perl csv compare columns

I have two csv files as below

Col1       col2
ABC         12
DSA         46
KJH          22
KEW         66
ADF         78

Now how can I write a perl script for file2 that should only include values from file1 and exclude remaining rows from col
From above example I want to update the File2 as below
Col1       col2
ABC         12
KJH          22
KEW        66

If you observe "ABC, KJH, KEW" are the only 3 values that are in File1 so we should keep only those entries in File2.
So how can I achieve this using perl

Who is Participating?
wilcoxonConnect With a Mentor Commented:
Or, in script form...
use strict;
use warnings;
use Tie::File; # not strictly required but I like it *MUCH* better for editing files
# put the values from File1 into hash %keep
open IN, 'File1' or die "could not open File1: $!";
my %keep = map { chomp; $_ => 1 } <IN>;
close IN;
# Tie::File lets us edit the file as if it was a simple list
tie my @lines, 'Tie::File', 'File2' or die "could not edit File2: $!";
# only keep the lines that match hash %keep
@lines = grep { m{^(\w+)} and exists $keep{$1} } @lines

Open in new window

perl -ane 'if( 1..eof&&($^I=".bak") ){ ++$v{$F[0]} }elsif( $v{$F[0]} ){ print; }' File1 File2
@lines = grep seems to defeat the Tie::File feature of not loading the file into memory, without taking advantage of any of its other features
chomp and m{^(\w+)} may not always have consistent concepts of what is a value
The exists test would be better than the truth test if one of the values can be "0"
I would expect the grep (under Tie::File) to only load pieces of the file into memory at a time (although I haven't tested on a very large file).  I'm usually doing more complicated editing of files but I pretty much just always use Tie::File when editing a file (or when reading a file if it seems to make more sense).

I don't follow your comment about "chomp and m{^(\w+)} may not always have consistent concepts of what is a value.  Can you elaborate?

I thought exists was always (slightly) more efficient than the truth test for hashes?
If you had values like AB.C, or if there was whitespace at the end of the File1 lines, then the values found by line 6 and line 11 may not match.

I have not tested the relative efficiency of exists and truth, but it seems reasonable to expect exists to be more efficient.
exists can also be more correct than truth in the case where an existing value may be false.
All Courses

From novice to tech pro — start learning today.