Most of the characters appear as question marks. I tried "$b = iconv("UTF-8","ISO-8859-1",$b);" but I got an error saying,
"Detected an illegal character in input string".
On Sunday, 17 November 2013 16:19:45 UTC-8, Pierre Abbat wrote:
On Sunday, November 17, 2013 15:02:38 qx4...@gmail.com wrote:
> Sorry, but it doesn't look like a keyword search will work. From the
> research that I have done, full text indices can only be added to text
> columns and not to blob columns, and the data for the posts and
> translations are blobs not text. When I changed the column from a blob to a
> text on a test database on my local computer, it converted the unicode
> characters to random looking ASCII characters, so I had to change it back.
If you take the random-looking characters and run them through iconv, do you
get Unicode? Are they actually ASCII, or do they have accents?
$ iconv -f latin1 -t utf-8
A vonaton egy őrült, mellette egy őr ült. Örült az őrült, hogy mellette egy őr
ült.
A vonaton egy Årült, mellette egy År ült. Ãrült az Årült, hogy mellette
egy År ült.
The second byte of "ő", 0x91, is invisible because it's a control character.
(lo fenki cu zvati lo trene .i lo zgaku'i cu mlana ra .i lo fenki cu gleki le
nu lo zgaku'i cu mlana ra)
mu'omi'e .pier.
--
loi mintu se ckaji danlu cu jmaji