1

I would like to anonymize several related files by replacing the names with an id number.

Example input

file1:

2 simon
4 wesley
5 nikesh
8 sarah
1 jenny

file2:

3 casey
8 jenny
1 alan
1 jessie
5 gordon
9 bill
1 sue

file3

4 jack
2 jenny
6 wesley
8 sarah
6 gordon
1 derek
6 sue
1 jessie
2 pritesh
7 bill

Desired Output

file 1

2 user1
4 user2
5 user3
8 user4
1 user5

file2:

3 user6
8 user5
1 user7
1 user8
5 user9
9 user10
1 user11

file3

4 user12
2 user5
6 user2
8 user4
6 user9
1 user13
6 user11
1 user8
2 user14
7 user10

What I have tried

Though I expect the solution involves arrays, I am struggling to get my head around this problem.

Improve this question
2
  • 3 user6 - why? What's the algorithm for matching ID and names? Commented Jan 2, 2022 at 20:26
  • The first number is unrelated to user[n] Users are just numbered in sequence. user6 just refers to the 6th user. The following user in the file is user5 as this user (name) is a duplicate of the fifth user. Commented Jan 2, 2022 at 20:33

2 Answers 2

Reset to default
3

You could do something like

awk '!($2 in a){a[$2] = "user"(++n)} {$2 = a[$2]; print > (FILENAME".anon")}' file1 file2 file3

  • if key $2 is not already in array a, then add it, with value being the concatenation of string user and prefix-incremented variable ++n

  • replace the value of $2 with its corresponding value from array a, then print the result to a new file whose name is the concatenation of the current filename plus some arbitrary suffix

Improve this answer
2
  • That appears to do exactly what I was looking for. If you have the time, I would be interested to know how it works. Commented Jan 2, 2022 at 21:13
  • 1
    @BumblingBadger I've added a brief explanation Commented Jan 2, 2022 at 22:30
0

The following is essentially steeldriver's answer, reformulated as a single awk code block (in fact, as a single print statement):

awk '{ print $1, ($2 in m) ? m[$2] : m[$2] = "user" ++n >(FILENAME ".anon") }' file[123]

The code uses the fact that the result of an assignment is the value that was assigned.

The associative array m contains a mapping between names and the anonymous user strings. For each record of two fields read from an input file, the first output field is reproduced as-is while the second field is picked from the value associated with the second input field used as a key in the m array. If no such key exists in m, a new user string is produced and stored in m.

Just like steeldriver decided to do, I write out the generated record to a new file named like the input file, with .anon added as a filename suffix.

If you want zero-filled numbers in the user strings, you could replace m[$2] = "user" ++n by e.g. m[$2] = sprintf("user%.3d", ++n), which would give you the strings user001 through to user014 for the given data.

Improve this answer

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.