HOW-TO: Set-up, configure and use Riak Search on existing buckets with Python

1. Enable search on nodes

Riak Search is enabled in the app.config file. Simply change the setting to “true” in Riak Search Config section (shown below).

%% Riak Search Config  
{riak_search, [  
           %% To enable Search functionality set this 'true'.  
           {enabled, false}  
          ]},

Restart riak node:

$ sudo riak restart

2. Enable bucket for search

From Riak docs:

$ search-cmd install my_bucket_name

3. Re-load existing data for indexing

Because we just added a post-commit hook the data already in there won't be in the index (sadtrombone.wav)

Warning: Probably not wise to do this if you have millions of keys already ... :)

With simple Python script:

import riak
rc = riak.RiakClient(host="127.0.0.1", port=8087, prefix="riak", transport_class=riak.RiakPbcTransport)
keys = rc.bucket("my_bucket_name").get_keys()
for key in keys:
	key1 = rc.bucket("my_bucket_name").get(key)
	key1.store()

4. Manually check the result

Using the command line (docs):

Example json data in my_bucket_name:

{
	"titlet":"Laptop bag",
	"price":10.0
}

Test command:

$ search-cmd search my_bucket_name "title:Laptop"

5. Retrieving results with Python

Simple:

import riak
rc = riak.RiakClient(host="127.0.0.1", port=8087, prefix="riak", transport_class=riak.RiakPbcTransport)
search_query = rc.search("my_bucket_name", "title:Laptop")
    for result in search_query.run():
        key = result.get()
        key_data = key.get_data()
        print "Loaded key:", key.get_key()
        print "Key data:\n", key_data

Just return the field we search on:

import riak
rc = riak.RiakClient(host="127.0.0.1", port=8087, prefix="riak", transport_class=riak.RiakPbcTransport)
field = "title"
keyword = "Laptop"
search_query = rc.search("my_bucket_name", field + ":" + keyword)
for result in search_query.run():
    key = result.get()
    key_data = key.get_data()
    print "Loaded key:", key.get_key()
    print field + ":", key_data[field]