RE: UTF-7 - I'm not really smarter

From: Keutgen, Walter (walter.keutgen@be.unisys.com)
Date: Tue Mar 28 2006 - 11:31:17 CST

  • Next message: Richard Wordingham: "Re: UTF-7 - I'm not really smarter"

    Otto,

    the reason is that Unicode refuses any escape sequence mechanism of any kind, here + ... -.

    Walter

    THIS COMMUNICATION MAY CONTAIN CONFIDENTIAL AND/OR OTHERWISE PROPRIETARY MATERIAL and is thus for use only by the intended recipient. If you received this in error, please contact the sender and delete the e-mail and its attachments from all computers.

    -----Original Message-----
    From: unicode-bounce@unicode.org [mailto:unicode-bounce@unicode.org] On Behalf Of Otto Stolz
    Sent: Dienstag, den 28. März 2006 19:15
    To: Kornkreismuster@web.de
    Cc: unicode@unicode.org
    Subject: Re: UTF-7 - I'm not really smarter

    Hello,

    Kornkreismuster@web.de schrieb:
    > Reading this [RFC 2152], I got the feeling it only encodes UTF-16 encoded Texts,
    > but I think that's not true.

    The description in RFC 2152, chapter 4, is probably misleading to
    the uninitiated. The key to understanding is that all UTFs are
    equivalent: they encode the same character set, viz. the whole Uni-
    code, and any string encoded in one UTF can be easily transformed
    into any other.

    So, all references in chapter 4 of RFC 2152 to UTF-16, and to
    16-bit code elements, are only meant to facilitate the description
    of the algorithm. You can describe the UTF-7 encoding algorithm
    (with a grain of salt) thusly:
    1. encode the source string in UTF-16 (regardless of its previous
        encoding);
    2. convert every three UTF-16 code units into 8 bytes using a modified
        base-64 algorithm (hence, every byte encodes 6 bit);
    3. enclose the result between a plus and a minus sign.
    Alternatively, runs of "harmless" characters may be encoded in ASCII,
    instead of applying steps 1..3, above.

    The latter alternative renders UTF-7 indeterminate: a character
    string may be encoded in several ways, cf. my example in
    <http://www.systems.uni-konstanz.de/Otto/Vortrag/Charset/Unicode-Grundlagen.html#UU-7>
    -- in contrast to UTF-8, UTF-16, and UTF-32. I guess, this is the
    main reason for not having UTF-7 in the Uncode standard.

    Regards,
       Otto Stolz

    This archive was generated by hypermail 2.1.5 : Tue Mar 28 2006 - 11:34:03 CST