Prefix aggregation

Hi,

This topic was already discussed in:

Although there was no solution/workaround in the discussion, I am asking the question again here.

Is there a way to have prefix buckets for aggregation, for example (inspired from the github discussion), the results would be like:

{
     buckets: [
    {
       "key": "/blog",
       "doc_count": 1000
    } ,
    {
       "key": "/user",
       "doc_count": 100
    }
  ]
}

/blog and /user are the keys of prefix buckets.

I didn't find any way to do this with the filtered aggregation.