts", Work in Progress, July 2008.
   
   [8]   Faltstrom, P., "The Unicode Codepoints and IDNA", Work in
         Progress, July 2008.
   
   [9]   United Nations Economic and Social Commission for Western Asia
         (UN-ESCWA), "Guidelines for an Arabic Domain Name System
         (ADNS)", Work in Progress, November 2007.
   
   [10]  Al-Zoman, A., "Supporting the Arabic Language in Domain Names",
         October 2003, <http://www.arabic-domains.org/docs/
         NIC-docs/SupportingArabicDomainNmaes.pdf>.
   
   [11]  Al-Zoman, A., "Arabic Top-Level Domains", Paper presented in
         Expert Group Meeting on Promotion of Digital Arabic Content,
         the United Nations, Economic and Social Commission for Western
         Asia, Beirut, June 2003.
   
   [12]  League of Arab States, "Report of the first meeting of AWG-ADN,
         Damascus", February 2005, <http://www.arabic-
         domains.org/ar/intrnational-entites.php>.

Authors' Addresses

Ayman El-Sherbiny
Information and Communication Technology Division ESCWA
UN-House
P.O. Box 11-8575
Beirut
Lebanon

   EMail: El-sherbiny@un.org

Mansour Farah
Information and Communication Technology Division ESCWA
UN-House
P.O. Box 11-8575
Beirut
Lebanon

EMail:

          farah14@un.org

Ibaa Oueichek
Syrian Telecom Establishment
Damascus
Syria

   EMail: oueichek@scs-net.org
   
   Abdulaziz H. Al-Zoman, PhD
   SaudiNIC, General Directorate of Internet Services
   IT Sector, CITC
   King Abdulaziz City for Science and Technology
   PO Box 6086
   Riyadh  11442
   Saudi Arabia
   
   EMail: azoman@citc.gov.sa