Shell Programming and Scripting

Hi,

I need to extract values from a CSV file based on some conditions as explained below:

File format details:

1. each set starts with AAA only
2. number of columns is fixed
3. number of rows per set may vary (as they are having different CCC rows)

Now, i need to extract 3rd column of AAA and 4th columns of the lines starting with CCC of that set. These values of 1 dataset to form one line.

For. e.g, i have data for 3 sets and file has the below:

AAA,1,a,b,c,d
BBB,1,j,k,l,m
CCC,1,p,q,r,s
CCC,1,w,x,y,z
AAA,2,j,k,l,m
BBB,2,a,b,c,d
CCC,2,p,q,r,s
AAA,3,w,x,y,z
BBB,3,a,b,c,d
CCC,3,p,q,r,s
CCC,3,m,n,o,p
CCC,3,i,j,k,l

then the output must be (3 lines only as 3 data sets exist)

a,q,x
j,q
w,q,n,j


Please advise.

Thanks
Prvn

Use nawk or /usr/xpg4/bin/awk on Solaris:

Code:

awk -F, 'END { print _ }
/^(A|C)/ { 
  if (/^A/ && _) { print _; _ = "" }  
  _ = _ ? _ FS $4 : $3 
  }' infile

something to start with:
nawk -F, -f prv.awk OFS=, myFile

prv.awk:

Code:

$1=="AAA" { printf("%c%s",(FNR==1)?"":ORS, $3); next }
$1=="CCC" { printf("%c%s",OFS, $4)}
END{print}

Ah, radoulov's solution is much nicer - congrats!!!

Thank you radoulov and vgersh99,

I tested radoulov's solution and worked great. Thank you very much!!


Prvn