/*-------------------------------------------------------
* bt_page_print_tuples()
*
* Form a tuple describing index tuple at a given offset
* ------------------------------------------------------
*/
static Datum
bt_page_print_tuples(FuncCallContext *fctx, Page page, OffsetNumber offset)
{
char *values[6];
HeapTuple tuple;
ItemId id;
IndexTuple itup;
int j;
int off;
int dlen;
char *dump;
char *ptr;
id = PageGetItemId(page, offset);
if (!ItemIdIsValid(id))
elog(ERROR, "invalid ItemId");
itup = (IndexTuple) PageGetItem(page, id);
j = 0;
values[j++] = psprintf("%d", offset);
values[j++] = psprintf("(%u,%u)",
ItemPointerGetBlockNumberNoCheck(&itup->t_tid),
ItemPointerGetOffsetNumberNoCheck(&itup->t_tid));
values[j++] = psprintf("%d", (int) IndexTupleSize(itup));
values[j++] = psprintf("%c", IndexTupleHasNulls(itup) ? 't' : 'f');
values[j++] = psprintf("%c", IndexTupleHasVarwidths(itup) ? 't' : 'f');
ptr = (char *) itup + IndexInfoFindDataOffset(itup->t_info);
dlen = IndexTupleSize(itup) - IndexInfoFindDataOffset(itup->t_info);
dump = palloc0(dlen * 3 + 1);
values[j] = dump;
for (off = 0; off < dlen; off++)
{
if (off > 0)
*dump++ = ' ';
sprintf(dump, "%02x", *(ptr + off) & 0xff);
dump += 2;
}
tuple = BuildTupleFromCStrings(fctx->attinmeta, values);
return HeapTupleGetDatum(tuple);
}
/* ----------------
* nocache_index_getattr
*
* This gets called from index_getattr() macro, and only in cases
* where we can't use cacheoffset and the value is not null.
*
* This caches attribute offsets in the attribute descriptor.
*
* An alternative way to speed things up would be to cache offsets
* with the tuple, but that seems more difficult unless you take
* the storage hit of actually putting those offsets into the
* tuple you send to disk. Yuck.
*
* This scheme will be slightly slower than that, but should
* perform well for queries which hit large #'s of tuples. After
* you cache the offsets once, examining all the other tuples using
* the same attribute descriptor will go much quicker. -cim 5/4/91
* ----------------
*/
Datum
nocache_index_getattr(IndexTuple tup,
int attnum,
TupleDesc tupleDesc)
{
Form_pg_attribute *att = tupleDesc->attrs;
char *tp; /* ptr to data part of tuple */
bits8 *bp = NULL; /* ptr to null bitmap in tuple */
bool slow = false; /* do we have to walk attrs? */
int data_off; /* tuple data offset */
int off; /* current offset within data */
/* ----------------
* Three cases:
*
* 1: No nulls and no variable-width attributes.
* 2: Has a null or a var-width AFTER att.
* 3: Has nulls or var-widths BEFORE att.
* ----------------
*/
data_off = IndexInfoFindDataOffset(tup->t_info);
attnum--;
if (IndexTupleHasNulls(tup))
{
/*
* there's a null somewhere in the tuple
*
* check to see if desired att is null
*/
/* XXX "knows" t_bits are just after fixed tuple header! */
bp = (bits8 *) ((char *) tup + sizeof(IndexTupleData));
/*
* Now check to see if any preceding bits are null...
*/
{
int byte = attnum >> 3;
int finalbit = attnum & 0x07;
/* check for nulls "before" final bit of last byte */
if ((~bp[byte]) & ((1 << finalbit) - 1))
slow = true;
else
{
/* check for nulls in any "earlier" bytes */
int i;
for (i = 0; i < byte; i++)
{
if (bp[i] != 0xFF)
{
slow = true;
break;
}
}
}
}
}
tp = (char *) tup + data_off;
if (!slow)
{
/*
* If we get here, there are no nulls up to and including the target
* attribute. If we have a cached offset, we can use it.
*/
if (att[attnum]->attcacheoff >= 0)
{
return fetchatt(att[attnum],
tp + att[attnum]->attcacheoff);
}
/*
* Otherwise, check for non-fixed-length attrs up to and including
* target. If there aren't any, it's safe to cheaply initialize the
* cached offsets for these attrs.
*/
if (IndexTupleHasVarwidths(tup))
{
int j;
for (j = 0; j <= attnum; j++)
{
if (att[j]->attlen <= 0)
{
slow = true;
break;
}
}
}
}
if (!slow)
{
int natts = tupleDesc->natts;
int j = 1;
/*
* If we get here, we have a tuple with no nulls or var-widths up to
* and including the target attribute, so we can use the cached offset
* ... only we don't have it yet, or we'd not have got here. Since
* it's cheap to compute offsets for fixed-width columns, we take the
* opportunity to initialize the cached offsets for *all* the leading
* fixed-width columns, in hope of avoiding future visits to this
* routine.
*/
att[0]->attcacheoff = 0;
/* we might have set some offsets in the slow path previously */
while (j < natts && att[j]->attcacheoff > 0)
j++;
off = att[j - 1]->attcacheoff + att[j - 1]->attlen;
for (; j < natts; j++)
{
if (att[j]->attlen <= 0)
break;
off = att_align_nominal(off, att[j]->attalign);
att[j]->attcacheoff = off;
off += att[j]->attlen;
}
Assert(j > attnum);
off = att[attnum]->attcacheoff;
}
else
{
bool usecache = true;
int i;
/*
* Now we know that we have to walk the tuple CAREFULLY. But we still
* might be able to cache some offsets for next time.
*
* Note - This loop is a little tricky. For each non-null attribute,
* we have to first account for alignment padding before the attr,
* then advance over the attr based on its length. Nulls have no
* storage and no alignment padding either. We can use/set
* attcacheoff until we reach either a null or a var-width attribute.
*/
off = 0;
for (i = 0;; i++) /* loop exit is at "break" */
{
if (IndexTupleHasNulls(tup) && att_isnull(i, bp))
{
usecache = false;
continue; /* this cannot be the target att */
}
/* If we know the next offset, we can skip the rest */
if (usecache && att[i]->attcacheoff >= 0)
off = att[i]->attcacheoff;
else if (att[i]->attlen == -1)
{
/*
* We can only cache the offset for a varlena attribute if the
* offset is already suitably aligned, so that there would be
* no pad bytes in any case: then the offset will be valid for
* either an aligned or unaligned value.
*/
if (usecache &&
off == att_align_nominal(off, att[i]->attalign))
att[i]->attcacheoff = off;
else
{
off = att_align_pointer(off, att[i]->attalign, -1,
tp + off);
usecache = false;
}
}
else
{
/* not varlena, so safe to use att_align_nominal */
off = att_align_nominal(off, att[i]->attalign);
if (usecache)
att[i]->attcacheoff = off;
}
if (i == attnum)
break;
off = att_addlength_pointer(off, att[i]->attlen, tp + off);
if (usecache && att[i]->attlen <= 0)
usecache = false;
}
}
return fetchatt(att[attnum], tp + off);
}
Datum
bt_page_items(PG_FUNCTION_ARGS)
{
text *relname = PG_GETARG_TEXT_P(0);
uint32 blkno = PG_GETARG_UINT32(1);
RangeVar *relrv;
Datum result;
char *values[BTPAGEITEMS_NCOLUMNS];
BTPageOpaque opaque;
HeapTuple tuple;
ItemId id;
FuncCallContext *fctx;
MemoryContext mctx;
struct user_args *uargs = NULL;
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
(errmsg("must be superuser to use pgstattuple functions"))));
if (blkno == 0)
elog(ERROR, "Block 0 is a meta page.");
if (SRF_IS_FIRSTCALL())
{
fctx = SRF_FIRSTCALL_INIT();
mctx = MemoryContextSwitchTo(fctx->multi_call_memory_ctx);
uargs = palloc(sizeof(struct user_args));
uargs->tupd = RelationNameGetTupleDesc(BTPAGEITEMS_TYPE);
uargs->offset = FirstOffsetNumber;
relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
uargs->rel = relation_openrv(relrv, AccessShareLock);
CHECK_RELATION_BLOCK_RANGE(uargs->rel, blkno);
uargs->buffer = ReadBuffer(uargs->rel, blkno);
if (!IS_INDEX(uargs->rel) || !IS_BTREE(uargs->rel))
elog(ERROR, "bt_page_items() can be used only on b-tree index.");
uargs->page = BufferGetPage(uargs->buffer);
opaque = (BTPageOpaque) PageGetSpecialPointer(uargs->page);
if (P_ISDELETED(opaque))
elog(NOTICE, "bt_page_items(): this page is deleted.");
fctx->max_calls = PageGetMaxOffsetNumber(uargs->page);
fctx->user_fctx = uargs;
MemoryContextSwitchTo(mctx);
}
fctx = SRF_PERCALL_SETUP();
uargs = fctx->user_fctx;
if (fctx->call_cntr < fctx->max_calls)
{
IndexTuple itup;
id = PageGetItemId(uargs->page, uargs->offset);
if (!ItemIdIsValid(id))
elog(ERROR, "Invalid ItemId.");
itup = (IndexTuple) PageGetItem(uargs->page, id);
{
int j = 0;
BlockNumber blkno = BlockIdGetBlockNumber(&(itup->t_tid.ip_blkid));
values[j] = palloc(32);
snprintf(values[j++], 32, "%d", uargs->offset);
values[j] = palloc(32);
snprintf(values[j++], 32, "(%u,%u)", blkno, itup->t_tid.ip_posid);
values[j] = palloc(32);
snprintf(values[j++], 32, "%d", (int) IndexTupleSize(itup));
values[j] = palloc(32);
snprintf(values[j++], 32, "%c", IndexTupleHasNulls(itup) ? 't' : 'f');
values[j] = palloc(32);
snprintf(values[j++], 32, "%c", IndexTupleHasVarwidths(itup) ? 't' : 'f');
{
int off;
char *dump;
char *ptr = (char *) itup + IndexInfoFindDataOffset(itup->t_info);
dump = palloc(IndexTupleSize(itup) * 3);
memset(dump, 0, IndexTupleSize(itup) * 3);
for (off = 0;
off < IndexTupleSize(itup) - IndexInfoFindDataOffset(itup->t_info);
off++)
{
if (dump[0] == '\0')
sprintf(dump, "%02x", *(ptr + off) & 0xff);
else
{
char buf[4];
sprintf(buf, " %02x", *(ptr + off) & 0xff);
strcat(dump, buf);
}
}
values[j] = dump;
}
tuple = BuildTupleFromCStrings(TupleDescGetAttInMetadata(uargs->tupd), values);
result = TupleGetDatum(TupleDescGetSlot(uargs->tupd), tuple);
}
uargs->offset = uargs->offset + 1;
SRF_RETURN_NEXT(fctx, result);
}
else
{
ReleaseBuffer(uargs->buffer);
relation_close(uargs->rel, AccessShareLock);
SRF_RETURN_DONE(fctx);
}
}
Datum
bt_page_items(PG_FUNCTION_ARGS)
{
text *relname = PG_GETARG_TEXT_P(0);
uint32 blkno = PG_GETARG_UINT32(1);
Datum result;
char *values[6];
HeapTuple tuple;
FuncCallContext *fctx;
MemoryContext mctx;
struct user_args *uargs;
if (!superuser())
ereport(ERROR,
(errcode(ERRCODE_INSUFFICIENT_PRIVILEGE),
(errmsg("must be superuser to use pageinspect functions"))));
if (SRF_IS_FIRSTCALL())
{
RangeVar *relrv;
Relation rel;
Buffer buffer;
BTPageOpaque opaque;
TupleDesc tupleDesc;
fctx = SRF_FIRSTCALL_INIT();
relrv = makeRangeVarFromNameList(textToQualifiedNameList(relname));
rel = relation_openrv(relrv, AccessShareLock);
if (!IS_INDEX(rel) || !IS_BTREE(rel))
elog(ERROR, "relation \"%s\" is not a btree index",
RelationGetRelationName(rel));
/*
* Reject attempts to read non-local temporary relations; we would be
* likely to get wrong data since we have no visibility into the
* owning session's local buffers.
*/
if (RELATION_IS_OTHER_TEMP(rel))
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot access temporary tables of other sessions")));
if (blkno == 0)
elog(ERROR, "block 0 is a meta page");
CHECK_RELATION_BLOCK_RANGE(rel, blkno);
buffer = ReadBuffer(rel, blkno);
LockBuffer(buffer, BUFFER_LOCK_SHARE);
/*
* We copy the page into local storage to avoid holding pin on the
* buffer longer than we must, and possibly failing to release it at
* all if the calling query doesn't fetch all rows.
*/
mctx = MemoryContextSwitchTo(fctx->multi_call_memory_ctx);
uargs = palloc(sizeof(struct user_args));
uargs->page = palloc(BLCKSZ);
memcpy(uargs->page, BufferGetPage(buffer), BLCKSZ);
UnlockReleaseBuffer(buffer);
relation_close(rel, AccessShareLock);
uargs->offset = FirstOffsetNumber;
opaque = (BTPageOpaque) PageGetSpecialPointer(uargs->page);
if (P_ISDELETED(opaque))
elog(NOTICE, "page is deleted");
fctx->max_calls = PageGetMaxOffsetNumber(uargs->page);
/* Build a tuple descriptor for our result type */
if (get_call_result_type(fcinfo, NULL, &tupleDesc) != TYPEFUNC_COMPOSITE)
elog(ERROR, "return type must be a row type");
fctx->attinmeta = TupleDescGetAttInMetadata(tupleDesc);
fctx->user_fctx = uargs;
MemoryContextSwitchTo(mctx);
}
fctx = SRF_PERCALL_SETUP();
uargs = fctx->user_fctx;
if (fctx->call_cntr < fctx->max_calls)
{
ItemId id;
IndexTuple itup;
int j;
int off;
int dlen;
char *dump;
char *ptr;
id = PageGetItemId(uargs->page, uargs->offset);
if (!ItemIdIsValid(id))
elog(ERROR, "invalid ItemId");
itup = (IndexTuple) PageGetItem(uargs->page, id);
j = 0;
values[j++] = psprintf("%d", uargs->offset);
values[j++] = psprintf("(%u,%u)",
BlockIdGetBlockNumber(&(itup->t_tid.ip_blkid)),
itup->t_tid.ip_posid);
values[j++] = psprintf("%d", (int) IndexTupleSize(itup));
values[j++] = psprintf("%c", IndexTupleHasNulls(itup) ? 't' : 'f');
values[j++] = psprintf("%c", IndexTupleHasVarwidths(itup) ? 't' : 'f');
ptr = (char *) itup + IndexInfoFindDataOffset(itup->t_info);
dlen = IndexTupleSize(itup) - IndexInfoFindDataOffset(itup->t_info);
dump = palloc0(dlen * 3 + 1);
values[j] = dump;
for (off = 0; off < dlen; off++)
{
if (off > 0)
*dump++ = ' ';
sprintf(dump, "%02x", *(ptr + off) & 0xff);
dump += 2;
}
tuple = BuildTupleFromCStrings(fctx->attinmeta, values);
result = HeapTupleGetDatum(tuple);
uargs->offset = uargs->offset + 1;
SRF_RETURN_NEXT(fctx, result);
}
else
{
pfree(uargs->page);
pfree(uargs);
SRF_RETURN_DONE(fctx);
}
}
/* ----------------
* nocache_index_getattr
*
* This gets called from index_getattr() macro, and only in cases
* where we can't use cacheoffset and the value is not null.
*
* This caches attribute offsets in the attribute descriptor.
*
* An alternate way to speed things up would be to cache offsets
* with the tuple, but that seems more difficult unless you take
* the storage hit of actually putting those offsets into the
* tuple you send to disk. Yuck.
*
* This scheme will be slightly slower than that, but should
* perform well for queries which hit large #'s of tuples. After
* you cache the offsets once, examining all the other tuples using
* the same attribute descriptor will go much quicker. -cim 5/4/91
* ----------------
*/
Datum
nocache_index_getattr(IndexTuple tup,
int attnum,
TupleDesc tupleDesc,
bool *isnull)
{
Form_pg_attribute *att = tupleDesc->attrs;
char *tp; /* ptr to att in tuple */
bits8 *bp = NULL; /* ptr to null bitmask in tuple */
bool slow = false; /* do we have to walk nulls? */
int data_off; /* tuple data offset */
(void) isnull; /* not used */
/*
* sanity checks
*/
/* ----------------
* Three cases:
*
* 1: No nulls and no variable-width attributes.
* 2: Has a null or a var-width AFTER att.
* 3: Has nulls or var-widths BEFORE att.
* ----------------
*/
#ifdef IN_MACRO
/* This is handled in the macro */
Assert(PointerIsValid(isnull));
Assert(attnum > 0);
*isnull = false;
#endif
data_off = IndexInfoFindDataOffset(tup->t_info);
attnum--;
if (!IndexTupleHasNulls(tup))
{
#ifdef IN_MACRO
/* This is handled in the macro */
if (att[attnum]->attcacheoff != -1)
{
return fetchatt(att[attnum],
(char *) tup + data_off +
att[attnum]->attcacheoff);
}
#endif
}
else
{
/*
* there's a null somewhere in the tuple
*
* check to see if desired att is null
*/
/* XXX "knows" t_bits are just after fixed tuple header! */
bp = (bits8 *) ((char *) tup + sizeof(IndexTupleData));
#ifdef IN_MACRO
/* This is handled in the macro */
if (att_isnull(attnum, bp))
{
*isnull = true;
return (Datum) NULL;
}
#endif
/*
* Now check to see if any preceding bits are null...
*/
{
int byte = attnum >> 3;
int finalbit = attnum & 0x07;
/* check for nulls "before" final bit of last byte */
if ((~bp[byte]) & ((1 << finalbit) - 1))
slow = true;
else
{
/* check for nulls in any "earlier" bytes */
int i;
for (i = 0; i < byte; i++)
{
if (bp[i] != 0xFF)
{
slow = true;
break;
}
}
}
}
}
tp = (char *) tup + data_off;
/*
* now check for any non-fixed length attrs before our attribute
*/
if (!slow)
{
if (att[attnum]->attcacheoff != -1)
{
return fetchatt(att[attnum],
tp + att[attnum]->attcacheoff);
}
else if (IndexTupleHasVarwidths(tup))
{
int j;
for (j = 0; j < attnum; j++)
{
if (att[j]->attlen <= 0)
{
slow = true;
break;
}
}
}
}
/*
* If slow is false, and we got here, we know that we have a tuple with no
* nulls or var-widths before the target attribute. If possible, we also
* want to initialize the remainder of the attribute cached offset values.
*/
if (!slow)
{
int j = 1;
long off;
/*
* need to set cache for some atts
*/
att[0]->attcacheoff = 0;
while (j < attnum && att[j]->attcacheoff > 0)
j++;
off = att[j - 1]->attcacheoff + att[j - 1]->attlen;
for (; j <= attnum; j++)
{
off = att_align(off, att[j]->attalign);
att[j]->attcacheoff = off;
off += att[j]->attlen;
}
return fetchatt(att[attnum], tp + att[attnum]->attcacheoff);
}
else
{
bool usecache = true;
int off = 0;
int i;
/*
* Now we know that we have to walk the tuple CAREFULLY.
*/
for (i = 0; i < attnum; i++)
{
if (IndexTupleHasNulls(tup))
{
if (att_isnull(i, bp))
{
usecache = false;
continue;
}
}
/* If we know the next offset, we can skip the rest */
if (usecache && att[i]->attcacheoff != -1)
off = att[i]->attcacheoff;
else
{
off = att_align(off, att[i]->attalign);
if (usecache)
att[i]->attcacheoff = off;
}
off = att_addlength(off, att[i]->attlen, tp + off);
if (usecache && att[i]->attlen <= 0)
usecache = false;
}
off = att_align(off, att[attnum]->attalign);
return fetchatt(att[attnum], tp + off);
}
}
