Shell Programming and Scripting

Hi

I have a text file with rows like this:

Code:

7 Herman ASI-40 Jungle (L) Blueprint (L) Weapon Herman ASI-40 Jungle (L) 215.00 57 65.21 114.41

and

Code:

9 Herman CAP-505 (L) Blueprint (L) Weapon Herman CAP-505 (L) 220.00 46.84 49.1 104.82

and

Code:

2 ClericDagger 1C blueprint Melee - Shortblade ClericDagger 1C 24.00 5.06 5.61 110.88

you note format is not fixed, numbers comes at different positions in row.
How could I make a script that extract numbers at different "positions"
the rows above have numbers

Code:

215.00 57 65.21 114.41

and

Code:

220.00 46.84 49.1 104.82

and

Code:

24.00 5.06 5.61 110.88

Example:
myscript1 (first pos) would generate 3 lines

Code:

215.00
220.00
24.00

myscript2 (second pos) would generate this:

Code:

57
46.88
5.06

any help welcome!
Pelle

Moderator's Comments:

Use code tags, see PM, thanks.

If I am not wrong, the numbers have fixed positions. In the 3 example lines, the values are always the 4th-last. The next are the 3rd-last.
How should the output look like? Want all numbers together, or just 1 position on all lines per file?

As a start:

Code:

$ awk '{print $(NF-3)}' infile
215.00
220.00
24.00
$ awk '{print $(NF-2)}' infile
57
46.84
5.06

Btw. is the dual wielding barbarian berserker already taken by a player?

Thanks alot! works like a charm.
Say I want to sort rows according to value from awk '{print $(NF-3)}' infile
Then I have to build a new file where every line is same as original file except every row has value from awk '{print $(NF-3)}' infile inserted at the beginning.
Could I do that by first vreating the file awk '{print $(NF-3)}' infile > infile2 and then ssomehow merge these 2 files?
I cant find any easy way to do this in a script,is there one?

pelle