Обсуждение: OR clause - check code
Vadim, would you please review this code, and let me know if it is
correct. I am unsure about the calls to ExecStoreTuple(), ExecQual()
(is proper context used?), and placement of ReleaseBuffer(). Do I need
to free the tuple buffer if I decide it doesn't meet my ExecQual test?
You will see the old code, then the new for IndexNext().
---------------------------------------------------------------------------
***************
*** 120,155 ****
* appropriate heap tuple.. else return NULL.
* ----------------
*/
! while ((result = index_getnext(scandesc, direction)) != NULL)
{
! tuple = heap_fetch(heapRelation, false, &result->heap_iptr, &buffer);
! /* be tidy */
! pfree(result);
!
! if (tuple != NULL)
{
! /* ----------------
! * store the scanned tuple in the scan tuple slot of
! * the scan state. Eventually we will only do this and not
! * return a tuple. Note: we pass 'false' because tuples
! * returned by amgetnext are pointers onto disk pages and
! * were not created with palloc() and so should not be pfree()'d.
! * ----------------
! */
! ExecStoreTuple(tuple, /* tuple to store */
! slot, /* slot to store in */
! buffer, /* buffer associated with tuple */
! false); /* don't pfree */
!
! return slot;
! }
! else
! {
if (BufferIsValid(buffer))
ReleaseBuffer(buffer);
}
}
-
/* ----------------
* if we get here it means the index scan failed so we
* are at the end of the scan..
--- 119,170 ----
* appropriate heap tuple.. else return NULL.
* ----------------
*/
! while (indexstate->iss_IndexPtr < numIndices)
{
! scandesc = scanDescs[indexstate->iss_IndexPtr];
! while ((result = index_getnext(scandesc, direction)) != NULL)
{
! tuple = heap_fetch(heapRelation, false, &result->heap_iptr, &buffer);
! /* be tidy */
! pfree(result);
!
! if (tuple != NULL)
! {
! bool prev_matches = false;
! int prev_index;
!
! /* ----------------
! * store the scanned tuple in the scan tuple slot of
! * the scan state. Eventually we will only do this and not
! * return a tuple. Note: we pass 'false' because tuples
! * returned by amgetnext are pointers onto disk pages and
! * were not created with palloc() and so should not be pfree()'d.
! * ----------------
! */
! ExecStoreTuple(tuple, /* tuple to store */
! slot, /* slot to store in */
! buffer, /* buffer associated with tuple */
! false); /* don't pfree */
!
! for (prev_index = 0; prev_index < indexstate->iss_IndexPtr;
! prev_index++)
! {
! if (ExecQual(nth(prev_index, node->indxqual),
! scanstate->cstate.cs_ExprContext))
! {
! prev_matches = true;
! break;
! }
! }
! if (!prev_matches)
! return slot;
! }
if (BufferIsValid(buffer))
ReleaseBuffer(buffer);
}
+ if (indexstate->iss_IndexPtr < numIndices)
+ indexstate->iss_IndexPtr++;
}
/* ----------------
* if we get here it means the index scan failed so we
* are at the end of the scan..
--
Bruce Momjian | 830 Blythe Avenue
maillist@candle.pha.pa.us | Drexel Hill, Pennsylvania 19026
+ If your life is a hard drive, | (610) 353-9879(w)
+ Christ can be your backup. | (610) 853-3000(h)
Bruce Momjian wrote:
>
> Vadim, would you please review this code, and let me know if it is
> correct. I am unsure about the calls to ExecStoreTuple(), ExecQual()
> (is proper context used?), and placement of ReleaseBuffer(). Do I need
^^^^^^^^^^^^^^^^^^^^^^
Yes.
execScan.c:ExecScan() sets econtext->ecxt_scantuple = slot <--
returned by Seq/Index scan -- before check Scan qual and
so econtext->ecxt_scantuple already points to the slot where
you store new tuple. It seems that no one change this pointer...
But did you try to use multi-index scan in inner plan of
Nestloop? It seems that you should add this to ExecIndexReScan():
if (exprCtxt != NULL)
node->scan.scanstate->cstate.cs_ExprContext->ecxt_outertuple =
exprCtxt->ecxt_outertuple;
Also, ExecIndexReScan() evaluates run-time keys for _current_
index scan only - this must be changed and indexstate->iss_IndexPtr
must be setted to 0.
> to free the tuple buffer if I decide it doesn't meet my ExecQual test?
Yes, you have to free buffer.
Vadim
> Bruce Momjian wrote: > > > > Vadim, would you please review this code, and let me know if it is > > correct. I am unsure about the calls to ExecStoreTuple(), ExecQual() > > (is proper context used?), and placement of ReleaseBuffer(). Do I need > ^^^^^^^^^^^^^^^^^^^^^^ > Yes. > execScan.c:ExecScan() sets econtext->ecxt_scantuple = slot <-- > returned by Seq/Index scan -- before check Scan qual and > so econtext->ecxt_scantuple already points to the slot where > you store new tuple. It seems that no one change this pointer... > > But did you try to use multi-index scan in inner plan of > Nestloop? It seems that you should add this to ExecIndexReScan(): No. I just don't understand the behavior of the various join types on index use. > > if (exprCtxt != NULL) > node->scan.scanstate->cstate.cs_ExprContext->ecxt_outertuple = > exprCtxt->ecxt_outertuple; Added. > > Also, ExecIndexReScan() evaluates run-time keys for _current_ > index scan only - this must be changed and indexstate->iss_IndexPtr > must be setted to 0. I have added this to the code. > > to free the tuple buffer if I decide it doesn't meet my ExecQual test? > > Yes, you have to free buffer. Done. Vadim, I am going to send you the current nodeIndexscan.c for your review, to make sure I made the proper changes. -- Bruce Momjian | 830 Blythe Avenue maillist@candle.pha.pa.us | Drexel Hill, Pennsylvania 19026 + If your life is a hard drive, | (610) 353-9879(w) + Christ can be your backup. | (610) 853-3000(h)