On Dec 20, 2013, at 10:46 AM, Michael Lum <michael.lum@xxxxxxxxxxxxxxxxx> wrote:
> Is there a goal to remove g_iconv calls from Wireshark.
I'd certainly like to do so to the maximum extent possible. I would like to have *all* code set handling done by using ENC_ arguments to proto_tree_add_item() or tvb_get_string_enc().
The code in dissectors would be much simpler, it wouldn't depend on particular g_iconv() implementations handling particular character sets, and it would allow us to handle invalid strings as we choose.
> I checked charsets.c/.h and there are two encodings that are not available that are used in the ANSI SMS dissector.
>
> iso-8859-8 (Latin/Hebrew) is pretty easy to add I believe I can follow the pattern for the code that is there now.
Yes, the ISO 8859-x character sets and encodings are fairly straightforward.
> The other is EUC-KR (Korean). I tried to find a code page that looks like the ISO ones but I'm not how these
> conversions are supposed to work.
DBCS encodings, such as the EUC encodings, will be more work, but we should do them eventually as well.