UNIX for Advanced & Expert Users

BeginDate 07/01/06
End: 07/31/06
Cust: A02991 - Burnham
0002000 5,829,773 145.3
0009701 4,043,850 267.3
2005000 286,785.13 100.0


BeginDate 07/01/06
End: 07/31/06
Cust: A01239 - East Track PSE
Index A
0009902 317,356.82 890.2
0020021 94,694.23 90989.1


BeginDate 07/01/06
End: 07/31/06
Cust: A04067 - Fidelity Growth
investement
0009902 1,063,852 2990.3
0020021 699,709.35 4590.3


Above is my incoming file format.
I need to create a file with 6 columns( BeginDate,EndDate,Cust and the three associated columns with each Customer).The customer record
some times may be shown as 2 records.I need to make as one record.

I am kind of lost how to solve

can you show us the output you needed?

Begin End Cust: Col1 Col2 Col3
07/01/06 07/31/06 A02991 - Burnham 0002000 5,829,773 145.3
07/01/06 07/31/06 A02991 - Burnham 0009701 4,043,850 267.3
07/01/06 07/31/06 A02991 - Burnham 2005000 286,785.13 100.0


This is how i need the output.

Code:

awk ' 
/BeginDate/ {
getline end_dt
getline cust
getline cust1
sub(".* ","",$0)
bgn_dt=$0
sub(".*: ","",end_dt)
sub(".*: ","",cust)
if ( match ( cust1 , "^[0-9][0-9]*" ) )
        print $0 " " end_dt " " cust " " cust1
else
{
        cust = cust cust1
}
}
/^[0-9]+/{
print bgn_dt " " end_dt " " cust " " $0 }' file

here's a slightly different approach:

nawk -f kris.awk myFile.txt

kris.awk:

Code:

BEGIN {
  FS=RS=""
}

{
   head=sprintf("%s%s", substr($1, index($1, " ")+1), OFS)
   head=head sprintf("%s%s", substr($2, index($2, " ")+1), OFS)
   head=head sprintf("%s%s", substr($3, index($3, " ")+1), OFS)
   for(f=4; f<=NF; f++)
      if( $f ~ /^[0-9]/ )
         printf("%s%s\n", head, $f)
}

Thank you for the help