liuyu/pgvector
1
0
Fork 0
mirror of https://github.com/pgvector/pgvector.git synced 2026-07-14 16:46:54 +08:00
Code Issues Packages Projects Releases Wiki Activity

Improved support functions for HNSW - #527

Browse Source
This commit is contained in:
Andrew Kane
2024-04-25 13:00:40 -07:00
parent 2bf1175ab0
commit 47d5b2896e
7 changed files with 104 additions and 107 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
src/hnswbuild.c
View File

@@ -488,8 +488,8 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
Datum value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
/* Check value */
if (buildstate->checkvalueprocinfo != NULL)
HnswCheckValue(buildstate->checkvalueprocinfo, buildstate->collation, value);
if (buildstate->typeInfo->checkValue != NULL)
buildstate->typeInfo->checkValue(DatumGetPointer(value));
/* Normalize if needed */
if (buildstate->normprocinfo != NULL)
@@ -672,32 +672,17 @@ HnswSharedMemoryAlloc(Size size, void *state)
return chunk;
}
/*
* Get max dimensions
*/
static int
GetMaxDimensions(Relation index)
{
FmgrInfo *procinfo = HnswOptionalProcInfo(index, HNSW_MAX_DIMS_PROC);
if (procinfo == NULL)
return HNSW_MAX_DIM;
return DatumGetInt32(FunctionCall1(procinfo, PointerGetDatum(NULL)));
}
/*
* Initialize the build state
*/
static void
InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, IndexInfo *indexInfo, ForkNumber forkNum)
{
int maxDimensions;
buildstate->heap = heap;
buildstate->index = index;
buildstate->indexInfo = indexInfo;
buildstate->forkNum = forkNum;
buildstate->typeInfo = HnswGetTypeInfo(index);
buildstate->m = HnswGetM(index);
buildstate->efConstruction = HnswGetEfConstruction(index);
@@ -707,14 +692,12 @@ InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, Index
if (TupleDescAttr(index->rd_att, 0)->atttypid == VARBITOID)
elog(ERROR, "type not supported for hnsw index");
maxDimensions = GetMaxDimensions(index);
/* Require column to have dimensions to be indexed */
if (buildstate->dimensions < 0)
elog(ERROR, "column does not have dimensions");
if (buildstate->dimensions > maxDimensions)
elog(ERROR, "column cannot have more than %d dimensions for hnsw index", maxDimensions);
if (buildstate->dimensions > buildstate->typeInfo->maxDimensions)
elog(ERROR, "column cannot have more than %d dimensions for hnsw index", buildstate->typeInfo->maxDimensions);
if (buildstate->efConstruction < 2 * buildstate->m)
elog(ERROR, "ef_construction must be greater than or equal to 2 * m");
@@ -726,7 +709,6 @@ InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, Index
buildstate->procinfo = index_getprocinfo(index, 1, HNSW_DISTANCE_PROC);
buildstate->normprocinfo = HnswOptionalProcInfo(index, HNSW_NORM_PROC);
buildstate->normalizeprocinfo = HnswOptionalProcInfo(index, HNSW_NORMALIZE_PROC);
buildstate->checkvalueprocinfo = HnswOptionalProcInfo(index, HNSW_CHECK_VALUE_PROC);
buildstate->collation = index->rd_indcollation[0];
InitGraph(&buildstate->graphData, NULL, maintenance_work_mem * 1024L);