acessing awk array element while getline < "file"

Shell Programming and Scripting
1

I am attempting to write a awk script that reads in a file after awk array elements are assigned and using those elements while reading in the new file. Does this make sense?

/pattern/ {tst[$2]=$3}
(( getline < "file" ) > 0 ) {
        x[$2]=x[$2] " "tst[$2]
}

When I print tst[$2] in the END statement it is assigned, but when I try to print tst[$2] in the while statement with getline is shows unasigned. How do I reference tst[$2] while reading in file with getline???

Thanks in advance.

2

I suppose FS has been modified,
try to force the split (assuming default OFS value, otherwise use whatever you need):

/pattern/ { tst[$2] = $3 }
(( getline < "file" ) > 0 ) {
        split($0, t, OFS)
        x[t[2]] = x[t[2]] FS tst[t[2]]
}
3

Thanks for the reply radoulov, still can't get this figured out. I think if I could reference it right, it will work. How come the following code prints out null, but if I use the same print statement in the END statement it will be successful

/pattern/ {tst[$2]=$3}
(( getline < "file" ) > 0 ) {
        print tst[23]   #will print null
}
END {
printf "%s--%s\n",tst[23],"END"  # is successful
4

Where is all the input coming from? Are you using only a file named "file" for all your input needs and can you post the entire awk script along with some input data?

5

In first part of awk script tst array is set for use in the getfile section of script. Problem I have got is, it looks like I can't use tst array while reading in and setting x array.

#!/bin/ksh
tput clear

dir=/my/dir/here/ #all files are gzip'd
ifile=/tmp/ifile
ofile=/tmp/ofile
$file=$dir$file.gz
echo "Enter the date and hour to run this script on [YYMMDDHH]: \c"
read file


#tool is run to generate $ofile here, file is semicolon delimited


/usr/bin/gzcat $file | /usr/xpg4/bin/awk '

BEGIN {
print "headers go here"
}


NR == 3 { print $2,$3"  --  "$4,$5 ; print "" ; next }

/^dise / {
        dipce = substr($2,4,1) * 1000 ; next
           }
/^LC/ {
        lec = substr($0,3,3) + dipce
        f++
        s = ""
        for ( i in arr ) { delete arr }
        }
	   f {
              if ( $1 == "RRAC" ) {
                s++
                 }
                if ( s ) {
                    if ( $1 == "tsec" ) {
                       lectsec = $2 * 1
                       }
                    if ( $1 == "VAR" ) {
                        car[$2 * 1]+=$8
                        lecCar = $2 * 1
                        c[lec,lectsec,BCar] = 0
                        }
                     c[lec,lectsec,BCar]++
                     if ( c[lec,lectsec,BCar] == 8 ) {
                        car[BCar]+=$8
                        tst[lec"-"BCar]=sprintf("%5.2f",car[BCar] / 12)
                        }
                    }
                if ( $1 == "RRAC-L" ) {
                        f = "" ; next
                   }
                   }


###This section not working because I can't access###
###array tst, to "test" I tried printing tst array, but can't get it to work.

{
while (( getline < "'"$ofile"'" ) > 0 ) {
FS = "\;" 
	x[$2"-"$4]+=tst[$5"-1"] ;
        x[$5"-"$7]+=tst[$5"-2"] ;
        x[$8"-"$10]+=tst[$5"-3"];
        x[$11"-"$13]+=tst[$5"-4] ;
}
}
#####################################################
END {
for ( i in x ) { print i,x }

    } '

/usr/bin/rm $ifile
/usr/bin/rm $ofile
6

This action is executed for every record (while, I suppose, your array in not yet build):

{
  while (( getline < "'"$ofile"'" ) > 0 ) {
  FS = "\;" 
    x[$2"-"$4]+=tst[$5"-1"] ;
        x[$5"-"$7]+=tst[$5"-2"] ;
        x[$8"-"$10]+=tst[$5"-3"];
        x[$11"-"$13]+=tst[$5"-4] ;
  }
}

Try moving it into the END block.

You can read both files in parallel if you need, the array infile is accessible consider this:

zsh-4.3.4% head *file
==> getlinefile <==
1 getlinefile
2 getlinefile

==> infile <==
1 infile
2 infile
zsh-4.3.4% awk '{ infile[$1] = $0 }
quote> (getline < "getlinefile") {
quote>   print infile[$1], $0
quote>   }' infile
1 infile 1 getlinefile
2 infile 2 getlinefile

This is not a good habit:

"'"$ofile"'"

Use this instead:

awk -v ofile="$ofile" ...
7

Thanks, I did not think of that. Getting closer but still not there. I moved to end statement like you sugested, but can you explain the following code:

(( getline < "file" ) > 0 ) {
        printf "%s--%s\n",tst["23"],"WHILE"     #this prints the element of tst[23]
	printf "%s\n",tst["\42"20+3"\42"]	#this print null???  why???

}
8

Because they are different:

zsh-4.3.4% awk 'BEGIN {
quote>   tst["23"]
quote>   tst["\42"20+3"\42"]
quote>   for (k in tst)
quote>     print k
quote> }'
23
"23"

If you want them equal:

zsh-4.3.4% awk 'BEGIN {
quote>   tst["\42"23"\42"]
quote>   tst["\42"20+3"\42"]
quote>   for (k in tst)
quote>     print k
quote> }'
"23"

And I suppose you need just:

zsh-4.3.4% awk 'BEGIN {
quote>   tst[23]
quote>   tst[20+3]
quote>   for (k in tst)
quote>     print k
quote> }'
23
9

Thanks ALOT I guess I got confused and frustrated. I was testing it wrong. I will try it on my script.

10

It's working! Can't thank you enough. No matter what I tried, I just couldn't get it going.

Thanks again.