I have the following issue with a UTF8 files structured as following:

FIELD1§FIELD2§FIELD3§FIELD4

Looking at hexadecimal values of the file it uses A7 to codify §. So according to this codify it should be UTF8, but it's strange because A7 > 7F so 1 byte shouldn't be enough to codify §.

So I tried using directly a BufferedReader with a specified charset:

BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(input), utf8))

but when I try to tokenize the string with

SmartTokenizer st = new SmartTokenizer(toTokenize, "§")

(the SmartTokenizer is a modified version of the StringTokenizer that keeps empty tokens)

no splitting occurs, and if I try to print the string I obtain

FIELD1?FIELD2?FIELD3?...

so § used in the file is different from the one specified as a the delimiter, and it's not able to print out it too.

So what's the problem here? Maybe the original file should use 2 bytes to store §?