Martha,
You need to change the ascii translation to specify code pages (e2a is
not reversible), for example:
|xlate from 37 to 819
This ascii is not utf anything. I think
|vchar 8 16
gets you utf-16; certainly, the utf filter converts between the utf
encodings, it does not convert to/from utf, so you do not want that.
j.
On 05/07/2015 12:39 AM, Martha McConaghy wrote:
I need a bit of help to get past a final hurdle in a project I've been working
on. The overall goal is to be able to take a text string and make a NTLM hash
out of it. Everything I've read says that NTLM is really a UTF-16 string put
through a MD4 hash. So, that is what I'm trying to do.
I have a MD4 hash generator stage written in REXX (a port from one that was
originally written for OS/2) and it works fine for creating a straight MD4
hash. (Many thanks to Neale Ferguson for getting me that!)
I convert the string to ASCII first and then put it through the
generator:
'PIPE var String1|' ,
'xlate e2a |' ,
'rexx md4|' ,
'cons'
I've compared the results with web based MD4 generators and the results
match.
Now, the hard part. Getting the string from ASCII to UTF-16. I
have John's new PIPELINE module from January and have been playing with
the UTF stage. However, I'm not sure if I'm doing the right thing, I'm
definitely not getting a result that matches any of the online NTLM
generators.
'PIPE var String1|' ,
'xlate e2a |' ,
'utf 8 to 16|' ,
'fblock 1|' ,
'spec x00 1 1 n|' ,
'join *|' ,
'strip|' ,
'rexx md4|' ,
'cons'
Any ideas? Is a conversion from UTF 8 to 16 even valid for this? Please, I
am definitely NOT a "code table geek", so use small words...%-)
Martha
Martha McConaghy
System Architect/Technical Lead
Marist College
