• open file 1 and 2
• read one line from each and build a $hash{line1}=line2
• once done open file 3
• output lines from $hash{line3}

update

in case 1 and 2 are very large and 3 is not, then do it the other way round:

• read file 3 into a %seen hash
• iterate again simultanously over 1 and 2
• output lines seen in 3