index.c

PostgreSQL 8.1.4的源码 适用于Linux下的开源数据库系统

	Assert(PointerIsValid(indexRelation->rd_am));

	procedure = indexRelation->rd_am->ambuild;
	Assert(RegProcedureIsValid(procedure));

	/*
	 * Call the access method's build procedure
	 */
	OidFunctionCall3(procedure,
					 PointerGetDatum(heapRelation),
					 PointerGetDatum(indexRelation),
					 PointerGetDatum(indexInfo));
}


/*
 * IndexBuildHeapScan - scan the heap relation to find tuples to be indexed
 *
 * This is called back from an access-method-specific index build procedure
 * after the AM has done whatever setup it needs.  The parent heap relation
 * is scanned to find tuples that should be entered into the index.  Each
 * such tuple is passed to the AM's callback routine, which does the right
 * things to add it to the new index.  After we return, the AM's index
 * build procedure does whatever cleanup is needed; in particular, it should
 * close the heap and index relations.
 *
 * The total count of heap tuples is returned.	This is for updating pg_class
 * statistics.	(It's annoying not to be able to do that here, but we can't
 * do it until after the relation is closed.)  Note that the index AM itself
 * must keep track of the number of index tuples; we don't do so here because
 * the AM might reject some of the tuples for its own reasons, such as being
 * unable to store NULLs.
 */
double
IndexBuildHeapScan(Relation heapRelation,
				   Relation indexRelation,
				   IndexInfo *indexInfo,
				   IndexBuildCallback callback,
				   void *callback_state)
{
	HeapScanDesc scan;
	HeapTuple	heapTuple;
	Datum		values[INDEX_MAX_KEYS];
	bool		isnull[INDEX_MAX_KEYS];
	double		reltuples;
	List	   *predicate;
	TupleTableSlot *slot;
	EState	   *estate;
	ExprContext *econtext;
	Snapshot	snapshot;
	TransactionId OldestXmin;

	/*
	 * sanity checks
	 */
	Assert(OidIsValid(indexRelation->rd_rel->relam));

	/*
	 * Need an EState for evaluation of index expressions and partial-index
	 * predicates.	Also a slot to hold the current tuple.
	 */
	estate = CreateExecutorState();
	econtext = GetPerTupleExprContext(estate);
	slot = MakeSingleTupleTableSlot(RelationGetDescr(heapRelation));

	/* Arrange for econtext's scan tuple to be the tuple under test */
	econtext->ecxt_scantuple = slot;

	/* Set up execution state for predicate, if any. */
	predicate = (List *)
		ExecPrepareExpr((Expr *) indexInfo->ii_Predicate,
						estate);

	/*
	 * Ok, begin our scan of the base relation.  We use SnapshotAny because we
	 * must retrieve all tuples and do our own time qual checks.
	 */
	if (IsBootstrapProcessingMode())
	{
		snapshot = SnapshotNow;
		OldestXmin = InvalidTransactionId;
	}
	else
	{
		snapshot = SnapshotAny;
		OldestXmin = GetOldestXmin(heapRelation->rd_rel->relisshared);
	}

	scan = heap_beginscan(heapRelation, /* relation */
						  snapshot,		/* seeself */
						  0,	/* number of keys */
						  NULL);	/* scan key */

	reltuples = 0;

	/*
	 * Scan all tuples in the base relation.
	 */
	while ((heapTuple = heap_getnext(scan, ForwardScanDirection)) != NULL)
	{
		bool		tupleIsAlive;

		CHECK_FOR_INTERRUPTS();

		if (snapshot == SnapshotAny)
		{
			/* do our own time qual check */
			bool		indexIt;

			/*
			 * We could possibly get away with not locking the buffer here,
			 * since caller should hold ShareLock on the relation, but let's
			 * be conservative about it.
			 */
			LockBuffer(scan->rs_cbuf, BUFFER_LOCK_SHARE);

			switch (HeapTupleSatisfiesVacuum(heapTuple->t_data, OldestXmin,
											 scan->rs_cbuf))
			{
				case HEAPTUPLE_DEAD:
					indexIt = false;
					tupleIsAlive = false;
					break;
				case HEAPTUPLE_LIVE:
					indexIt = true;
					tupleIsAlive = true;
					break;
				case HEAPTUPLE_RECENTLY_DEAD:

					/*
					 * If tuple is recently deleted then we must index it
					 * anyway to preserve MVCC semantics.  (Pre-existing
					 * transactions could try to use the index after we
					 * finish building it, and may need to see such tuples.)
					 */
					indexIt = true;
					tupleIsAlive = false;
					break;
				case HEAPTUPLE_INSERT_IN_PROGRESS:

					/*
					 * Since caller should hold ShareLock or better, we should
					 * not see any tuples inserted by open transactions ---
					 * unless it's our own transaction. (Consider INSERT
					 * followed by CREATE INDEX within a transaction.)	An
					 * exception occurs when reindexing a system catalog,
					 * because we often release lock on system catalogs before
					 * committing.
					 */
					if (!TransactionIdIsCurrentTransactionId(
								   HeapTupleHeaderGetXmin(heapTuple->t_data))
						&& !IsSystemRelation(heapRelation))
						elog(ERROR, "concurrent insert in progress");
					indexIt = true;
					tupleIsAlive = true;
					break;
				case HEAPTUPLE_DELETE_IN_PROGRESS:

					/*
					 * Since caller should hold ShareLock or better, we should
					 * not see any tuples deleted by open transactions ---
					 * unless it's our own transaction. (Consider DELETE
					 * followed by CREATE INDEX within a transaction.)	An
					 * exception occurs when reindexing a system catalog,
					 * because we often release lock on system catalogs before
					 * committing.
					 */
					Assert(!(heapTuple->t_data->t_infomask & HEAP_XMAX_IS_MULTI));
					if (!TransactionIdIsCurrentTransactionId(
								   HeapTupleHeaderGetXmax(heapTuple->t_data))
						&& !IsSystemRelation(heapRelation))
						elog(ERROR, "concurrent delete in progress");
					indexIt = true;
					tupleIsAlive = false;
					break;
				default:
					elog(ERROR, "unexpected HeapTupleSatisfiesVacuum result");
					indexIt = tupleIsAlive = false;		/* keep compiler quiet */
					break;
			}

			LockBuffer(scan->rs_cbuf, BUFFER_LOCK_UNLOCK);

			if (!indexIt)
				continue;
		}
		else
		{
			/* heap_getnext did the time qual check */
			tupleIsAlive = true;
		}

		reltuples += 1;

		MemoryContextReset(econtext->ecxt_per_tuple_memory);

		/* Set up for predicate or expression evaluation */
		ExecStoreTuple(heapTuple, slot, InvalidBuffer, false);

		/*
		 * In a partial index, discard tuples that don't satisfy the
		 * predicate.
		 */
		if (predicate != NIL)
		{
			if (!ExecQual(predicate, econtext, false))
				continue;
		}

		/*
		 * For the current heap tuple, extract all the attributes we use in
		 * this index, and note which are null.  This also performs evaluation
		 * of any expressions needed.
		 */
		FormIndexDatum(indexInfo,
					   slot,
					   estate,
					   values,
					   isnull);

		/*
		 * You'd think we should go ahead and build the index tuple here, but
		 * some index AMs want to do further processing on the data first.	So
		 * pass the values[] and isnull[] arrays, instead.
		 */

		/* Call the AM's callback routine to process the tuple */
		callback(indexRelation, heapTuple, values, isnull, tupleIsAlive,
				 callback_state);
	}

	heap_endscan(scan);

	ExecDropSingleTupleTableSlot(slot);

	FreeExecutorState(estate);

	/* These may have been pointing to the now-gone estate */
	indexInfo->ii_ExpressionsState = NIL;
	indexInfo->ii_PredicateState = NIL;

	return reltuples;
}


/*
 * IndexGetRelation: given an index's relation OID, get the OID of the
 * relation it is an index on.	Uses the system cache.
 */
static Oid
IndexGetRelation(Oid indexId)
{
	HeapTuple	tuple;
	Form_pg_index index;
	Oid			result;

	tuple = SearchSysCache(INDEXRELID,
						   ObjectIdGetDatum(indexId),
						   0, 0, 0);
	if (!HeapTupleIsValid(tuple))
		elog(ERROR, "cache lookup failed for index %u", indexId);
	index = (Form_pg_index) GETSTRUCT(tuple);
	Assert(index->indexrelid == indexId);

	result = index->indrelid;
	ReleaseSysCache(tuple);
	return result;
}

/*
 * reindex_index - This routine is used to recreate a single index
 */
void
reindex_index(Oid indexId)
{
	Relation	iRel,
				heapRelation;
	Oid			heapId;
	bool		inplace;

	/*
	 * Open and lock the parent heap relation.	ShareLock is sufficient since
	 * we only need to be sure no schema or data changes are going on.
	 */
	heapId = IndexGetRelation(indexId);
	heapRelation = heap_open(heapId, ShareLock);

	/*
	 * Open the target index relation and get an exclusive lock on it, to
	 * ensure that no one else is touching this particular index.
	 */
	iRel = index_open(indexId);
	LockRelation(iRel, AccessExclusiveLock);

	/*
	 * If it's a shared index, we must do inplace processing (because we have
	 * no way to update relfilenode in other databases).  Otherwise we can do
	 * it the normal transaction-safe way.
	 *
	 * Since inplace processing isn't crash-safe, we only allow it in a
	 * standalone backend.	(In the REINDEX TABLE and REINDEX DATABASE cases,
	 * the caller should have detected this.)
	 */
	inplace = iRel->rd_rel->relisshared;

	if (inplace && IsUnderPostmaster)
		ereport(ERROR,
				(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
				 errmsg("shared index \"%s\" can only be reindexed in stand-alone mode",
						RelationGetRelationName(iRel))));

	PG_TRY();
	{
		IndexInfo  *indexInfo;

		/* Suppress use of the target index while rebuilding it */
		SetReindexProcessing(heapId, indexId);

		/* Fetch info needed for index_build */
		indexInfo = BuildIndexInfo(iRel);

		if (inplace)
		{
			/* Truncate the actual file (and discard buffers) */
			RelationTruncate(iRel, 0);
		}
		else
		{
			/*
			 * We'll build a new physical relation for the index.
			 */
			setNewRelfilenode(iRel);
		}

		/* Initialize the index and rebuild */
		index_build(heapRelation, iRel, indexInfo);

		/*
		 * index_build will close both the heap and index relations (but not
		 * give up the locks we hold on them).	So we're done.
		 */
	}
	PG_CATCH();
	{
		/* Make sure flag gets cleared on error exit */
		ResetReindexProcessing();
		PG_RE_THROW();
	}
	PG_END_TRY();
	ResetReindexProcessing();
}

/*
 * reindex_relation - This routine is used to recreate all indexes
 * of a relation (and optionally its toast relation too, if any).
 *
 * Returns true if any indexes were rebuilt.  Note that a
 * CommandCounterIncrement will occur after each index rebuild.
 */
bool
reindex_relation(Oid relid, bool toast_too)
{
	Relation	rel;
	Oid			toast_relid;
	bool		is_pg_class;
	bool		result;
	List	   *indexIds,
			   *doneIndexes;
	ListCell   *indexId;

	/*
	 * Open and lock the relation.	ShareLock is sufficient since we only need
	 * to prevent schema and data changes in it.
	 */
	rel = heap_open(relid, ShareLock);

	toast_relid = rel->rd_rel->reltoastrelid;

	/*
	 * Get the list of index OIDs for this relation.  (We trust to the
	 * relcache to get this with a sequential scan if ignoring system
	 * indexes.)
	 */
	indexIds = RelationGetIndexList(rel);

	/*
	 * reindex_index will attempt to update the pg_class rows for the relation
	 * and index.  If we are processing pg_class itself, we want to make sure
	 * that the updates do not try to insert index entries into indexes we
	 * have not processed yet.	(When we are trying to recover from corrupted
	 * indexes, that could easily cause a crash.) We can accomplish this
	 * because CatalogUpdateIndexes will use the relcache's index list to know
	 * which indexes to update. We just force the index list to be only the
	 * stuff we've processed.
	 *
	 * It is okay to not insert entries into the indexes we have not processed
	 * yet because all of this is transaction-safe.  If we fail partway
	 * through, the updated rows are dead and it doesn't matter whether they
	 * have index entries.	Also, a new pg_class index will be created with an
	 * entry for its own pg_class row because we do setNewRelfilenode() before
	 * we do index_build().
	 *
	 * Note that we also clear pg_class's rd_oidindex until the loop is done,
	 * so that that index can't be accessed either.  This means we cannot
	 * safely generate new relation OIDs while in the loop; shouldn't be a
	 * problem.
	 */
	is_pg_class = (RelationGetRelid(rel) == RelationRelationId);
	doneIndexes = NIL;

	/* Reindex all the indexes. */
	foreach(indexId, indexIds)
	{
		Oid			indexOid = lfirst_oid(indexId);

		if (is_pg_class)
			RelationSetIndexList(rel, doneIndexes, InvalidOid);

		reindex_index(indexOid);

		CommandCounterIncrement();

		if (is_pg_class)
			doneIndexes = lappend_oid(doneIndexes, indexOid);
	}

	if (is_pg_class)
		RelationSetIndexList(rel, indexIds, ClassOidIndexId);

	/*
	 * Close rel, but continue to hold the lock.
	 */
	heap_close(rel, NoLock);

	result = (indexIds != NIL);

	/*
	 * If the relation has a secondary toast rel, reindex that too while we
	 * still hold the lock on the master table.
	 */
	if (toast_too && OidIsValid(toast_relid))
		result |= reindex_relation(toast_relid, false);

	return result;
}