Unique records into single line

[skp]

Hi All,

Below is my source file format.

Account name,city,value
ABC,hyd,1000
DEF,ban,2000
ABC,Chn,3000
GHI,US,3000
JKL,UK,4000
DEF,us,6600

I need to find unique records I. Account name column and all unique records to populate in single line like below.

Output:
/'ABC/',/'DEF/',/'GHI/',/'JKL/'

Please let me know how to get the desired output.

[rkashyap]

Create a DataStage job to sort/unique followed by Vertical Pivot on first column; replace delimiters as needed.

This can also be done in External Source stage or external filter of Sequential File stage by passing following command:

Code: Select all

awk -F',' '{print $1|"sort -u"}'|awk '{printf "/'"'"'" $1} END {print "/'"'"'"}'

[skp]

Hi rkashyap,

I tried to do this in unix itself and tried below command

awk -F',' '{print $1}' test2.txt|sort -u |awk '{printf "\\'"'"'" $1} END {print "\\'"'"'"}'

but it's displaying output like
\'A\'B\'C\'

But I need output like below
\'A\',\'B\',\'C\'

[chulett]

And... the Awk Clinic is back in session.

ps. What you "need" seems to have changed.

[rkashyap]

As Craig noted above, the requirement seems to have changed. For initial requirement try

Code: Select all

awk -F',' '{print $1|"sort -u"}'|awk 'BEGIN {SS=PS="/'"'"'"}{if (NR>1){PS="," SS}} {printf PS $0 SS}'

You can also code a dataStage job for this/these req (see previous post).