An automatic regex learner?

Question

I have strings with recurrent pattern. Like that:

/socket.io/1/xhr-polling/993bcoZK7UkiqsNYpbja?t=1407502307019
/socket.io/1/xhr-polling/993bcoZK7UkiqsNYpbja?t=2222222222222
/socket.io/1/xhr-polling/993bco56465456465465a?t=333333333333

And also thing like that:

/api/bucket/53e4ce6584df65130e7ead66/data/metadata
/api/bucket/465456456456465456445456/data/metadata
/api/bucket/898989898989898989898989/data/metadata

And much more like that.

How would be the best way to find patterns in these strings and to aggregate them ?

Like getting some sort of json:

{
   pattern : "/api/bucket/*/data/metadata"
   routes : ["/api/bucket/53e4ce6584df65130e7ead66/data/metadata",
             "/api/bucket/465456456456465456445456/data/metadata",
             "/api/bucket/898989898989898989898989/data/metadata"]
}

better now ? I want to get some sort of regex from string that match a common pattern — Unitech, Aug 08 '14 at 20:44
I found something like this for Node.JS, a regex-trie that can generate a regex from strings. Now I try to find an easy way to find the matching probability of string then pass them to regex-trie. — Unitech, Aug 09 '14 at 14:58

score -1 · Answer 1 · answered Aug 09 '14 at 15:33

-1

Ok found the solution, the Aho-Corasick matching algorithm referenced in the answer is not adapted. Levenshtein distance algorithm is. If found this module for Node.JS to do that: https://github.com/jefarmstrong/sortzzy

answered Aug 09 '14 at 15:33

Unitech

5,781
5
40
47

An automatic regex learner?

1 Answers1