On 2 Apr, 14:05, "Tine M=FCller" <ti...@email.dk> wrote:
> Can someone help me what code to change so it's functioning with danish ch=
ar
> which is in the database?

I had a good answer, but the accursed Google Groups interface ate it
8-(

Lose the <meta> in your page. You're sending a HTTP-Header too (better
idea anyway) and that will take precedence.  As the page is claiming
to be UTF-8, then you'd better send UTF-8 / Unicode characters in it,
not ISO-8859 characters. If your DB is working in ISO-8859 already,
then you might find it easier to change the page generation code to
send the right HTTP header.

Read up (Wikipedia is good) on Unicode, ISO-8859 and the distinctions
between "character sets" (lists of squiggles that humans use) and
"character encoding" (sequences of bytes that computers use to refer
to these squiggles).

You _must_ use a character set that includes the squiggle you need.
Danish is easy (commonplace ISO-8859-1 suffices), Czech or Polish is
harder. In general these days, go straight to Unicode. This works for
all languages at once, ISO-8859 requires you to swap between
ISO-8859-1 and ISO-8859-4 depending on which side of the Baltic you
are.

You then need an encoding that works for the character set you use and
is configured so that your DB content matches your page (or gets
translated) and your page's header matches your page's content. Use
UTF-8 for Unicode, use the relevant ISO-8859-* for ISO-8859 (and you
would have to swap between sides of the Baltic).


Alternatively (a very bad idea, I only mention it to save a follow-up
pinhead doing so), use plain old ASCII as an encoding and HTML numeric
character entities or HTML entity references. You will find that
numeric character entities are no easier to generate than UTF-8 (if
your DB contains ISO-8859) and that entity references are a big chunk
of tedious look-up code to create (maybe a PHP library exists for
this). This approach is tedious, but it does work in cases where you
can't fix the web server's HTTP-Headers to indicate a correct or
useful encoding.