join lines (2 or 3) & extracting fileds that have a specific word in i

[ancadmihai]

hi all

please give me a hand in solving the following problem

i have a file that contains the following lines


40765300012 226030100315050 CONNECTED AVAILABLE
OCSI-1 OFA-20 PWD-0000 OICK-20
SCHAR-4-0
40765300023 226030100315061 CONNECTED AVAILABLE
OCSI-1 OFA-20 PWD-0000 OICK-20
SCHAR-4-0
40765300031 226030100315069 CONNECTED AVAILABLE
OCSI-1 OFA-20 PWD-0000 OICK-20
SCHAR-4-0
40765300039 226030100315077 CONNECTED AVAILABLE



each msisdn(4076******) may have a OICK & SCHAR or a SCHAR. each of this info is printed in diferent lines



i need a script that has as an output lines that look like this

4076****** oick=** (if it exists) SCHAR-*-0(where * may be 8 or 4)


please give me a hand. i've managed to give myself a head acke with this.

thank's a lot in advance.

 

[bigoldbulldog]

How about this sed soln? Awk would probably be better.

#! /bin/sed -nf

1{
  /^4076/{s/ .*//;h;}
}
1!{
/^4076/{x;s/\n//g;p;g;s/ .*//;h;}
/^OCSI/{s/.*\( OICK.*\)/\L\1/;H;}
/^SCHAR/{s/^/ /;H;}
}
${
/^4076/{p;q}
g;s/\n//g;p
}

$ /path/sedfile /path/input

Cheers,
ND

 

[Annihilannic]

An awk solution:

awk '
        function printrec () { print num,oick,schar ; num=oick=schar="" }
        /^[0-9]/ && NR>1 { printrec() }
        /^[0-9]/ { num=$1 }
        /^OCSI/ { oick=$NF ; sub("OICK-","oick=",oick) }
        /^SCHAR/ { schar=$1 }
        END { printrec() }
' inputfile > outputfile

Annihilannic.

 

[ancadmihai]

thanks a lot annihilannic
you;ve saved my day.