hash.c

PostgreSQL 8.1.4的源码 适用于Linux下的开源数据库系统

/*-------------------------------------------------------------------------
 *
 * hash.c
 *	  Implementation of Margo Seltzer's Hashing package for postgres.
 *
 * Portions Copyright (c) 1996-2005, PostgreSQL Global Development Group
 * Portions Copyright (c) 1994, Regents of the University of California
 *
 *
 * IDENTIFICATION
 *	  $PostgreSQL: pgsql/src/backend/access/hash/hash.c,v 1.81 2005/10/15 02:49:08 momjian Exp $
 *
 * NOTES
 *	  This file contains only the public interface routines.
 *
 *-------------------------------------------------------------------------
 */

#include "postgres.h"

#include "access/genam.h"
#include "access/hash.h"
#include "access/heapam.h"
#include "access/xlogutils.h"
#include "catalog/index.h"
#include "commands/vacuum.h"
#include "executor/executor.h"
#include "miscadmin.h"


/* Working state for hashbuild and its callback */
typedef struct
{
	double		indtuples;
} HashBuildState;

static void hashbuildCallback(Relation index,
				  HeapTuple htup,
				  Datum *values,
				  bool *isnull,
				  bool tupleIsAlive,
				  void *state);


/*
 *	hashbuild() -- build a new hash index.
 */
Datum
hashbuild(PG_FUNCTION_ARGS)
{
	Relation	heap = (Relation) PG_GETARG_POINTER(0);
	Relation	index = (Relation) PG_GETARG_POINTER(1);
	IndexInfo  *indexInfo = (IndexInfo *) PG_GETARG_POINTER(2);
	double		reltuples;
	HashBuildState buildstate;

	/*
	 * We expect to be called exactly once for any index relation. If that's
	 * not the case, big trouble's what we have.
	 */
	if (RelationGetNumberOfBlocks(index) != 0)
		elog(ERROR, "index \"%s\" already contains data",
			 RelationGetRelationName(index));

	/* initialize the hash index metadata page */
	_hash_metapinit(index);

	/* build the index */
	buildstate.indtuples = 0;

	/* do the heap scan */
	reltuples = IndexBuildHeapScan(heap, index, indexInfo,
								   hashbuildCallback, (void *) &buildstate);

	/* since we just counted the # of tuples, may as well update stats */
	IndexCloseAndUpdateStats(heap, reltuples, index, buildstate.indtuples);

	PG_RETURN_VOID();
}

/*
 * Per-tuple callback from IndexBuildHeapScan
 */
static void
hashbuildCallback(Relation index,
				  HeapTuple htup,
				  Datum *values,
				  bool *isnull,
				  bool tupleIsAlive,
				  void *state)
{
	HashBuildState *buildstate = (HashBuildState *) state;
	IndexTuple	itup;
	HashItem	hitem;

	/* form an index tuple and point it at the heap tuple */
	itup = index_form_tuple(RelationGetDescr(index), values, isnull);
	itup->t_tid = htup->t_self;

	/* Hash indexes don't index nulls, see notes in hashinsert */
	if (IndexTupleHasNulls(itup))
	{
		pfree(itup);
		return;
	}

	hitem = _hash_formitem(itup);

	_hash_doinsert(index, hitem);

	buildstate->indtuples += 1;

	pfree(hitem);
	pfree(itup);
}

/*
 *	hashinsert() -- insert an index tuple into a hash table.
 *
 *	Hash on the index tuple's key, find the appropriate location
 *	for the new tuple, and put it there.
 */
Datum
hashinsert(PG_FUNCTION_ARGS)
{
	Relation	rel = (Relation) PG_GETARG_POINTER(0);
	Datum	   *values = (Datum *) PG_GETARG_POINTER(1);
	bool	   *isnull = (bool *) PG_GETARG_POINTER(2);
	ItemPointer ht_ctid = (ItemPointer) PG_GETARG_POINTER(3);

#ifdef NOT_USED
	Relation	heapRel = (Relation) PG_GETARG_POINTER(4);
	bool		checkUnique = PG_GETARG_BOOL(5);
#endif
	HashItem	hitem;
	IndexTuple	itup;

	/* generate an index tuple */
	itup = index_form_tuple(RelationGetDescr(rel), values, isnull);
	itup->t_tid = *ht_ctid;

	/*
	 * If the single index key is null, we don't insert it into the index.
	 * Hash tables support scans on '='. Relational algebra says that A = B
	 * returns null if either A or B is null.  This means that no
	 * qualification used in an index scan could ever return true on a null
	 * attribute.  It also means that indices can't be used by ISNULL or
	 * NOTNULL scans, but that's an artifact of the strategy map architecture
	 * chosen in 1986, not of the way nulls are handled here.
	 */
	if (IndexTupleHasNulls(itup))
	{
		pfree(itup);
		PG_RETURN_BOOL(false);
	}

	hitem = _hash_formitem(itup);

	_hash_doinsert(rel, hitem);

	pfree(hitem);
	pfree(itup);

	PG_RETURN_BOOL(true);
}


/*
 *	hashgettuple() -- Get the next tuple in the scan.
 */
Datum
hashgettuple(PG_FUNCTION_ARGS)
{
	IndexScanDesc scan = (IndexScanDesc) PG_GETARG_POINTER(0);
	ScanDirection dir = (ScanDirection) PG_GETARG_INT32(1);
	HashScanOpaque so = (HashScanOpaque) scan->opaque;
	Relation	rel = scan->indexRelation;
	Page		page;
	OffsetNumber offnum;
	bool		res;

	/*
	 * We hold pin but not lock on current buffer while outside the hash AM.
	 * Reacquire the read lock here.
	 */
	if (BufferIsValid(so->hashso_curbuf))
		_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_NOLOCK, HASH_READ);

	/*
	 * If we've already initialized this scan, we can just advance it in the
	 * appropriate direction.  If we haven't done so yet, we call a routine to
	 * get the first item in the scan.
	 */
	if (ItemPointerIsValid(&(scan->currentItemData)))
	{
		/*
		 * Check to see if we should kill the previously-fetched tuple.
		 */
		if (scan->kill_prior_tuple)
		{
			/*
			 * Yes, so mark it by setting the LP_DELETE bit in the item flags.
			 */
			offnum = ItemPointerGetOffsetNumber(&(scan->currentItemData));
			page = BufferGetPage(so->hashso_curbuf);
			PageGetItemId(page, offnum)->lp_flags |= LP_DELETE;

			/*
			 * Since this can be redone later if needed, it's treated the same
			 * as a commit-hint-bit status update for heap tuples: we mark the
			 * buffer dirty but don't make a WAL log entry.
			 */
			SetBufferCommitInfoNeedsSave(so->hashso_curbuf);
		}

		/*
		 * Now continue the scan.
		 */
		res = _hash_next(scan, dir);
	}
	else
		res = _hash_first(scan, dir);

	/*
	 * Skip killed tuples if asked to.
	 */
	if (scan->ignore_killed_tuples)
	{
		while (res)
		{
			offnum = ItemPointerGetOffsetNumber(&(scan->currentItemData));
			page = BufferGetPage(so->hashso_curbuf);
			if (!ItemIdDeleted(PageGetItemId(page, offnum)))
				break;
			res = _hash_next(scan, dir);
		}
	}

	/* Release read lock on current buffer, but keep it pinned */
	if (BufferIsValid(so->hashso_curbuf))
		_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_READ, HASH_NOLOCK);

	PG_RETURN_BOOL(res);
}


/*
 *	hashgetmulti() -- get multiple tuples at once
 *
 * This is a somewhat generic implementation: it avoids lock reacquisition
 * overhead, but there's no smarts about picking especially good stopping
 * points such as index page boundaries.
 */
Datum
hashgetmulti(PG_FUNCTION_ARGS)
{
	IndexScanDesc scan = (IndexScanDesc) PG_GETARG_POINTER(0);
	ItemPointer tids = (ItemPointer) PG_GETARG_POINTER(1);
	int32		max_tids = PG_GETARG_INT32(2);
	int32	   *returned_tids = (int32 *) PG_GETARG_POINTER(3);
	HashScanOpaque so = (HashScanOpaque) scan->opaque;
	Relation	rel = scan->indexRelation;
	bool		res = true;
	int32		ntids = 0;

	/*
	 * We hold pin but not lock on current buffer while outside the hash AM.
	 * Reacquire the read lock here.
	 */
	if (BufferIsValid(so->hashso_curbuf))
		_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_NOLOCK, HASH_READ);

	while (ntids < max_tids)
	{
		/*
		 * Start scan, or advance to next tuple.
		 */
		if (ItemPointerIsValid(&(scan->currentItemData)))
			res = _hash_next(scan, ForwardScanDirection);
		else
			res = _hash_first(scan, ForwardScanDirection);

		/*
		 * Skip killed tuples if asked to.
		 */
		if (scan->ignore_killed_tuples)
		{
			while (res)
			{
				Page		page;
				OffsetNumber offnum;

				offnum = ItemPointerGetOffsetNumber(&(scan->currentItemData));
				page = BufferGetPage(so->hashso_curbuf);
				if (!ItemIdDeleted(PageGetItemId(page, offnum)))
					break;
				res = _hash_next(scan, ForwardScanDirection);
			}
		}

		if (!res)
			break;
		/* Save tuple ID, and continue scanning */
		tids[ntids] = scan->xs_ctup.t_self;
		ntids++;
	}

	/* Release read lock on current buffer, but keep it pinned */
	if (BufferIsValid(so->hashso_curbuf))
		_hash_chgbufaccess(rel, so->hashso_curbuf, HASH_READ, HASH_NOLOCK);

	*returned_tids = ntids;
	PG_RETURN_BOOL(res);
}


/*
 *	hashbeginscan() -- start a scan on a hash index
 */
Datum
hashbeginscan(PG_FUNCTION_ARGS)
{
	Relation	rel = (Relation) PG_GETARG_POINTER(0);
	int			keysz = PG_GETARG_INT32(1);
	ScanKey		scankey = (ScanKey) PG_GETARG_POINTER(2);
	IndexScanDesc scan;
	HashScanOpaque so;

	scan = RelationGetIndexScan(rel, keysz, scankey);
	so = (HashScanOpaque) palloc(sizeof(HashScanOpaqueData));
	so->hashso_bucket_valid = false;
	so->hashso_bucket_blkno = 0;
	so->hashso_curbuf = so->hashso_mrkbuf = InvalidBuffer;
	scan->opaque = so;

	/* register scan in case we change pages it's using */
	_hash_regscan(scan);

	PG_RETURN_POINTER(scan);