UTF-8 interop testing, take 2

[Francois Yergeau <FYergeau@alis.com>]

This message is the initial part of a test of Take 2 of interoperability of
UTF-8 on
the Internet.  The reason for having a second test is that the initial round
did not test any characters beyond the BMP (character number > 0xFFFF).

The test is based on a test file of plain text encoded in UTF-8, containing
text in a few languages and scripts. The test file was composed in Windows
2000 Notepad and is attached to this message as test-utf-8bis.txt.  It has
an initial BOM (Byte Order Mark).  The content of the test file, copy-pasted
into this message is:

------------------>snip<---------------------
UTF-8 interop test
===================
???: ??
???????: ????????????!
?????????: ???? ???
Español: ¡Hola!
Türkçe: Merhaba
????: ?????? ?????
??????????: ??????????????
===================
------------------>snip<---------------------

Also attached is test-utf-8bis.jpg, a JPEG screen shot showing the test file
in Windows 2000 Wordpad.  Wordpad is used here because Notepad allows only a
single font for the whole file, which is insufficient as I do not have a
single font containing glyphs for all the scripts in the test file.  To get
correct display in Wordpad, it is sufficient to set an adequate font (James
Kass' Code2001 here) on the last line, which is in the non-BMP Gothic
script).

Also attached is test-utf-8bis.html, an HTML version of the same text.
Instead
of a BOM, this version uses a <meta> element to identify the charset as
UTF-8.

An interop report will follow as a reply to this message.

-- 
François Yergeau