mirror of
https://github.com/pgvector/pgvector.git
synced 2026-07-22 12:07:34 +08:00
Compare commits
20 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
71ee682ed4 | ||
|
|
b64a1482d9 | ||
|
|
a5f2d70bc2 | ||
|
|
f3fcb5e005 | ||
|
|
3a6e0afb9c | ||
|
|
183d50bdbd | ||
|
|
bd776fee68 | ||
|
|
d30b113e4b | ||
|
|
fd3200f718 | ||
|
|
02c815d876 | ||
|
|
4b2a7cc49d | ||
|
|
da0ff998e9 | ||
|
|
cb36e24289 | ||
|
|
b1d0d4c7a3 | ||
|
|
1dc6514b66 | ||
|
|
6c53f7ca02 | ||
|
|
0d35a14198 | ||
|
|
3ea2ce89be | ||
|
|
62350b1589 | ||
|
|
dd57309281 |
1
.github/workflows/build.yml
vendored
1
.github/workflows/build.yml
vendored
@@ -73,6 +73,7 @@ jobs:
|
|||||||
postgres-version: 14
|
postgres-version: 14
|
||||||
- run: |
|
- run: |
|
||||||
call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars64.bat" && ^
|
call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars64.bat" && ^
|
||||||
|
cd %TEMP% && ^
|
||||||
nmake /NOLOGO /F Makefile.win && ^
|
nmake /NOLOGO /F Makefile.win && ^
|
||||||
nmake /NOLOGO /F Makefile.win install && ^
|
nmake /NOLOGO /F Makefile.win install && ^
|
||||||
nmake /NOLOGO /F Makefile.win installcheck && ^
|
nmake /NOLOGO /F Makefile.win installcheck && ^
|
||||||
|
|||||||
@@ -1,3 +1,11 @@
|
|||||||
|
## 0.7.0 (unreleased)
|
||||||
|
|
||||||
|
- Added subscript function for vectors
|
||||||
|
|
||||||
|
## 0.6.2 (unreleased)
|
||||||
|
|
||||||
|
- Reduced lock contention with parallel HNSW index builds
|
||||||
|
|
||||||
## 0.6.1 (2024-03-04)
|
## 0.6.1 (2024-03-04)
|
||||||
|
|
||||||
- Fixed error with `ANALYZE` and vectors with different dimensions
|
- Fixed error with `ANALYZE` and vectors with different dimensions
|
||||||
|
|||||||
116
README.md
116
README.md
@@ -26,7 +26,7 @@ make
|
|||||||
make install # may need sudo
|
make install # may need sudo
|
||||||
```
|
```
|
||||||
|
|
||||||
See the [installation notes](#installation-notes) if you run into issues
|
See the [installation notes](#installation-notes---linux-and-mac) if you run into issues
|
||||||
|
|
||||||
You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), [APT](#apt), [Yum](#yum), [pkg](#pkg), or [conda-forge](#conda-forge), and it comes preinstalled with [Postgres.app](#postgresapp) and many [hosted providers](#hosted-postgres). There are also instructions for [GitHub Actions](https://github.com/pgvector/setup-pgvector).
|
You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), [APT](#apt), [Yum](#yum), [pkg](#pkg), or [conda-forge](#conda-forge), and it comes preinstalled with [Postgres.app](#postgresapp) and many [hosted providers](#hosted-postgres). There are also instructions for [GitHub Actions](https://github.com/pgvector/setup-pgvector).
|
||||||
|
|
||||||
@@ -44,12 +44,15 @@ Then use `nmake` to build:
|
|||||||
|
|
||||||
```cmd
|
```cmd
|
||||||
set "PGROOT=C:\Program Files\PostgreSQL\16"
|
set "PGROOT=C:\Program Files\PostgreSQL\16"
|
||||||
|
cd %TEMP%
|
||||||
git clone --branch v0.6.1 https://github.com/pgvector/pgvector.git
|
git clone --branch v0.6.1 https://github.com/pgvector/pgvector.git
|
||||||
cd pgvector
|
cd pgvector
|
||||||
nmake /F Makefile.win
|
nmake /F Makefile.win
|
||||||
nmake /F Makefile.win install
|
nmake /F Makefile.win install
|
||||||
```
|
```
|
||||||
|
|
||||||
|
See the [installation notes](#installation-notes---windows) if you run into issues
|
||||||
|
|
||||||
You can also install it with [Docker](#docker) or [conda-forge](#conda-forge).
|
You can also install it with [Docker](#docker) or [conda-forge](#conda-forge).
|
||||||
|
|
||||||
## Getting Started
|
## Getting Started
|
||||||
@@ -410,13 +413,39 @@ You can use [Reciprocal Rank Fusion](https://github.com/pgvector/pgvector-python
|
|||||||
|
|
||||||
## Performance
|
## Performance
|
||||||
|
|
||||||
|
### Tuning
|
||||||
|
|
||||||
|
Use a tool like [PgTune](https://pgtune.leopard.in.ua/) to set initial values for Postgres server parameters.
|
||||||
|
|
||||||
|
### Loading
|
||||||
|
|
||||||
|
Use `COPY` for bulk loading data ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/bulk_loading.py)).
|
||||||
|
|
||||||
|
```sql
|
||||||
|
COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
|
||||||
|
```
|
||||||
|
|
||||||
|
Add any indexes *after* loading the initial data for best performance.
|
||||||
|
|
||||||
|
### Indexing
|
||||||
|
|
||||||
|
See index build time for [HNSW](#index-build-time) and [IVFFlat](#index-build-time-1).
|
||||||
|
|
||||||
|
In production environments, create indexes concurrently to avoid blocking writes.
|
||||||
|
|
||||||
|
```sql
|
||||||
|
CREATE INDEX CONCURRENTLY ...
|
||||||
|
```
|
||||||
|
|
||||||
|
### Querying
|
||||||
|
|
||||||
Use `EXPLAIN ANALYZE` to debug performance.
|
Use `EXPLAIN ANALYZE` to debug performance.
|
||||||
|
|
||||||
```sql
|
```sql
|
||||||
EXPLAIN ANALYZE SELECT * FROM items ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
|
EXPLAIN ANALYZE SELECT * FROM items ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
|
||||||
```
|
```
|
||||||
|
|
||||||
### Exact Search
|
#### Exact Search
|
||||||
|
|
||||||
To speed up queries without an index, increase `max_parallel_workers_per_gather`.
|
To speed up queries without an index, increase `max_parallel_workers_per_gather`.
|
||||||
|
|
||||||
@@ -430,7 +459,7 @@ If vectors are normalized to length 1 (like [OpenAI embeddings](https://platform
|
|||||||
SELECT * FROM items ORDER BY embedding <#> '[3,1,2]' LIMIT 5;
|
SELECT * FROM items ORDER BY embedding <#> '[3,1,2]' LIMIT 5;
|
||||||
```
|
```
|
||||||
|
|
||||||
### Approximate Search
|
#### Approximate Search
|
||||||
|
|
||||||
To speed up queries with an IVFFlat index, increase the number of inverted lists (at the expense of recall).
|
To speed up queries with an IVFFlat index, increase the number of inverted lists (at the expense of recall).
|
||||||
|
|
||||||
@@ -438,7 +467,7 @@ To speed up queries with an IVFFlat index, increase the number of inverted lists
|
|||||||
CREATE INDEX ON items USING ivfflat (embedding vector_l2_ops) WITH (lists = 1000);
|
CREATE INDEX ON items USING ivfflat (embedding vector_l2_ops) WITH (lists = 1000);
|
||||||
```
|
```
|
||||||
|
|
||||||
## Vacuuming
|
### Vacuuming
|
||||||
|
|
||||||
Vacuuming can take a while for HNSW indexes. Speed it up by reindexing first.
|
Vacuuming can take a while for HNSW indexes. Speed it up by reindexing first.
|
||||||
|
|
||||||
@@ -447,6 +476,41 @@ REINDEX INDEX CONCURRENTLY index_name;
|
|||||||
VACUUM table_name;
|
VACUUM table_name;
|
||||||
```
|
```
|
||||||
|
|
||||||
|
## Monitoring
|
||||||
|
|
||||||
|
Monitor performance with [pg_stat_statements](https://www.postgresql.org/docs/current/pgstatstatements.html) (be sure to add it to `shared_preload_libraries`).
|
||||||
|
|
||||||
|
```sql
|
||||||
|
CREATE EXTENSION pg_stat_statements;
|
||||||
|
```
|
||||||
|
|
||||||
|
Get the most time-consuming queries with:
|
||||||
|
|
||||||
|
```sql
|
||||||
|
SELECT query, calls, ROUND((total_plan_time + total_exec_time) / calls) AS avg_time_ms,
|
||||||
|
ROUND((total_plan_time + total_exec_time) / 60000) AS total_time_min
|
||||||
|
FROM pg_stat_statements ORDER BY total_plan_time + total_exec_time DESC LIMIT 20;
|
||||||
|
```
|
||||||
|
|
||||||
|
Note: Replace `total_plan_time + total_exec_time` with `total_time` for Postgres < 13
|
||||||
|
|
||||||
|
Monitor recall by comparing results from approximate search with exact search.
|
||||||
|
|
||||||
|
```sql
|
||||||
|
BEGIN;
|
||||||
|
SET LOCAL enable_indexscan = off; -- use exact search
|
||||||
|
SELECT ...
|
||||||
|
COMMIT;
|
||||||
|
```
|
||||||
|
|
||||||
|
## Scaling
|
||||||
|
|
||||||
|
Scale pgvector the same way you scale Postgres.
|
||||||
|
|
||||||
|
Scale vertically by increasing memory, CPU, and storage on a single instance. Use existing tools to [tune parameters](#tuning) and [monitor performance](#monitoring).
|
||||||
|
|
||||||
|
Scale horizontally with [replicas](https://www.postgresql.org/docs/current/hot-standby.html), or use [Citus](https://github.com/citusdata/citus) or another approach for sharding ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/citus.py)).
|
||||||
|
|
||||||
## Languages
|
## Languages
|
||||||
|
|
||||||
Use pgvector from any language with a Postgres client. You can even generate and store vectors in one language and query them in another.
|
Use pgvector from any language with a Postgres client. You can even generate and store vectors in one language and query them in another.
|
||||||
@@ -552,7 +616,17 @@ SELECT pg_size_pretty(pg_relation_size('index_name'));
|
|||||||
|
|
||||||
#### Why isn’t a query using an index?
|
#### Why isn’t a query using an index?
|
||||||
|
|
||||||
The cost estimation in pgvector < 0.4.3 does not always work well with the planner. You can encourage the planner to use an index for a query with:
|
The query needs to have an `ORDER BY` and `LIMIT`, and the `ORDER BY` must be the result of a distance operator, not an expression.
|
||||||
|
|
||||||
|
```sql
|
||||||
|
-- index
|
||||||
|
ORDER BY embedding <=> '[3,1,2]' LIMIT 5;
|
||||||
|
|
||||||
|
-- no index
|
||||||
|
ORDER BY 1 - (embedding <=> '[3,1,2]') DESC LIMIT 5;
|
||||||
|
```
|
||||||
|
|
||||||
|
You can encourage the planner to use an index for a query with:
|
||||||
|
|
||||||
```sql
|
```sql
|
||||||
BEGIN;
|
BEGIN;
|
||||||
@@ -581,6 +655,10 @@ or choose to store vectors inline:
|
|||||||
ALTER TABLE items ALTER COLUMN embedding SET STORAGE PLAIN;
|
ALTER TABLE items ALTER COLUMN embedding SET STORAGE PLAIN;
|
||||||
```
|
```
|
||||||
|
|
||||||
|
#### Why are there less results for a query after adding an HNSW index?
|
||||||
|
|
||||||
|
Results are limited by the size of the dynamic candidate list (`hnsw.ef_search`). There may be even less results due to dead tuples or filtering conditions in the query. We recommend setting `hnsw.ef_search` to at least twice the `LIMIT` of the query. If you need more than 500 results, use an IVFFlat index instead.
|
||||||
|
|
||||||
#### Why are there less results for a query after adding an IVFFlat index?
|
#### Why are there less results for a query after adding an IVFFlat index?
|
||||||
|
|
||||||
The index was likely created with too little data for the number of lists. Drop the index until the table has more data.
|
The index was likely created with too little data for the number of lists. Drop the index until the table has more data.
|
||||||
@@ -589,6 +667,8 @@ The index was likely created with too little data for the number of lists. Drop
|
|||||||
DROP INDEX index_name;
|
DROP INDEX index_name;
|
||||||
```
|
```
|
||||||
|
|
||||||
|
Results can also be limited by the number of probes (`ivfflat.probes`).
|
||||||
|
|
||||||
## Reference
|
## Reference
|
||||||
|
|
||||||
### Vector Type
|
### Vector Type
|
||||||
@@ -624,7 +704,7 @@ Function | Description | Added
|
|||||||
avg(vector) → vector | average |
|
avg(vector) → vector | average |
|
||||||
sum(vector) → vector | sum | 0.5.0
|
sum(vector) → vector | sum | 0.5.0
|
||||||
|
|
||||||
## Installation Notes
|
## Installation Notes - Linux and Mac
|
||||||
|
|
||||||
### Postgres Location
|
### Postgres Location
|
||||||
|
|
||||||
@@ -666,12 +746,24 @@ If compilation fails and the output includes `warning: no such sysroot directory
|
|||||||
|
|
||||||
### Portability
|
### Portability
|
||||||
|
|
||||||
|
By default, pgvector compiles with `-march=native` on some platforms for best performance. However, this can lead to `Illegal instruction` errors if trying to run the compiled extension on a different machine.
|
||||||
|
|
||||||
To compile for portability, use:
|
To compile for portability, use:
|
||||||
|
|
||||||
```sh
|
```sh
|
||||||
make OPTFLAGS=""
|
make OPTFLAGS=""
|
||||||
```
|
```
|
||||||
|
|
||||||
|
## Installation Notes - Windows
|
||||||
|
|
||||||
|
### Missing Header
|
||||||
|
|
||||||
|
If compilation fails with `Cannot open include file: 'postgres.h': No such file or directory`, make sure `PGROOT` is correct.
|
||||||
|
|
||||||
|
### Permissions
|
||||||
|
|
||||||
|
If installation fails with `Access is denied`, re-run the installation instructions as an administrator.
|
||||||
|
|
||||||
## Additional Installation Methods
|
## Additional Installation Methods
|
||||||
|
|
||||||
### Docker
|
### Docker
|
||||||
@@ -856,6 +948,12 @@ make installcheck REGRESS=functions # regression test
|
|||||||
make prove_installcheck PROVE_TESTS=test/t/001_ivfflat_wal.pl # TAP test
|
make prove_installcheck PROVE_TESTS=test/t/001_ivfflat_wal.pl # TAP test
|
||||||
```
|
```
|
||||||
|
|
||||||
|
To enable assertions:
|
||||||
|
|
||||||
|
```sh
|
||||||
|
make clean && PG_CFLAGS="-DUSE_ASSERT_CHECKING" make && make install
|
||||||
|
```
|
||||||
|
|
||||||
To enable benchmarking:
|
To enable benchmarking:
|
||||||
|
|
||||||
```sh
|
```sh
|
||||||
@@ -868,12 +966,6 @@ To show memory usage:
|
|||||||
make clean && PG_CFLAGS="-DHNSW_MEMORY -DIVFFLAT_MEMORY" make && make install
|
make clean && PG_CFLAGS="-DHNSW_MEMORY -DIVFFLAT_MEMORY" make && make install
|
||||||
```
|
```
|
||||||
|
|
||||||
To enable assertions:
|
|
||||||
|
|
||||||
```sh
|
|
||||||
make clean && PG_CFLAGS="-DUSE_ASSERT_CHECKING" make && make install
|
|
||||||
```
|
|
||||||
|
|
||||||
To get k-means metrics:
|
To get k-means metrics:
|
||||||
|
|
||||||
```sh
|
```sh
|
||||||
|
|||||||
7
sql/vector--0.6.2--0.7.0.sql
Normal file
7
sql/vector--0.6.2--0.7.0.sql
Normal file
@@ -0,0 +1,7 @@
|
|||||||
|
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
|
||||||
|
\echo Use "ALTER EXTENSION vector UPDATE TO '0.7.0'" to load this file. \quit
|
||||||
|
|
||||||
|
CREATE FUNCTION vector_subscript(internal) RETURNS internal
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
ALTER TYPE vector SET (SUBSCRIPT = vector_subscript);
|
||||||
@@ -20,12 +20,16 @@ CREATE FUNCTION vector_recv(internal, oid, integer) RETURNS vector
|
|||||||
CREATE FUNCTION vector_send(vector) RETURNS bytea
|
CREATE FUNCTION vector_send(vector) RETURNS bytea
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION vector_subscript(internal) RETURNS internal
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
CREATE TYPE vector (
|
CREATE TYPE vector (
|
||||||
INPUT = vector_in,
|
INPUT = vector_in,
|
||||||
OUTPUT = vector_out,
|
OUTPUT = vector_out,
|
||||||
TYPMOD_IN = vector_typmod_in,
|
TYPMOD_IN = vector_typmod_in,
|
||||||
RECEIVE = vector_recv,
|
RECEIVE = vector_recv,
|
||||||
SEND = vector_send,
|
SEND = vector_send,
|
||||||
|
SUBSCRIPT = vector_subscript,
|
||||||
STORAGE = external
|
STORAGE = external
|
||||||
);
|
);
|
||||||
|
|
||||||
|
|||||||
@@ -129,7 +129,7 @@ HnswPtrDeclare(HnswNeighborArray, HnswNeighborArrayRelptr, HnswNeighborArrayPtr)
|
|||||||
HnswPtrDeclare(HnswNeighborArrayPtr, HnswNeighborsRelptr, HnswNeighborsPtr);
|
HnswPtrDeclare(HnswNeighborArrayPtr, HnswNeighborsRelptr, HnswNeighborsPtr);
|
||||||
HnswPtrDeclare(char, DatumRelptr, DatumPtr);
|
HnswPtrDeclare(char, DatumRelptr, DatumPtr);
|
||||||
|
|
||||||
typedef struct HnswElementData
|
struct HnswElementData
|
||||||
{
|
{
|
||||||
HnswElementPtr next;
|
HnswElementPtr next;
|
||||||
ItemPointerData heaptids[HNSW_HEAPTIDS];
|
ItemPointerData heaptids[HNSW_HEAPTIDS];
|
||||||
@@ -144,7 +144,7 @@ typedef struct HnswElementData
|
|||||||
BlockNumber neighborPage;
|
BlockNumber neighborPage;
|
||||||
DatumPtr value;
|
DatumPtr value;
|
||||||
LWLock lock;
|
LWLock lock;
|
||||||
} HnswElementData;
|
};
|
||||||
|
|
||||||
typedef HnswElementData * HnswElement;
|
typedef HnswElementData * HnswElement;
|
||||||
|
|
||||||
@@ -155,12 +155,12 @@ typedef struct HnswCandidate
|
|||||||
bool closer;
|
bool closer;
|
||||||
} HnswCandidate;
|
} HnswCandidate;
|
||||||
|
|
||||||
typedef struct HnswNeighborArray
|
struct HnswNeighborArray
|
||||||
{
|
{
|
||||||
int length;
|
int length;
|
||||||
bool closerSet;
|
bool closerSet;
|
||||||
HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
|
HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
|
||||||
} HnswNeighborArray;
|
};
|
||||||
|
|
||||||
typedef struct HnswPairingHeapNode
|
typedef struct HnswPairingHeapNode
|
||||||
{
|
{
|
||||||
@@ -185,6 +185,7 @@ typedef struct HnswGraph
|
|||||||
|
|
||||||
/* Entry state */
|
/* Entry state */
|
||||||
LWLock entryLock;
|
LWLock entryLock;
|
||||||
|
LWLock entryWaitLock;
|
||||||
HnswElementPtr entryPoint;
|
HnswElementPtr entryPoint;
|
||||||
|
|
||||||
/* Allocations state */
|
/* Allocations state */
|
||||||
|
|||||||
@@ -431,10 +431,15 @@ InsertTupleInMemory(HnswBuildState * buildstate, HnswElement element)
|
|||||||
HnswGraph *graph = buildstate->graph;
|
HnswGraph *graph = buildstate->graph;
|
||||||
HnswElement entryPoint;
|
HnswElement entryPoint;
|
||||||
LWLock *entryLock = &graph->entryLock;
|
LWLock *entryLock = &graph->entryLock;
|
||||||
|
LWLock *entryWaitLock = &graph->entryWaitLock;
|
||||||
int efConstruction = buildstate->efConstruction;
|
int efConstruction = buildstate->efConstruction;
|
||||||
int m = buildstate->m;
|
int m = buildstate->m;
|
||||||
char *base = buildstate->hnswarea;
|
char *base = buildstate->hnswarea;
|
||||||
|
|
||||||
|
/* Wait if another process needs exclusive lock on entry lock */
|
||||||
|
LWLockAcquire(entryWaitLock, LW_EXCLUSIVE);
|
||||||
|
LWLockRelease(entryWaitLock);
|
||||||
|
|
||||||
/* Get entry point */
|
/* Get entry point */
|
||||||
LWLockAcquire(entryLock, LW_SHARED);
|
LWLockAcquire(entryLock, LW_SHARED);
|
||||||
entryPoint = HnswPtrAccess(base, graph->entryPoint);
|
entryPoint = HnswPtrAccess(base, graph->entryPoint);
|
||||||
@@ -445,8 +450,10 @@ InsertTupleInMemory(HnswBuildState * buildstate, HnswElement element)
|
|||||||
/* Release shared lock */
|
/* Release shared lock */
|
||||||
LWLockRelease(entryLock);
|
LWLockRelease(entryLock);
|
||||||
|
|
||||||
/* Get exclusive lock */
|
/* Tell other processes to wait and get exclusive lock */
|
||||||
|
LWLockAcquire(entryWaitLock, LW_EXCLUSIVE);
|
||||||
LWLockAcquire(entryLock, LW_EXCLUSIVE);
|
LWLockAcquire(entryLock, LW_EXCLUSIVE);
|
||||||
|
LWLockRelease(entryWaitLock);
|
||||||
|
|
||||||
/* Get latest entry point after lock is acquired */
|
/* Get latest entry point after lock is acquired */
|
||||||
entryPoint = HnswPtrAccess(base, graph->entryPoint);
|
entryPoint = HnswPtrAccess(base, graph->entryPoint);
|
||||||
@@ -612,6 +619,7 @@ InitGraph(HnswGraph * graph, char *base, long memoryTotal)
|
|||||||
graph->indtuples = 0;
|
graph->indtuples = 0;
|
||||||
SpinLockInit(&graph->lock);
|
SpinLockInit(&graph->lock);
|
||||||
LWLockInitialize(&graph->entryLock, hnsw_lock_tranche_id);
|
LWLockInitialize(&graph->entryLock, hnsw_lock_tranche_id);
|
||||||
|
LWLockInitialize(&graph->entryWaitLock, hnsw_lock_tranche_id);
|
||||||
LWLockInitialize(&graph->allocatorLock, hnsw_lock_tranche_id);
|
LWLockInitialize(&graph->allocatorLock, hnsw_lock_tranche_id);
|
||||||
LWLockInitialize(&graph->flushLock, hnsw_lock_tranche_id);
|
LWLockInitialize(&graph->flushLock, hnsw_lock_tranche_id);
|
||||||
}
|
}
|
||||||
|
|||||||
181
src/vector.c
181
src/vector.c
@@ -4,11 +4,18 @@
|
|||||||
|
|
||||||
#include "catalog/pg_type.h"
|
#include "catalog/pg_type.h"
|
||||||
#include "common/shortest_dec.h"
|
#include "common/shortest_dec.h"
|
||||||
|
#include "executor/execExpr.h"
|
||||||
#include "fmgr.h"
|
#include "fmgr.h"
|
||||||
#include "hnsw.h"
|
#include "hnsw.h"
|
||||||
#include "ivfflat.h"
|
#include "ivfflat.h"
|
||||||
#include "lib/stringinfo.h"
|
#include "lib/stringinfo.h"
|
||||||
#include "libpq/pqformat.h"
|
#include "libpq/pqformat.h"
|
||||||
|
#include "nodes/makefuncs.h"
|
||||||
|
#include "nodes/nodeFuncs.h"
|
||||||
|
#include "nodes/subscripting.h"
|
||||||
|
#include "parser/parse_coerce.h"
|
||||||
|
#include "parser/parse_expr.h"
|
||||||
|
#include "parser/parse_node.h"
|
||||||
#include "port.h" /* for strtof() */
|
#include "port.h" /* for strtof() */
|
||||||
#include "utils/array.h"
|
#include "utils/array.h"
|
||||||
#include "utils/builtins.h"
|
#include "utils/builtins.h"
|
||||||
@@ -418,6 +425,180 @@ vector_send(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
|
PG_RETURN_BYTEA_P(pq_endtypsend(&buf));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Transform the subscript expressions
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
vector_subscript_transform(SubscriptingRef *sbsref, List *indirection, ParseState *pstate, bool isSlice, bool isAssignment)
|
||||||
|
{
|
||||||
|
A_Indices *ai;
|
||||||
|
Node *subexpr;
|
||||||
|
|
||||||
|
if (list_length(indirection) != 1)
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
|
||||||
|
errmsg("vector allows only one subscript"),
|
||||||
|
parser_errposition(pstate,
|
||||||
|
exprLocation((Node *) indirection))));
|
||||||
|
|
||||||
|
ai = linitial_node(A_Indices, indirection);
|
||||||
|
|
||||||
|
if (isSlice)
|
||||||
|
{
|
||||||
|
if (ai->lidx)
|
||||||
|
{
|
||||||
|
subexpr = transformExpr(pstate, ai->lidx, pstate->p_expr_kind);
|
||||||
|
/* If it's not int4 already, try to coerce */
|
||||||
|
subexpr = coerce_to_target_type(pstate,
|
||||||
|
subexpr, exprType(subexpr),
|
||||||
|
INT4OID, -1,
|
||||||
|
COERCION_ASSIGNMENT,
|
||||||
|
COERCE_IMPLICIT_CAST,
|
||||||
|
-1);
|
||||||
|
if (subexpr == NULL)
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_DATATYPE_MISMATCH),
|
||||||
|
errmsg("vector subscript must have type integer"),
|
||||||
|
parser_errposition(pstate, exprLocation(ai->lidx))));
|
||||||
|
}
|
||||||
|
else if (!ai->is_slice)
|
||||||
|
{
|
||||||
|
/* Make a constant 1 */
|
||||||
|
subexpr = (Node *) makeConst(INT4OID,
|
||||||
|
-1,
|
||||||
|
InvalidOid,
|
||||||
|
sizeof(int32),
|
||||||
|
Int32GetDatum(1),
|
||||||
|
false,
|
||||||
|
true); /* pass by value */
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
/* Slice with omitted lower bound, put NULL into the list */
|
||||||
|
subexpr = NULL;
|
||||||
|
}
|
||||||
|
sbsref->reflowerindexpr = list_make1(subexpr);
|
||||||
|
}
|
||||||
|
else
|
||||||
|
Assert(ai->lidx == NULL && !ai->is_slice);
|
||||||
|
|
||||||
|
if (ai->uidx)
|
||||||
|
{
|
||||||
|
subexpr = transformExpr(pstate, ai->uidx, pstate->p_expr_kind);
|
||||||
|
/* If it's not int4 already, try to coerce */
|
||||||
|
subexpr = coerce_to_target_type(pstate,
|
||||||
|
subexpr, exprType(subexpr),
|
||||||
|
INT4OID, -1,
|
||||||
|
COERCION_ASSIGNMENT,
|
||||||
|
COERCE_IMPLICIT_CAST,
|
||||||
|
-1);
|
||||||
|
if (subexpr == NULL)
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_DATATYPE_MISMATCH),
|
||||||
|
errmsg("array subscript must have type integer"),
|
||||||
|
parser_errposition(pstate, exprLocation(ai->uidx))));
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
/* Slice with omitted upper bound, put NULL into the list */
|
||||||
|
Assert(isSlice && ai->is_slice);
|
||||||
|
subexpr = NULL;
|
||||||
|
}
|
||||||
|
sbsref->refupperindexpr = list_make1(subexpr);
|
||||||
|
|
||||||
|
if (isSlice)
|
||||||
|
sbsref->refrestype = sbsref->refcontainertype;
|
||||||
|
else
|
||||||
|
sbsref->refrestype = FLOAT4OID;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Fetch a vector element
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
vector_subscript_fetch(ExprState *state, ExprEvalStep *op, ExprContext *econtext)
|
||||||
|
{
|
||||||
|
SubscriptingRefState *sbsrefstate = op->d.sbsref.state;
|
||||||
|
Vector *vec = DatumGetVector(*op->resvalue);
|
||||||
|
int index = DatumGetInt32(sbsrefstate->upperindex[0]);
|
||||||
|
|
||||||
|
if (index < 1 || index > vec->dim)
|
||||||
|
*op->resnull = true;
|
||||||
|
else
|
||||||
|
*op->resvalue = Float4GetDatum(vec->x[index - 1]);
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Fetch a vector slice
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
vector_subscript_fetch_slice(ExprState *state, ExprEvalStep *op, ExprContext *econtext)
|
||||||
|
{
|
||||||
|
SubscriptingRefState *sbsrefstate = op->d.sbsref.state;
|
||||||
|
|
||||||
|
if (sbsrefstate->upperprovided[0] && sbsrefstate->upperindexnull[0])
|
||||||
|
*op->resnull = true;
|
||||||
|
else if (sbsrefstate->lowerprovided[0] && sbsrefstate->lowerindexnull[0])
|
||||||
|
*op->resnull = true;
|
||||||
|
else
|
||||||
|
{
|
||||||
|
Vector *vec = DatumGetVector(*op->resvalue);
|
||||||
|
int upperindex = sbsrefstate->upperprovided[0] ? DatumGetInt32(sbsrefstate->upperindex[0]) : vec->dim;
|
||||||
|
int lowerindex = sbsrefstate->lowerprovided[0] ? DatumGetInt32(sbsrefstate->lowerindex[0]) : 1;
|
||||||
|
int dim;
|
||||||
|
Vector *result;
|
||||||
|
|
||||||
|
if (upperindex > vec->dim)
|
||||||
|
upperindex = vec->dim;
|
||||||
|
|
||||||
|
if (lowerindex < 1)
|
||||||
|
lowerindex = 1;
|
||||||
|
|
||||||
|
dim = upperindex - lowerindex + 1;
|
||||||
|
|
||||||
|
CheckDim(dim);
|
||||||
|
|
||||||
|
result = InitVector(dim);
|
||||||
|
for (int i = 0; i < dim; i++)
|
||||||
|
result->x[i] = vec->x[lowerindex + i - 1];
|
||||||
|
|
||||||
|
*op->resvalue = PointerGetDatum(result);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Set up execution state for a vector subscript operation
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
vector_exec_setup(const SubscriptingRef *sbsref, SubscriptingRefState *sbsrefstate, SubscriptExecSteps *methods)
|
||||||
|
{
|
||||||
|
methods->sbs_check_subscripts = NULL;
|
||||||
|
if (sbsrefstate->numlower != 0)
|
||||||
|
methods->sbs_fetch = vector_subscript_fetch_slice;
|
||||||
|
else
|
||||||
|
methods->sbs_fetch = vector_subscript_fetch;
|
||||||
|
methods->sbs_assign = NULL;
|
||||||
|
methods->sbs_fetch_old = NULL;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Subscript handler
|
||||||
|
*/
|
||||||
|
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_subscript);
|
||||||
|
Datum
|
||||||
|
vector_subscript(PG_FUNCTION_ARGS)
|
||||||
|
{
|
||||||
|
static const SubscriptRoutines sbsroutines = {
|
||||||
|
.transform = vector_subscript_transform,
|
||||||
|
.exec_setup = vector_exec_setup,
|
||||||
|
.fetch_strict = true, /* fetch returns NULL for NULL inputs */
|
||||||
|
.fetch_leakproof = true, /* fetch returns NULL for bad subscript */
|
||||||
|
.store_leakproof = false /* ... but assignment throws error */
|
||||||
|
};
|
||||||
|
|
||||||
|
PG_RETURN_POINTER(&sbsroutines);
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Convert vector to vector
|
* Convert vector to vector
|
||||||
* This is needed to check the type modifier
|
* This is needed to check the type modifier
|
||||||
|
|||||||
@@ -24,6 +24,112 @@ SELECT '[1e37]'::vector * '[1e37]';
|
|||||||
ERROR: value out of range: overflow
|
ERROR: value out of range: overflow
|
||||||
SELECT '[1e-37]'::vector * '[1e-37]';
|
SELECT '[1e-37]'::vector * '[1e-37]';
|
||||||
ERROR: value out of range: underflow
|
ERROR: value out of range: underflow
|
||||||
|
SELECT ('[1,2,3]'::vector)[0];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[1];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[2];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
2
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[3];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
3
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[4];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[1:1];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[1]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[1:2];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[1,2]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:4];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[-2:2];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[1,2]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:1];
|
||||||
|
ERROR: vector must have at least 1 dimension
|
||||||
|
SELECT ('[1,2,3]'::vector)[:];
|
||||||
|
vector
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[:2];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[1,2]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
[2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[:4];
|
||||||
|
vector
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[-2:];
|
||||||
|
vector
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[NULL];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[NULL:2];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:NULL];
|
||||||
|
vector
|
||||||
|
--------
|
||||||
|
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[1][1];
|
||||||
|
ERROR: vector allows only one subscript
|
||||||
SELECT '[1,2,3]'::vector = '[1,2,3]';
|
SELECT '[1,2,3]'::vector = '[1,2,3]';
|
||||||
?column?
|
?column?
|
||||||
----------
|
----------
|
||||||
|
|||||||
@@ -6,6 +6,26 @@ SELECT '[1,2,3]'::vector * '[4,5,6]';
|
|||||||
SELECT '[1e37]'::vector * '[1e37]';
|
SELECT '[1e37]'::vector * '[1e37]';
|
||||||
SELECT '[1e-37]'::vector * '[1e-37]';
|
SELECT '[1e-37]'::vector * '[1e-37]';
|
||||||
|
|
||||||
|
SELECT ('[1,2,3]'::vector)[0];
|
||||||
|
SELECT ('[1,2,3]'::vector)[1];
|
||||||
|
SELECT ('[1,2,3]'::vector)[2];
|
||||||
|
SELECT ('[1,2,3]'::vector)[3];
|
||||||
|
SELECT ('[1,2,3]'::vector)[4];
|
||||||
|
SELECT ('[1,2,3]'::vector)[1:1];
|
||||||
|
SELECT ('[1,2,3]'::vector)[1:2];
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:4];
|
||||||
|
SELECT ('[1,2,3]'::vector)[-2:2];
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:1];
|
||||||
|
SELECT ('[1,2,3]'::vector)[:];
|
||||||
|
SELECT ('[1,2,3]'::vector)[:2];
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:];
|
||||||
|
SELECT ('[1,2,3]'::vector)[:4];
|
||||||
|
SELECT ('[1,2,3]'::vector)[-2:];
|
||||||
|
SELECT ('[1,2,3]'::vector)[NULL];
|
||||||
|
SELECT ('[1,2,3]'::vector)[NULL:2];
|
||||||
|
SELECT ('[1,2,3]'::vector)[2:NULL];
|
||||||
|
SELECT ('[1,2,3]'::vector)[1][1];
|
||||||
|
|
||||||
SELECT '[1,2,3]'::vector = '[1,2,3]';
|
SELECT '[1,2,3]'::vector = '[1,2,3]';
|
||||||
SELECT '[1,2,3]'::vector = '[1,2]';
|
SELECT '[1,2,3]'::vector = '[1,2]';
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user