Your point is right but these code point does not exist in the subject
string so this isn't the issue here.
I'm really stuck at this one :S
Thank you again for trying to help!
On Fri, May 29, 2009 at 2:40 PM, Tom Worster <fsb@xxxxxxxxxx> wrote:
> On 5/28/09 2:06 PM, "Nitsan Bin-Nun" <nitsan@xxxxxxxxxxxx> wrote:
>
> > preg_replace("/([\xE0-\xFA])/e","chr(215).chr(ord(\${1})-80)",$s);
>
> ...
>
> > The preg_replace() above convert the Hebrew chars into UTF8.
>
> that preg_replace takes a byte string $s and:
>
> - leaves bytes with value 0-127 intact
> - converts bytes with value 224-250 to the utf8 multibyte character code of
> the corresponding win-1255 character
> - produces an invalid utf8 output for all other code points, for which see:
> http://en.wikipedia.org/wiki/Windows-1255
>
> so it's a win-1255 to utf-8 converter only so long as you are confident
> that
> code points 128-223 and 251-255 are never in the subject string.
>
> try iconv('CP1255', 'UTF-8', $s) instead.
>
>
>
>
>
