RE: pb with implementation of greek characters

From: Marco.Cimarosti@icl.com
Date: Fri Mar 10 2000 - 10:35:38 EST


> I try to mix greek characters, in HTML text areas, to latin
> characters. One
> should send these pages to the server and read them with any navigator
> (netscape 4.0 or Internet explorer4.0). I search the best and
> simplest way to do it with UNICODE.

IMHO, the simplest way is to use UTF-8. You need an editor or word processor
that can save in this Unicode format (there is an increasing number of
packages supporting UTF-8), and you must instruct the client about the
document's text encoding. You do this by adding this meta tag in the HTML
header (<head> ... </head>):

        <meta http-equiv="content-type" content="text/html; charset=utf-8">

>My questions are:
> - What is the best representation (in HTML pages) for the
> scalar value
> of the character? In my examples, only the decimal notation
> (&#956; for ex.)
> seems to be taken into account. The hexadecimal notations (as
> 03BC , U+03BC,
> U-03BC, 0x......) don't work, when I'd prefer directly insert
> this one.

I thought that current HTML standards required decimal notation, but I see
from Markus Kuhn's reply that this is only a matter of some browsers not
supporting the HTML 4.0 hex syntax.

> - From an example found on the web, UltraEdit gives for the greek
> character "Tau" the hexadecimal value CE A4, when the ISO
> 10646-1 gives 03
> A4. Both work. So far as I know, one character may be encoded
> in several
> ways; but are both of them work? Or is one of them more common?

Octets 0xCE 0xA4 are the UTF-8 representation of U+03A4. The extra bits set
in 0xCE are there to tell the decoder how many octets follow to complete the
character (UTF-8 is a variable-length encoding that represents Unicode
characters with sequences of 1 to 4 octets).

0x03 0xA4 is a 16-bit little-endian representation (UTF-16, probably) for
the same character.

> - According to the Windows (or Mac) configuration, and even with
> Nescape 4.0 or Explorer 4.0, some characters appear on the
> screen, others
> not. I'd like to set up the greek font only on the server and
> to sent it
> from the server to the customer without a "plug in" process?
> Is it possible
> and How could I do ?

I don't know. But this should not be necessary for several recent versions
of Windows (NT, 2000, maybe also 98), that are equipped with Greek fonts by
default.

_ Marco



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:59 EDT