> Given 3 columns of text with tabs in between, what grep pattern (I assume that's the best way to handle this) would select the first character after the 2nd tab?
>
> 112 000 PDQ
> 987 000 DEF
> 090 000 XYZ
> 678 000 ABC
>
> would become
> 678 000 ABC
> 987 000 DEF
> 112 000 PDQ
> 090 000 XYZ
Use Text -> Sort Lines...
Check "Sort using pattern"
Searching pattern: ^(.+?)\t(.+?)\t(.+)
Sort using: Specific sub-patterns:
\3
--
Rod Buchanan
KDSI / Kelly Supply Co / ISCO
308 382-8764 x220
[^\t]*\t[^\t]*\t([^\t]*)
Sort on \1
Basically what you are saying is, I want any number of characters (*) that are not a tab ([^\t]) followed by a tab, twice. Then, I want you to grab the next set of characters that are not tabs into a buffer () and since that's the first buffer, it will be named \1. Now sort on that buffer.