PureBasic Forums - English

First of all: I don't know too much about unicode .

Anyway I realized that unicode strings normally begin with $FFFE. But now I have one that has $FF00FE at the beginning and I don't understand what this means.

Any ideas?

It is a file header for unicode or other code types.
You can for example use notepad to open this file and use save as to check the current filetype.

The 2 a 3 bytes are in specific order and value to inidicate the specific type.
Old fashion ansifiles don't have this header.

not seeing yours

http://en.wikipedia.org/wiki/Byte-order_mark

Unicode strings don't have any special beginning.

$FFFE is at the start of some unicode files. It means the file uses UCS2 encoding (the same as PB uses for unicode strings).

$FF00FE is something different entirely. It probably isn't a unicode file, or it's a unicode file without a byte order mark. Or you simply read the value wrong, and it's really $FF FE 00 00 or $00 00 FE FF, which means UTF-32.

pdwyer wrote:not seeing yours

http://en.wikipedia.org/wiki/Byte-order_mark

That's why I'm confused.

Trond wrote:Unicode strings don't have any special beginning.

$FFFE is at the start of some unicode files. It means the file uses UCS2 encoding (the same as PB uses for unicode strings).

$FF00FE is something different entirely. It probably isn't a unicode file, or it's a unicode file without a byte order mark. Or you simply read the value wrong, and it's really $FF FE 00 00 or $00 00 FE FF, which means UTF-32.

To be more precise: The string I'm reading is a ID3-Tag. The tag has a unicode flag so the whole string must be in unicode as far as I know.
So normally the unicode tags I found began with $FFFE but now I found a mp3 file where every tag has $FF00FE at the beginning.

I really don't know what this means.

c4s wrote:

pdwyer wrote:not seeing yours

http://en.wikipedia.org/wiki/Byte-order_mark

That's why I'm confused.

Trond wrote:Unicode strings don't have any special beginning.

$FFFE is at the start of some unicode files. It means the file uses UCS2 encoding (the same as PB uses for unicode strings).

$FF00FE is something different entirely. It probably isn't a unicode file, or it's a unicode file without a byte order mark. Or you simply read the value wrong, and it's really $FF FE 00 00 or $00 00 FE FF, which means UTF-32.

To be more precise: The string I'm reading is a ID3-Tag. The tag has a unicode flag so the whole string must be in unicode as far as I know.
So normally the unicode tags I found began with $FFFE but now I found a mp3 file where every tag has $FF00FE at the beginning.

I really don't know what this means.

Well ID3 is messed up by a lot of editors (including windows media player - it doesn't safe the tagsize of APIC as syncsafe integer). Maybe its not your fault, but the fault of the editor you have used for the file.

I'm working from memory, but I'm fairly sure I've seen this used to determine whether the file uses Little-Endian or big-Endian byte ordering.

Hm, I just found out that the following will handle the string as I need it: ...Because it will display the "best" with (not) unicode-executable and (not) unicode-string, right?
At least it can handle $FF00FE and stuff.