UNIX for Advanced & Expert Users

Hi,
I have this file which has 3 columns, District , stores and unit. What I want is all rows belonging to one district to be created separately under each district, the districts may vary every day , the source file may have 3 districts today and may have 160 tomorrow, so what I need is a script that puts all district rows in one file and so on , do you have any ideas ? Attached is the sample of what I am asking for


Source File
District Store UnitSales
D1 ST101 4
D1 ST102 2
D1 ST103 3
D2 ST104 45
D2 ST105 6
D2 ST106 7
D1 ST107 8
D1 ST108 89
D2 ST109 8
D3 ST111 0
D3 ST101 7


D1 file
D1 ST101 4
D1 ST102 2
D1 ST103 3
D1 ST107 8
D1 ST108 89

D2 file
D2 ST104 45
D2 ST105 6
D2 ST106 7
D2 ST109 8

D3 file
D3 ST111 0
D3 ST101 7

Thanks, very much

Code:

while read district store unit; do
    echo "$district $store $unit" >> ${district}.txt
done < file

Try....

Code:

awk 'NF{print $0 > "outfile." $1}' infile

I did not understand how this works??

Could you please explain a little bit more?

I understand the redirect of the record to outfile.field1 works...

I dont know how NF(no of fields) is applicable here... and also how they get grouped is what I wanted to know.

Code:

NF{print $0 > "outfile." $1}

The block is executed if NF is greater than one.In case of blank line NF is zero and the block is not executed.