Comparing text files using grep/diff/whatever...

cornbb

Hello all, a bit of an easy exercise but anyway it has me stumped anyway I think... I need to compare 2 text files and output any lines that each file has in common - so if a line exists in file A and file B, it needs to end up in file C. What's the most elegant way to achieve this in Perl? I would have a preference for using Unix CLI tools like grep, diff etc.

An example of what I'm talking about:

File A

banana
orange
apple
pear

File B

mango
banana
blueberry
pear

output will consist of the lines that each file contains:

banana
pear

The files are not likely to be very large (maybe 20-30 lines each max).
I could write something that iterates through each line of File A, comparing it with each line of File B but if someone has done that already, then that would be great.

fuse

2nd answer here looks like a perl solution
http://stackoverflow.com/questions/746458/how-to-show-lines-in-common-reverse-diff

E39MSport

I'd use one file as a search pattern file using sed

# /usr/xpg4/bin/grep -f fileB fileA
banana
pear
#

Ross

Should do the trick:

grep -f fileA fileB > result

Or does it have to use Perl?

PhantomBeaker

My own take in perl (in python, I'd just use sets ) would be to use hashes.

Iterate over your first file, A, and read each line. Use that line as a key for your hash. Put anything in as the value, just so long as you have the key there.

Then you can read over each line in B and check if the key is in your hash ( I'd use something like "if defined($myhash{$line})" ) and then print out your line.

It'll partially preserve ordering in that you'll get matching lines as they occur in B, but if you were to rewrite A as:

pear
orange
apple
banana

you'd still get a result of:

banana
pear

This may matter in some applications that want lines that match in sequence (as in, if you want to match "pear" but not "banana" because in A, banana is after pear and in B it's before pear)

Hope that helps (if you haven't already cracked it)

Aoife