tarantool bloomfilter concept

bloomfilter_concept_example.lua

local bloom = require('bloomfilter')

local bloom_options = {size = 512, hash = 3}
local bf_instance = bloom.new(bloom_options)

local init_state = nil
local somedata = "Data"
local somedata2 = "Data2"

-- now the filter is filled with zeros
bf_instance:merge(init_state)

bf_instance:add(somedata, somedata2)    -- same as bf_instance:merge(bf_instance.hash(somedata), bf_instance.hash(somedata))

-- exists : bool
local exists = bf_instance:intersect(bf_instance.hash("Data"))  -- true

------------ 
-- Integrate with tarantool box

local bloom_box = require('bloomfilter.box')

-- create common space
local my_space = box.schema.space.create('my_bloom_filter', {   temporary = false , if_not_exists = true })
my_space:create_index('primary', { type = 'TREE', unique = true, parts = { 1, 'unsigned' }, if_not_exists = true })


-- make space with bloomfilter index
local filed_num = 2
local index_name = 'bloom_bitset'
local my_bloom_space = bloom_box.create(bloom_options, my_space, filed_num, index_name)  -- gererate: -- my_space:create_index('bloom_bitset', { unique=false, type='BITSET', parts = {2, 'string'}})


local tuple_primary_index_key = {1}

-- add new data to bloom filter 
my_bloom_space:merge(tuple_primary_index_key, bf_instance) -- insert or update space  --  box.space.my_bloom_filter:update( {id}, { {'|', 2, tostring(bf_instance) } } )

-- iterate over tuples which may contain "Data"
for _, tuple in my_bloom_space:pairs( bf_instance.hash("Data") , { iterator = box.index.BITS_ALL_SET }) do   -- box.space.my_bloom_filter.index.bloom_bitset:pairs(...
  -- do some with tuple 
end	

-- etc ...