databases Domain Model
| Contents | |
|---|---|
| bfilter | Contains the definition of the Bloom Filter operation and its implementations (algorithm parallelbfilter, and primitive pbfilter). |
| BFILTER | Operation that filters a stream based on a Bloom bitmap. |
| parallelbfilter | Parallel implementation of the Bloom Filter operation. |
| pbfilter | Primitive implementation of the Bloom Filter operation. |
| bloom | Contains the definition of the Bloom operation and its implementations (algorithm parallelbloom, and primitive pbloom). |
| BLOOM | Operation that creates a Bloom bitmap from a stream of tuples. |
| parallelbloom | Parallel implementation of the Bloom operation. |
| pbloom | Primitive implementation of the Bloom operation. |
| hjoin | Contains the definition of the Hash Join operation and its implementations (algorithms bloomfilterhjoin and parallelhjoin, and primitive phjoin). |
| HJOIN | Operation that joins (natural join) two relations. |
| bloomfilterhjoin | Implementation of the Hash Join operation, that starts by building a bloom filter for input stream A, and then filters input stream B before applying the Hash Join operation to both streams. |
| parallelhjoin | Parallel implementation of the Hash Join operation. |
| phjoin | Primitive implementation of the Hash Join operation. |
| m-merge-split | Contains the definition of the Bitmap Split and Bitmap Merge operations, their implementations (primitives pmsplit and pmmerge), and their optimizations. |
| IMMERGESPLIT | Operation that receives two bitmaps and returns the two bitmaps. |
| mms_mmergemsplit | Implementation of the Bitmap Merge-Hash Split operation, that starts by merging the input bitmaps, and then splits the resulting bitmap. |
| mms_identity | Implementation of the Bitmap Merge-Hash Split operation as the identity. |
| MMERGE | Operation that merges two bitmaps into a single one. |
| pmmerge | Primitive implementation of the Bitmap Merge operation. |
| MSPLIT | Operation that splits a Bloom bitmap. |
| pmsplit | Primitive implementation of the Bitmap Split operation. |
| merge-hsplit | Contains the definition of the Hash Split and Merge operations, their implementations (primitives phsplit and pmerge), and their optimizations. |
| HSPLIT | Operation that splits a stream based on an hash function. |
| phsplit | Primitive implementation of the Hash Split operation. |
| IMERGEHSPLIT | Operation that receives two streams of tuples, and returns another two streams, where the input tuples were split according to an hash function. |
| mhs_mergehsplit | Implementation of the Merge-Hash Split operation, that starts by merging the input streams, and then splits the resulting stream. |
| mhs_identity | Implementation of the Merge-Hash Split operation as the identity (possible when the input streams are already "correctly" split). |
| mhs_hsplitmerge | Implementation of the Merge-Hash Split operation, that starts by spliting the input streams, and then merges the resulting stream. |
| MERGE | Operation that merges several streams into a single one. |
| pmerge | Primitive implementation of the Merge operation. |
Rules Set: bfilter
Description
Contains the definition of the Bloom Filter operation and its implementations (algorithm parallelbfilter, and primitive pbfilter).
Contents
[Interface] BFILTER
(A : Relation) = BFILTER(A : Relation, M : BitMap, BFilterKey)
Operation that filters a stream based on a Bloom bitmap.
Its additional parameter represents the attribute that is to be used as hash key when filtering the stream.
- Inputs
- A - Stream of tuples.
- M - Bloom bitmap.
- Outputs
- A - Filtered stream of tuples.
- Additional Parameters
- BFilterKey
[Algorithm] parallelbfilter
(A : Relation) = parallelbfilter(A : Relation, M : BitMap, BFilterKey)
Parallel implementation of the Bloom Filter operation. The input stream is split (using an hash function), as well as the Bloom bitmap. Then they are processed in parallel, and finally the result of the different BFilter operations are merged.
[Primitive] pbfilter
(A : Relation) = pbfilter(A : Relation, M : BitMap, BFilterKey)
Primitive implementation of the Bloom Filter operation.
Rules Set: bloom
Description
Contains the definition of the Bloom operation and its implementations (algorithm parallelbloom, and primitive pbloom).
Contents
[Interface] BLOOM
(A : Relation, M : BitMap) = BLOOM(A : Relation, BloomKey)
Operation that creates a Bloom bitmap from a stream of tuples.
Its additional parameter represents the attribute that is to be used as hash key when creating the bitmap.
- Inputs
- A - Stream of tuples.
- Outputs
- A - Stream of tuples (equal to the input stream).
- M - Bitmap.
- Additional Parameters
- BloomKey
[Algorithm] parallelbloom
(A : Relation, M : BitMap) = parallelbloom(A : Relation, BloomKey)
Parallel implementation of the Bloom operation. The input stream is split (using an hash function), then they are processed in parallel, and finally the results of the different Bloom operations are merged.
[Primitive] pbloom
(A : Relation, M : BitMap) = pbloom(A : Relation, BloomKey)
Primitive implementation of the Bloom operation.
Rules Set: hjoin
Description
Contains the definition of the Hash Join operation and its implementations (algorithms bloomfilterhjoin and parallelhjoin, and primitive phjoin).
Contents
[Interface] HJOIN
(AB : Relation) = HJOIN(A : Relation, B : Relation, JoinKeyA, JoinKeyB)
Operation that joins (natural join) two relations.
Its additional parameters represent the attributes that are to be used as join keys.
- Inputs
- A - Left stream of tuples to be joined.
- B - Right stream of tuples to be joined.
- Outputs
- AB - The stream of tuples that results from the join operation.
- Additional Parameters
- JoinKeyA, JoinKeyB
[Algorithm] bloomfilterhjoin
(AB : Relation) = bloomfilterhjoin(A : Relation, B : Relation, JoinKeyA, JoinKeyB)
Implementation of the Hash Join operation, that starts by building a bloom filter for input stream A, and then filters input stream B before applying the Hash Join operation to both streams.
[Algorithm] parallelhjoin
(AB : Relation) = parallelhjoin(A : Relation, B : Relation, JoinKeyA, JoinKeyB)
Parallel implementation of the Hash Join operation. Each input stream is split (using an hash function), then they are joined in parallel, and finally the result of the different Hash Joins are merged.
[Primitive] phjoin
(AB : Relation) = phjoin(A : Relation, B : Relation, JoinKeyA, JoinKeyB)
Primitive implementation of the Hash Join operation.
Rules Set: m-merge-split
Description
Contains the definition of the Bitmap Split and Bitmap Merge operations, their implementations (primitives pmsplit and pmmerge), and their optimizations.
Contents
[Interface] IMMERGESPLIT
(M1 : BitMap, M2 : BitMap) = IMMERGESPLIT(M1 : BitMap, M2 : BitMap)
Operation that receives two bitmaps and returns the two bitmaps.
- Inputs
- M1 - Bitmap.
- M2 - Bitmap.
- Outputs
- M1 - Bitmap.
- M2 - Bitmap.
[Pattern] mms_mmergemsplit
(M1 : BitMap, M2 : BitMap) = mms_mmergemsplit(M1 : BitMap, M2 : BitMap)
Implementation of the Bitmap Merge-Hash Split operation, that starts by merging the input bitmaps, and then splits the resulting bitmap.
[Algorithm] mms_identity
(M1 : BitMap, M2 : BitMap) = mms_identity(M1 : BitMap, M2 : BitMap)
Implementation of the Bitmap Merge-Hash Split operation as the identity.
Rules Set: merge-hsplit
Description
Contains the definition of the Hash Split and Merge operations, their implementations (primitives phsplit and pmerge), and their optimizations.
Contents
[Interface] HSPLIT
(A1 : Relation, A2 : Relation) = HSPLIT(A : Relation, SplitKey)
Operation that splits a stream based on an hash function.
Its additional parameter represents the attribute to be used to choose the output stream.
- Inputs
- A - Stream of tuples.
- Outputs
- A1 - Streams of tuples, resulting from splitting the input stream.
- A2 - Streams of tuples, resulting from splitting the input stream.
- Additional Parameters
- SplitKey
[Primitive] phsplit
(A1 : Relation, A2 : Relation) = phsplit(A : Relation, SplitKey)
Primitive implementation of the Hash Split operation.
[Interface] IMERGEHSPLIT
(A1 : Relation, A2 : Relation) = IMERGEHSPLIT(A1 : Relation, A2 : Relation, SplitKey)
Operation that receives two streams of tuples, and returns another two streams, where the input tuples were split according to an hash function.
Its additional parameter represents the attribute to be used when splitting the streams.
- Inputs
- A1 - Stream of tuples.
- A2 - Stream of tuples.
- Outputs
- A1 - Stream of tuples.
- A2 - Stream of tuples.
- Additional Parameters
- SplitKey
[Pattern] mhs_mergehsplit
(A1 : Relation, A2 : Relation) = mhs_mergehsplit(A1 : Relation, A2 : Relation, SplitKey)
Implementation of the Merge-Hash Split operation, that starts by merging the input streams, and then splits the resulting stream.
[Algorithm] mhs_identity
(A1 : Relation, A2 : Relation) = mhs_identity(A1 : Relation, A2 : Relation, SplitKey)
Implementation of the Merge-Hash Split operation as the identity (possible when the input streams are already "correctly" split).
[Pattern] mhs_hsplitmerge
(A1 : Relation, A2 : Relation) = mhs_hsplitmerge(A1 : Relation, A2 : Relation, SplitKey)
Implementation of the Merge-Hash Split operation, that starts by spliting the input streams, and then merges the resulting stream.
[Interface] MERGE
(A : Relation) = MERGE(A1 : Relation, A2 : Relation)
Operation that merges several streams into a single one.
- Inputs
- A1 - Stream of tuples.
- A2 - Stream of tuples.
- Outputs
- A - Stream of tuples, resulting from merging the input streams.
[Primitive] pmerge
(A : Relation) = pmerge(A1 : Relation, A2 : Relation)
Primitive implementation of the Merge operation.