Chris Moore
2006-Jul-10  13:30 UTC
[Samba] nmblookup fails to look up hostnames containing non-ASCII characters
I'm trying to use nmblookup to find the IP address of a Windows machine
called "MARKETKA" (only with an accent over the 'E').  If I do
a reverse
lookup on the IP address, with debugging enabled, I can see that the hex
code for the accented 'E' character is 0x90:
  $ nmblookup -d 4 -A 192.168.1.15 | grep MARK
  answers   0 char .MARK.TKA     hex 044D41524B90544B4120202020202020
  answers  20 char   ...MARK.TKA hex 20200084004D41524B90544B41202020
      MARK..TKA       <00> -         B <ACTIVE>
      MARK..TKA       <20> -         B <ACTIVE>
If I try to lookup that name, however, it can't be found (I'm using
Python to run nmblookup because I know how to insert characters given
their hex codes in Python):
  $ python -c 'import os; print os.popen("nmblookup MARK%cTKA" %
0x90).readlines()'
  ['querying MARK\x90TKA on 192.168.1.255\n', 'name_query failed to
find
name MARK\x90TKA\n']
nmblookup works fine for all the machines on the network with ASCII
names:
  $ nmblookup
  tom querying tom on 192.168.1.255
  192.168.1.10 tom<00>
Is this a known limitation of nmblookup?  Is there any way I can work
around it?
Thanks.
Chris.
