On 20/05/08 09:11, Nikolai Weibull wrote:
> On Sun, May 18, 2008 at 8:44 PM, Tony Mechelynck
> <[EMAIL PROTECTED]> wrote:
>
>>> Note, that you probably do not want to use BOM with UTF-8.
>>> See http://unicode.org/faq/utf_bom.html#29 (Q: Can a UTF-8 data stream
>>> contain the BOM character (in UTF-8 form)? If yes, then can I still
>>> assume the remaining UTF-8 bytes are in big-endian order?)
>
>> The BOM can also be used in UTF-8, not to determine endianness (which is
>> not relevant for UTF-8 -- one could argue that UTF-8 is always
>> big-endian) but to distinguish UTF-8 from other encodings including
>> UTF-16 and UTF-32.
>
> How can you argue that? UTF-8 is neither big-endian nor
> little-endian. It's just a sequence of 8-bit bytes.
Exactly. Some people (including you, apparently) say that endianness is
not a property of bytes but of words (or doublewords, quadwords, etc.)
when written to disk. Since UTF-8 doesn't use 2^n-word data items (with
n in the set {0, 1, 2, ...}), those people would say that it is neither
big-endian nor little-endian. According to a different definition of
endianness (used, apparently, by Ilya Bobir), any sequence of two or
more bytes representing a single integer (and for instance a Unicode
codepoint number) can be big-endian (if the bits of higher weight come
first) or little-endian (if it's the bits of lower weight). According to
this latter definition, UTF-8 is always big-endian.
Best regards,
Tony.
--
Everything should be built top-down, except the first time.
--~--~---------~--~----~------------~-------~--~----~
You received this message from the "vim_dev" maillist.
For more information, visit http://www.vim.org/maillist.php
-~----------~----~----~----~------~----~------~--~---
