Linux bash command that returns where two strings differ

Question

So, I've been googling around, and also searching in more detail on stack overflow, but I just can't seem to find an easy way of doing exactly this:

I want to know in what way two strings (without whitespace) differ, and simply print what that exact difference is.

E.g.:

Input 1 > "Chocolatecakeflavour"
Input 2 > "Chocolateflavour"

Output: "cake"

I've tried doing this with diff and dwdiff, cmp, and other known bash commands that popped into mind, but I just couldn't get this exact result.

Any ideas?

That is a hard problem, depending on the amount of differences and the length of the input text. Have a look at diff, match and patch algorithms they might set you on the right path. — Steve
– Steve, Commented Oct 23, 2014 at 9:45
What should the output be if input 1 is "foo" and input 2 is "bar"? — Anton
– Anton, Commented Oct 23, 2014 at 9:49

anubhava · Accepted Answer · 2014-10-23 10:07:14Z

9

You can use diff with fold and awk like ths:

s="Chocolatecakeflavour"
r="Chocolateflavour"

diff <(fold -w1 <<< "$s") <(fold -w1 <<< "$r") | awk '/[<>]/{printf $2}'
cake

fold -w1 is to split input string character by character (one in each line)
diff is to get difference in both lists (1 char in each line)
awk '/[<>]/{printf $2}' is to suppress < OR > from diff'e output and print everything in same line

EDIT: As per OP's comments below if strings are in different lines of a file then use:

f=file
diff <(fold -w1 <(sed '2q;d' $f)) <(fold -w1 <(sed '3q;d' $f)) | awk '/[<>]/{printf $2}'
cake

edited Oct 23, 2014 at 10:07

answered Oct 23, 2014 at 9:47

anubhava

791k67 gold badges604 silver badges671 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

user2810298 Over a year ago

Somehow, for my specific needs this doesn't do what I want. (although I dont understand how) My input 1 and input 2 are on different lines in a file, so I use sed '2q;d' and sed '3q;d' to get them from the file like this: diff --normal <(fold -w1 <<< <(echo sed '2q;d' AdaptiveMutants)) <(fold -w1 <<< <(echo sed '3q;d' AdaptiveMutants)) | awk '/[<>]/{printf $2}' Somehow though, this returns a '32', which is definitely not part of the input strings.

Collectives™ on Stack Overflow

Linux bash command that returns where two strings differ

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related