UNIX for Dummies Questions & Answers

First, get used to using vB Codes when quoting either code or other's posts - it really makes it easier to read and would improve your chances of people pay attention to you posts.

Secondly, you really don't need to

Code:

arr[$1]=1

OR

Code:

arr[$1] != 1

both of the above is redundant: you don't need to assign '1' reading the first file as you're just building the hash the values of the hash don't matter as long as there's a key/index in a hash.

and you don't need to 'arr[$1] != 1' as the sheer presence of a key in a hash is enough. you re-code the second piece for ease of reading/understanding as:

Code:

$1 in arr

I did try it w/o =1 or !=1. When I did I got no results. As soon as I put the =1 and !=1. It displayed results.

one awkward awk you've got there - must be M$ or something.....

COuld be the version of awk. More likely me and something I am doing. But it does what i need for now. And it is all thanks to your help.

Again, thank you for taking the time to help.

Dear vgersh99,

One more followup question if you do not mind? I may need to compare other files where the static data is not just in the first field arr[$1].

It could be the 1st and 3rd arr[$1,$3] or 2nd and 5th etc..

Is there a way to write this so I have one script and I can pass the value to it? So somehting like: arr[$NUMS]. Where $NUMS may equal $1 or $1,$3?

NOTE: fields are defined on the command line by the value of a variable 'fields' - list of fields separated by a ','

awk -v flds='5,6,7' -f eja.awk FILE1 FILE2

eja.awk:

Code:

BEGIN {
   SEP_list=","
   n=split(flds, fldsA, SEP_list)
}
{
   idx=""
   for(i=1; i<=n; i++) idx=(i==1) ? $(fldsA[i]) : idx SUBSEP $(fldsA[i])
   #printf("[%f]: idx->[%s]\n", FNR, idx)
}
NR==FNR { arr[idx]=1; next }
arr[idx] !=1