Help with AWK SED

Hi guys,

I have a question regarding AWK or SED.

I have a data file which reads like this;

1,03005 41460 73215 10018 20000 39990 40092 51035 78782 879// 90850
2,
3,333 41998 55300 20000 83915 85650 90710 91129=
.........

There are about 100 or so lines like this.

What I need to do is to append the line with the = on the end to the line above so the data file reads like this

1,03005 41460 73215 10018 20000 39990 40092 51035 78782 879// 90850 333 41998 55300 20000 83915 85650 90710 91129=
2,
3,
........

Any ideas guys?

My pc is about to get throw out of the window....

Cheers,
Ian

Is this happening on every other line or is the pattern not regular?

---------- Post updated at 05:36 AM ---------- Previous update was at 05:06 AM ----------

Regardless, try this:

awk '
$2~/[0-9][^=]$/ {nr=NR}
$2~/=$/ { store2=$2;  $2=""; b[a[nr]]=b[a[nr]] store2 } 
{ a[NR]=$1; b[$1]=$2 }
END{
  for(i=1;i<=NR; i++)
     print a,b[a] 
}' FS="," OFS="," input

Another awk script :

awk -F, '
function printLine() {
    for (i=1; i<=cnt; i++) print line;
    cnt = 0;
}
{
    cnt++
    if (cnt==1) {
        line[1] = $0;
    } else {
        line[1] = line[1] " " $2;
        line[cnt] = $1 FS;
    }
    if (/=$/) printLine();
}
END { printLine() }
' inputfile

Input file:

1,03005 41460 73215 10018 20000 39990 40092 51035 78782 879// 90850
2,
3,333 41998 55300 20000 83915 85650 90710 91129=
4,123 456 789 012
5,abc def ghi jkl
7,AAA BBB CCC DDD EEE FFF
6,xyz=

Output :

1,03005 41460 73215 10018 20000 39990 40092 51035 78782 879// 90850  333 41998 55300 20000 83915 85650 90710 91129=
2,
3,
4,123 456 789 012 abc def ghi jkl AAA BBB CCC DDD EEE FFF xyz=
5,
7,
6,

Jean-Pierre.

Hi Guys,
Thanks for replying.

First it's not a regular patten, as some of the lines go like this

1 03005 41460 73215 10018 20000 39990 40092 51035 78782 879// 90850
2
3 333 41998 55300 20000 83915 85650 90710 91129=
4
5 03089 41460 73215 10018 20000 39990 40092 333 51035=

Generally, if the line starts with 333 it needs to be appended to the line above..

BTW - As I'm using AWK to read and process these files, can I just read the line terminated with /=$/ to read though the whole line until it gets to = ?

If so how is this done with awk? I have not been able to find any documentation on it

Cheers,

---------- Post updated 01-13-12 at 03:27 AM ---------- Previous update was 01-12-12 at 02:41 PM ----------

Hi Guys,

I have tried both scripts, but I was unable to get any of them to work.

mirni,
When I tried yours it just printed out the same as what I had

And Jean-Pierre,

It produced syntax errors on the last printLine

END { printLine() }

I removed this and more error appeared throughout the script

Please show us your script code and the errors produced

Try to replace awk with nawk or gawk (depending of your unix).

Jean-Pierre.

Hi Jean-Pierre,

I don't have nawk installed but I do have gawk. However it came up with the same error as awk

END printLine() }
             ^ syntax error

thanks
Ian

END { printLine() }

Jean-Pierre.

Yeah,

END { printLine() }

Is in there, but still producers the same error

I am trying this from a slightly different approach

awk 'BEGIN {RS="  "} { for (i-1;i<NF;i++) printf $(i) " "} END { print ""}' infile

This kind of works apart from removing the last set of numbers from the top line and adding a space on the following line

infile

1 03005 12671 73505 10027 21024 30197 40301 52007 60001 81434 91150
2
3  333 41998 55307 20344 81632 86075=  
4

outfile

1 03005 12671 73505 10027 21024 30197 40301 52007 60001 81434 333 41998 55307 20344 81632 86075= 
2
3  03005 12671 73505 10027 21024 30197 40301 52007 60001 81434 333 41998 55307 20344 81632 86075=
4

As you can see the outfile has the 91150 missing and also on the following lines after line one, there is a extra space appeared on the start of the lines.

Any ideas on this one?

Cheers,
ian

Can you please post the output of

cat -A inputFile  | head

I am suspicious you have some spaces after the final '=' that's why the awk isn't matching the lines.
Is this

awk '/=$/' input

returning any lines?

You could try to change this

$2~/=$/

to

$2~/= *$/

in my code.