SENA S.A.
May 1996
Greek Character Encoding for Electronic Mail Messages
Status of This Memo
-
This memo provides information for the Internet community. This memo does not specify an Internet standard of any kind. Distribution of this memo is unlimited.
Overview and Rational
-
This document describes a standard encoding for electronic mail [RFC822] containing Greek text and provides implementation guide- lines. The standard is based on MIME [RFC1521] and the ISO 8859-7 character encoding. Although the implementation of this standard is straightforward several non-standard but "functional" - though unlikely to inter-operate - alternatives are in common use. For this reason we highlight common implementation and mail user agent setup errors.
Description
In order to transfer Greek text via electronic mail the text is first translated into the ISO 8859-7 character set, and then encoded using either the Base64 (preferable for text that is mainly Greek) or the Quoted-Printable (justifiable in cases where some Greek words appear inside predominately Latin text) method, as defined in MIME.
The following table provides most common Greek encodings (see also [RFC1345]):
0646 37 M7 51 MC 23 69 LG L1 G7 GO GC 28 97 Description ---- -- -- -- -- -- -- -- -- -- -- -- -- -- ----------- 0386 ea a2 86 cd 71 86 b6 Capital alpha with acute 0388 eb b8 8d ce 72 8d b8 Capital epsilon with acute 0389 ec b9 8f d7 73 8f b9 Capital eta with acute 038a ed ba 90 d8 75 90 ba Capital iota with acute 038c ee bc 92 d9 76 92 bc Capital omicron with acute 038e ef be 95 da 77 95 be Capital upsilon with acute 038f f0 bf 98 df 78 98 bf Capital omega with acute 0390 c0 a1 fd a1 c0 Small iota with acute and diaeresis 0391 80 c1 a4 b0 41 a4 61 41 61 41 41 c1 Capital alpha 0392 81 c2 a5 b5 42 a5 62 42 62 42 42 c2 Capital beta 0393 82 c3 a6 a1 43 a6 67 23 43 67 43 44 c3 Capital gamma 0394 83 c4 a7 a2 44 a7 64 40 44 64 44 45 c4 Capital delta 0395 84 c5 a8 b6 45 a8 65 45 65 45 46 c5 Capital epsilon 0396 85 c6 a9 b7 46 a9 7a 46 7a 46 49 c6 Capital zeta 0397 86 c7 aa b8 47 aa 68 47 68 47 4a c7 Capital eta 0398 87 c8 ac a3 48 ac 75 5c 48 75 48 4b c8 Capital theta 0399 88 c9 ad b9 49 ad 69 49 69 49 4c c9 Capital iota 039a 89 ca b5 ba 51 b5 6b 4b 6b 4a 4d ca Capital kappa 039b 8a cb b6 a4 52 b6 6c 5e 4c 6c 4b 4e cb Capital lamda 039c 8b cc b8 bb 53 b7 6d 4d 6d 4c 4f cc Capital mu 039d 8c cd b7 c1 54 b8 6e 4e 6e 4d 50 cd Capital nu 039e 8d ce bd a5 55 bd 6a 21 4f 6a 4e 51 ce Capital xi 039f 8e cf be c3 56 be 6f 50 6f 4f 52 cf Capital omicron 03a0 8f d0 c6 a6 57 c6 70 3f 51 70 50 53 d0 Capital pi 03a1 90 d1 c7 c4 58 c7 72 52 72 51 55 d1 Capital rho 03a3 91 d3 cf aa 59 cf 73 5f 53 73 53 56 d3 Capital sigma 03a4 92 d4 d0 c6 62 d0 74 54 74 54 58 d4 Capital tau 03a5 93 d5 d1 cb 63 d1 79 55 79 55 59 d5 Capital upsilon 03a6 94 d6 d2 bc 64 d2 66 5d 56 66 56 5a d6 Capital phi 03a7 95 d7 d3 cc 65 d3 78 58 78 57 5b d7 Capital chi 03a8 96 d8 d4 be 66 d4 63 3a 59 63 58 5c d8 Capital psi 03a9 97 d9 d5 bf 67 d5 76 5b 5a 76 59 5d d9 Capital omega 03aa da ab 91 da Capital iota with diaeresis 03ab db bd 96 db Capital upsilon with diaeresis 03ac e1 dc 9b c0 b1 9b dc Small alpha with acute 03ad e2 dd 9d db b2 9d dd Small epsilon with acute 03ae e3 de 9e dc b3 9e de Small eta with acute 03af e5 df 9f dd b5 9f df Small iota with acute 03b0 e0 fc fe fc e0 Small upsilon with acute and diaeresis 03b1 98 e1 d6 e1 8a d6 61 41 61 61 e1 Small alpha 03b2 99 e2 d7 e2 8b d7 62 42 62 62 e2 Small beta 03b3 9a e3 d8 e7 8c d8 63 47 63 64 e3 Small gamma 03b4 9b e4 dd e4 8d dd 64 44 64 65 e4 Small delta 03b5 9c e5 de e5 8e de 65 45 65 66 e5 Small epsilon 03b6 9d e6 e0 fa 8f e0 66 5a 66 69 e6 Small zeta 03b7 9e e7 e1 e8 9a e1 67 48 67 6a e7 Small eta 03b8 9f e8 e2 f5 9b e2 68 55 68 6b e8 Small theta 03b9 a0 e9 e3 e9 9c e3 69 49 69 6c e9 Small iota 03ba a1 ea e4 eb 9d e4 6b 4b 6a 6d ea Small kappa 03bb a2 eb e5 ec 9e e5 6c 4c 6b 6e eb Small lamda 03bc a3 ec e6 ed 9f e6 6d 4d 6c 6f ec Small mu 03bd a4 ed e7 ee aa e7 6e 4e 6d 70 ed Small nu 03be a5 ee e8 ea ab e8 6f 4a 6e 71 ee Small xi 03bf a6 ef e9 ef ac e9 70 4f 6f 72 ef Small omicron 03c0 a7 f0 ea f0 ad ea 71 50 70 73 f0 Small pi 03c1 a8 f1 eb f2 ae eb 72 52 71 75 f1 Small rho 03c2 aa f2 ed f7 af ed 77 57 72 77 f2 Small final sigma 03c3 a9 f3 ec f3 ba ec 73 53 73 76 f3 Small sigma 03c4 ab f4 ee f4 bb ee 74 54 74 78 f4 Small tau 03c5 ac f5 f2 f9 bc f2 75 59 75 79 f5 Small upsilon 03c6 ad f6 f3 e6 bd f3 76 46 76 7a f6 Small phi 03c7 ae f7 f4 f8 be f4 78 58 77 7b f7 Small chi 03c8 af f8 f6 e3 bf f6 79 43 78 7c f8 Small psi 03c9 e0 f9 fa f6 db fa 7a 56 79 7d f9 Small omega 03ca e4 fa a0 fb b4 a0 fa Small iota with diaeresis 03cb e8 fb fb fc b8 fb fb Small upsilon with diaeresis 03cc e6 fc a2 de b6 a2 fc Small omicron with acute 03cd e7 fd a3 e0 b7 a3 fd Small upsilon with acute 03ce e9 fe fd f1 b9 fd fe Small omega with acuteNote:
-
-
All values are in hexadecimal.
-
The column headers refer to the following character sets:
0646 The ISO 2DIS 10646 code.
37 PC code page 737 also known as 437G. Note that some implementa-
tions of this code page do not include capital letters with acute.
M7 Character set 8859-7 as implemented in Microsoft Windows 3.1,
Microsoft Windows 3.11, and Microsoft Windows 95.
51 IBM code page 851.
MC The Greek code page implemented on the Apple Macintosh computers.
23 IBM code page 423 (EBCDIC-CP-GR).
69 IBM code page 869.
LG Latin Greek (iso-ir-19).
L1 Latin Greek 1 (iso-ir-27). This page only contains the Greek cap-
ital letters whose glyphs do not exist in the Latin alphabet. The
other capital letters are rendered using the equivalent Latin let-
ter (e.g. "Greek capital letter alpha" is rendered as "Latin capi-
tal letter A"). When mapping "Latin Greek 1" text to ISO 8859-7
the Latin capital letters should only be transcribed to the
equivalent Greek ones if a suitable heuristic determines that the
-
-
specific Latin letters are used to represent Greek glyphs.
-
G7 7 bit Greek (iso-ir-88).
GO Old 7 bit Greek (iso-ir-18).
GC Greek CCITT (iso-ir-150).
28 Character set ISO 5428:1980 (iso-ir-55).
97 The target character set ISO 8859-7:1987 (ELOT-928) (iso-ir-126).
MIME Headers
-
-
A mail message that contains Greek text must contain at least the following MIME headers:
MIME-Version: 1.0 Content-type: text/plain; charset=ISO-8859-7 Content-transfer-encoding: BASE64 | Quoted-Printable In the future, when all email systems implement fully transparent 8-bit e-mail as defined in RFC 1425 and RFC 1426 the message body encoding phase described in this standard will be no longer needed. In this case the requisite MIME headers are modified as follows: MIME-Version: 1.0 Content-type: text/plain; charset=ISO-8859-7 Content-transfer-encoding: 8BIT Even when RFC 1425 is used, Q or B encoding will continue to apply to message headers as detailed in the following section. Optional-
It is recommended, although not required, to support Greek encod- ing in mail headers as specified in RFC 1522. Specifically, the B-encoding format is to be the default method used for encoding Greek text in RFC-822 mail headers, and the Q-encoding format the method to use for the exceptional case of encoding a single Greek word or letter in an otherwise Latin-character-based header.
-
Example
-
-
Below is a short example of Quoted-Printable encoded Greek email:
Date: Wed, 31 Jan 96 20:15:03 EET From: Diomidis Spinellis <dds@senanet.com> Subject: Sample Greek mail To: Achilleas Voliotis <achilles@theseas.ntua.gr> MIME-Version: 1.0 Content-ID: <Wed_Feb_14_18_49_50_EET_1996_0@senanet> Content-Type: Text/plain; charset=ISO-8859-7 Content-Transfer-Encoding: Base64 -
yuHr5+zd8eEsCgrU7yDl6+vn7enq/CDh6/bc4uf07yDh8O/05evl3/Th6SDh8PwgMjYg4/Hc
7Ozh9OEuCg==
Discussion
-
-
It is possible [RFC1428] (and unfortunately common practice) to set up an arrangement of mail user and transfer agents that allow end users to communicate with Greek e-mail messages while violating a number of standards. Such arrangements are unlikely to offer wide scale interoperability.
One common error is to arrange the rendering and composition of Greek messages by rigging a mail user agent hosted in an ISO 8859-1 environment to use a presentation font that contains Greek glyphs and a keyboard input method that generates Greek text using those glyphs. The resulting messages begin with header items indicating contents in the ISO 8859-1 character set and include text in a totally different encoding. Unfortunately this "solution" appears to "work" across similar systems and is widely used.
One other error is to tag Greek text generated on Microsoft Windows platforms as ISO 8859-7 without an intermediate translation phase. It is important to note that the character set used by the Microsoft Windows Greek implementations is NOT the same as the ISO 8859-7 representation. First of all, the character set used to represent Greek characters differs slightly from the ISO 8859-7 encoding (this difference was instrumented in order to rectify the appearance of an early version of Microsoft Word for Windows in which the end-of-section symbol clashed with the "Greek capital alpha with acute" glyph). In addition, a number of 8-bit characters available on Greek Windows implementations are not part of the ISO 8859-7 character set.
Note that the ISO 8859-7 encoding is equivalent to the Greek Standards Organisation ELOT-928 encoding.
-
References
-
[ISO-8859] Information Processing -- 8-bit Single-Byte Coded Graphic Character Sets, Part 7: Latin/Greek alphabet, ISO 8859-7, 1987. [RFC822] Crocker, D., "Standard for the Format of ARPA Internet Text Messages", STD 11, RFC 822, UDEL, August 1982. [RFC1345] Simonsen, K., "Character Mnemonics & Character Sets" RFC 1345, Rationel Almen Planlaegning, June 1992. [RFC1425] Klensin, J., Freed N., Rose M., Stefferud E., and D. Crocker, "SMTP Service Extensions", RFC 1425, United Nations University, Innosoft International, Inc., Dover Beach Consulting, Inc., Network Management Associates, Inc., The Branch Office, February 1993. [RFC1426] Klensin, J., Freed N., Rose M., Stefferud E., and D. Crocker, "SMTP Service Extension for 8bit-MIME Transport", RFC 1426, United Nations University, Innosoft International, Inc., Dover Beach Consulting, Inc., Network Management Associates, Inc., The Branch Office, February 1993. [RFC1428] Vaudreuil, G., "Transition of Internet Mail from Just-Send-8 to 8bit-SMTP/MIME", RFC 1428, CNRI, February 1993. [RFC1521] Borenstein N., and N. Freed, "MIME (Multipurpose Internet Mail Extensions) Part One: Mechanisms for Specifying and Describing the Format of Internet Message Bodies", Bellcore, Innosoft, September 1993. [RFC1522] Moore K., "MIME Part Two: Message Header Extensions for Non-ASCII Text", University of Tennessee, September 1993.
Security Considerations
-
Security issues are not discussed in this memo.
Author's Address
-
Diomidis Spinellis
SENA S.A.
Kyprou 27
GR-152 47 Filothei
GREECEPhone: +30 (1) 6854535 Fax: +30 (1) 6840631 EMail: D.Spinellis@senanet.com