greg/hanoidb
1
0
Fork 0
Code Issues 10 Pull requests 3 Projects Releases Packages Wiki Activity

Bloom filter: replace array with faster custom representation.

Browse source
This commit is contained in:
Erik Søe Sørensen 2012-09-25 17:12:10 +02:00
parent 1c6e0df15c
commit ed65b5a468
2 changed files with 47 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
src/hanoidb_bloom.erl
View file

@ -188,7 +188,7 @@ set_bits(Mask, I1, I, [H|T], Acc) ->
%%%========== Dispatch to appropriate representation:
bitmask_new(LogN) ->
if LogN >= 20 -> % Use sparse representation.
bitarray_new(1 bsl LogN);
hanoidb_sparse_bitmap:new(LogN);
true -> % Use dense representation.
hanoidb_dense_bitmap:new(1 bsl LogN)
end.
@ -196,6 +196,7 @@ bitmask_new(LogN) ->
bitmask_set(I, BM) ->
case element(1,BM) of
array -> bitarray_set(I, BM);
sparse_bitmap -> hanoidb_sparse_bitmap:set(I, BM);
dense_bitmap_ets -> hanoidb_dense_bitmap:set(I, BM);
dense_bitmap ->
%% Surprise - we need to mutate a built representation:
@ -206,12 +207,14 @@ bitmask_set(I, BM) ->
bitmask_build(BM) ->
case element(1,BM) of
array -> BM;
sparse_bitmap -> BM;
dense_bitmap_ets -> hanoidb_dense_bitmap:build(BM)
end.
bitmask_get(I, BM) ->
case element(1,BM) of
array -> bitarray_get(I, BM);
sparse_bitmap -> hanoidb_sparse_bitmap:member(I, BM);
dense_bitmap_ets -> hanoidb_dense_bitmap:member(I, BM);
dense_bitmap -> hanoidb_dense_bitmap:member(I, BM)
end.

43
src/hanoidb_sparse_bitmap.erl Normal file
View file

@ -0,0 +1,43 @@
-module(hanoidb_sparse_bitmap).
-export([new/1, set/2, member/2]).
-define(REPR_NAME, sparse_bitmap).
new(Bits) when is_integer(Bits), Bits>0 ->
{?REPR_NAME, Bits, []}.
set(N, {?REPR_NAME, Bits, Tree}) ->
{?REPR_NAME, Bits, set_to_tree(N, 1 bsl (Bits-1), Tree)}.
set_to_tree(N, HighestBit, Mask) when HighestBit<32 ->
Nbit = 1 bsl N,
case Mask of
[]-> Nbit;
_ -> Nbit bor Mask
end;
set_to_tree(N, _HighestBit, []) -> N;
set_to_tree(N, HighestBit, [TLo|THi]) ->
pushdown(N, HighestBit, TLo, THi);
set_to_tree(N, _HighestBit, N) -> N;
set_to_tree(N, HighestBit, M) when is_integer(M) ->
set_to_tree(N, HighestBit, pushdown(M, HighestBit, [], [])).
pushdown(N, HighestBit, TLo, THi) ->
NHigh = N band HighestBit,
if NHigh =:= 0 -> [set_to_tree(N, HighestBit bsr 1, TLo) | THi];
true -> [TLo | set_to_tree(N bxor NHigh, HighestBit bsr 1, THi)]
end.
member(N, {?REPR_NAME, Bits, Tree}) ->
member_in_tree(N, 1 bsl (Bits-1), Tree).
member_in_tree(_N, _HighestBit, []) -> false;
member_in_tree(N, HighestBit, Mask) when HighestBit<32 ->
Nbit = 1 bsl N,
Nbit band Mask > 0;
member_in_tree(N, _HighestBit, M) when is_integer(M) -> N =:= M;
member_in_tree(N, HighestBit, [TLo|THi]) ->
NHigh = N band HighestBit,
if NHigh =:= 0 -> member_in_tree(N, HighestBit bsr 1, TLo);
true -> member_in_tree(N bxor NHigh, HighestBit bsr 1, THi)
end.