New help tweaking awk...

giannicello · #1 (**permalink**) 03-13-2009

Guys,

My awk is not very good and I'm kind of stuck.
I have a file like so:

1,a,a,ab1234,ab1234,e,f
2,a,b,cd1234,ef5678,e,f
3,a,a,cd3456,gh5678,g,h
4,a,b,ef5678,ef1234,g,h
5,a,a,cd7890,ab5678,e,f
6.a,b,cd7890,jk1234,il

I don't care about any other columns other than col4 and col5.

#1 I can't have col4 = col5 (same line) so need to remove the the line from input file and write the bad record to a 'badrec.txt' (first record in my example).
#2 Similarly, I can't have any value in col4=col5 (two diff lines) and need remove them both from input file and spit both out to a badrec.txt (rec#2 and #4 in my example).
#3 I can't have col4 equal two different things in col5..all must be rejected (rec#5 and #6 are both bad).

I trying to work with this syntax:
awk -F"," '!x[$1,$2,$3,$6,$7]++ {if ( $4 == $5 ) { print $4,$5 } else { print $0 } }' inputfile.txt

I have a problem writing the whole line out to a file? I keep getting "awk: The statement cannot be correctly parsed." The { print $4,$5 } is just temporary. This does give me half the solution for #1 but not #2 when it's on another line.

Can anyone tweak what I have to solve any of my problems?

Thanks.
Gianni

giannicello · #2 (**permalink**) 03-13-2009

So far this will do #1. I had to put badrec.txt in double quotes:

awk -F"," '!x[$1,$2,$3,$6,$7]++ {if ( $4 == $5 ) { print $0 > "badrec.txt" } else
{ print $0 } }' inputfile.txt > outputfile.txt

Trying to figure out #2 and #3 still.
Thanks.

vgersh99 · #3 (**permalink**) 03-13-2009

here's #1 & #2

nawk -f gia.awk myFile

gia.awk:

Code:

BEGIN {
  FS=OFS=","
  badFile="badRec.txt"
}

$4 == $5 {print > badFile; next}

$4 in col5 {
  print > badFile ; print col5[$4] > badFile
  delete col5[$4]
  next;
}

{ col5[$5]=$0 }

END {
  for (i in col5)
     print col5[i]
}

giannicello · #4 (**permalink**) 03-13-2009

Wow. This is awesome. I was going to handle each rule one by one but this worked very well. I guess #3 has to be done separately then, but this is neat.

Thank you.

vgersh99 · #5 (**permalink**) 03-13-2009

actually, this will only work if you had ONE pair of matching records (for #2), but you might have more than ONE pair...

Code:

BEGIN {
  FS=OFS=","
  badFile="badRec.txt"
}

$4 == $5 {print > badFile; next}

$4 in col5 {
  print > badFile ; print col5[$4] > badFile
  #delete col5[$4]
  col5del[$4]
  next;
}

{ col5[$5]=$0 }

END {
  for (i in col5)
     if (!(i in col5del)) print col5[i]
}

summer_cherry · #6 (**permalink**) 03-16-2009

Code:

#!/usr/bin/perl
open $fh,"<","a.txt";
my (%h1,%h2);
while(<$fh>){
	chomp;
	my @tmp=split(",",$_);
	next if $tmp[3] eq $tmp[4];
	if(not exists $h1{$tmp[3]}){
		$h1{$tmp[3]}=$_;
	}
	else{
		$h1{$tmp[3]}="";
	}
	$h2{$tmp[4]}=$tmp[3];
}
foreach my $k1 (keys %h1){
	if(exists $h2{$k1}){
		delete $h1{$k1};
		delete $h1{$h2{$k1}};
	}
}
foreach my $key (keys %h1){
	print $h1{$key},"\n" if $h1{$key} ne "";
}

giannicello · #7 (**permalink**) 03-18-2009

Thanks. I tried the perl command to see what it will do (which one of the rules) and I received:

Too many arguments for open at a.pl line 2, near ""a.txt";"
Execution of a.pl aborted due to compilation errors.

So I tried open(fh, "< a.txt") or die "Can't open input file\n";
It did appeared to have run but I see no output of any kind and no errors.
I'm not a perl programmer so not sure what could be wrong.
Thank you.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes	Rate This Thread
Switch to Linear Mode Hybrid Mode Switch to Threaded Mode	Rate This Thread:


	Powered by