codec to parse raw UCS data?

[Oleg Leschov]

Where can I find a list and documentation for codecs?
What I want to do is to make a unicode string out of unicode data.
for example. I am parsing NTFS metadata, that contains filenames as
UCS-2 code, so I have a binary string that pretends to be a UCS-2.
Currently I am using hand-written module, a wrapper to iconv library, to make
translation of this data (that is, I used it for that task).
However, I am aware of python's codecs that are supposed to do the same
thing in much prettier way.. So, as long as I can do things like
st = u'\uxxxx' and such, I could construct and exec it, but this is ugly...
Shouldn't there be a simple thing like <bin.string>.decode("UCS-2")
that would return a python's unicode string?
Or perhaps I've missed something in the documentation...

[Fredrik Lundh]

Oleg Leschov wrote:

Where can I find a list and documentation for codecs?
What I want to do is to make a unicode string out of unicode data.
for example. I am parsing NTFS metadata, that contains filenames as
UCS-2 code, so I have a binary string that pretends to be a UCS-2.

the "utf-16-le" codec is probably what you want.

(utf-16 is basically ucs-2 plus mechanisms to encode characters outside
the 16-bit BMP set; IIRC, Windows 2k and later uses utf-16, not ucs-2).

</F>