comparing contents of two files (field comparison)

[AwkRookie]

I want to write a script which will compare field(s) from two different files. The data in these fields is numeric and I want to identify the row number and data element where there is a match.

I have sketched out some pseudo-code which may help explain my question; my explanations are denoted by **

(1) Use the following:
awk '{print $<field# from file#1>}' file#1>f1

* put field # from file #1 into output file f1

awk '{print $<field# from file#2>}' file#2>f2

* put field # from file #2 into output file fw

awk 'compare_file_fields' f1 f2

* run the script below using files f1 and f2

BEGIN

for (a=1; a<=NF, a++)

x(a)={ print f1 $<field# from file#1>}
y(a)={ print f2 $<field# from file#2>}

* put field elements from file 1 and file 2 into variables x and y
{
for (i=1; i<=NF; i++)
{
for (j=1; j<=NF; j++)
{if (x(j)==y(i))
<print data element from x that matches y and also print the row number>


** check to see if any element from x matches any element from y and print out that element along with the row # in file 2 where the match occurred

}
}
}

What is the best way to accomplish this numeric comparison of field elements between two different files?

 

[marsd]

Just data samples for me please, and a before and after snapshot.
I can't wade through pcode

 

[AwkRookie]

marsd,

Sorry.

Here's my example:

file 1
------
contains fields file1_field1 thru file1_field4 with numeric data (double-precision) as:


file1_field1 file1_field2 file1_field3 file1_field4
16:00 10 20 30
16:10 40 50 60
16:20 12 23 14

file 2
------
contains fields file2_field1 thru file2_field4 with numeric data (double-precision) as:

file2_field1 file2_field2 file2_field3 file2_field4
16:00 10 22 31
16:10 20 50 60
16:20 14 29 13

Now, I want to see if row 1, field 3 from file #1 matches row 1, field 3 from file #2 OR row2, field 3 from file #2 OR rowN, field 3, from file #2 (where N is the number of rows in file 2). Once I check to see if row 1, field 3 from file #1 matches all of the corresponding elements in field 3 of file #2, I would index to row 2. Now, I check to see if row 2, field 3 from file #1 matches row 1, field 3 from file #2 OR row 2, field 3 from file #2, and so on.

The output should be:

row 1, field 2 match between files #1 and #2 at row number 1, field #2 (both have a value of 10)

row 2, field 2 match between files #1 and #2 at row number 2, field #2 (both have a value of 50)

and so on ...

output to screen would be:

value of 10 found in row 1
value of 50 found in row 2

Also, I would like to store the row#s and values of the solution in a matrix as follows; for each value in field #1 of:

16:00 10 1
16:10 50 2
16:20 no match

The size of the output file for this script should correspond to the size of file #2, i.e., for every row in file #2, there should either be a value & row number or the text &quot;no match&quot;.

Let me know if this is a sufficient input/output description of what I am trying to do.

AwkRookie

 

[marsd]

Gawk only. I don't work with old awk or nawk.

function compare(array,FLDVAL,REC,FLDNUM) {
#print &quot;AT&quot;, FLDVAL, &quot;NR=&quot;, REC, &quot;FIELDNUMBER=&quot;, FLDNUM
for (all in array) {
if (all == REC) {
v = split(array[all],loc)
if (loc[FLDNUM - 1] == FLDVAL) {
printf &quot;Matching value row %d, field = %d, fieldval = %d:\n&quot;, REC, FLDNUM ,FLDVAL
return &quot;1&quot;
}
}
}
return &quot;0&quot;
}


BEGIN {
x = 1;
while ((getline < &quot;k1.txt&quot > 0) {
arr1[x++] = $2&quot; &quot;$3&quot; &quot;$4
}
close(&quot;/home/k1.txt&quot
}

{
for (m=2 ; m <= NF ; m++) {
#print &quot;Field no:&quot;, m , &quot;value=&quot;, $m
compare(arr1,$m,NR,m)
}
}

Output with your data sample:
Matching value row 1, field = 2, fieldval = 10:
Matching value row 2, field = 3, fieldval = 50:
Matching value row 2, field = 4, fieldval = 60: