Shell Programming and Scripting

Code:

sort -t '|' -k1n -k4n -k3rn -k2nr abc.txt

vgersh,

I'm sorry to say that it doesn't work...I'm trying to be clear here

1705614|FNP|2|99|C|1/20/2009 0:00:00
1705614|MSN|1|88|U|1/2/2009 0:00:00

should be sorted on 0th field i.e 1705614
1705614
Then on 4th field i.e C ( C is should come first)
U
Should both be the 'C' s,
sort on 3rd field i.e 99 (Ascending)
88
Should both be the same numbers,
sort on 2nd field i.e 2 (descending)
1
I hope I am clear...Thanks

Fields and characters within fields are numbered starting with 1

Code:

sort -t '|' -k1n,1 -k5,5 -k4n,4 -k3n,3r myFile

Vgersh,

I appreciate your help...Seems working...Thank you so much

vgersh,

Hope I'm not bothering you...one last question for now

1705614|FNP|2|99|C|1/2/2009 0:00:00
1705614|MSN|2|99|C|1/22/2009 0:00:00

Assuming all those previously sorted fields equal < can i sort the above
to get the second row up since the date is recent...

Thanks again...

Code:

nawk -F'|' '{n=split($NF, a, "[ /]"); gsub(":", "", a[n]);printf("%d%02d%02d%s%s\n", a[3], a[1], a[2], a[n], OFS $0)}' OFS='|' myFile | sort -rn -k1,1 | cut -d '|' -f2-

sample.txt file
1873072|PAC|2|3|C|1/4/2009 0:00:00
1873072|PA|2|3|U|1/9/2009 0:00:00

nawk -F'|' '{n=split($NF, a, "[ /]"); gsub(":", "", a[n]);print a[3] a[1] a[2] a[n] OFS $0}' OFS='|' sample.txt | sort -rn -k1,1 | cut -d '|' -f2-

gives

1873072|PA|2|3|U|1/9/2009 0:00:00
1873072|PAC|2|3|C|1/4/2009 0:00:00

but In the above case I need the data in reverse, because column 5 should be preferred for 'C', only in case if both are 'C's or 'U's then i want this to be sorted on the last column i.e date

Thanks
-venkat