Finding shortest unique prefix length using map/reduce

I have a list of strings (from documents in CouchDB).

I want to find the minimum prefix length so that all shortened strings (taking the first LEN characters) are unique.

For example:

• aabb
• aabc
• abcd

should give: LEN is three.

Is it possible to write this as 开发者_运维百科a map/reduce function ?

---

Doing it the brute force way:

MAP: Create for each input record "ABCDE" records with the keys - "A" - "AB" - "ABC" - "ABCD" - "ABCDE"

REDUCE:

• IF you have 1 value in the iterator output: "length(key)" "true"
• If you have more than one 1 value in the iterator output: "length(key)" "false"

MAP: Identity mapper

REDUCE: Output "true" if all input values are true. Else output false (or nothing);

That should result in a true for all lengths that are "unique"