database/postgres
1
0
Fork 0
mirror of https://github.com/postgres/postgres.git synced 2025-07-15 19:21:59 +03:00
Code Activity

Clean up messy clause-selectivity code in clausesel.c; repair bug

Browse Source 
identified by Hiroshi (incorrect cost attributed to OR clauses
after multiple passes through set_rest_selec()).  I think the code
was trying to allow selectivities of OR subclauses to be passed in
from outside, but noplace was actually passing any useful data, and
set_rest_selec() was passing wrong data.

Restructure representation of "indexqual" in IndexPath nodes so that
it is the same as for indxqual in completed IndexScan nodes: namely,
a toplevel list with an entry for each pass of the index scan, having
sublists that are implicitly-ANDed index qual conditions for that pass.
You don't want to know what the old representation was :-(

Improve documentation of OR-clause indexscan functions.

Remove useless 'notclause' field from RestrictInfo nodes.  (This might
force an initdb for anyone who has stored rules containing RestrictInfos,
but I do not think that RestrictInfo ever appears in completed plans.)
This commit is contained in:
Tom Lane
1999-07-24 23:21:14 +00:00
parent 348bdbce79
commit ac4913a0dd
17 changed files with 471 additions and 519 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
src/backend/optimizer/path/allpaths.c
View File

@ -7,7 +7,7 @@
*
*
* IDENTIFICATION
* $Header: /cvsroot/pgsql/src/backend/optimizer/path/allpaths.c,v 1.50 1999/07/17 20:17:11 momjian Exp $
* $Header: /cvsroot/pgsql/src/backend/optimizer/path/allpaths.c,v 1.51 1999/07/24 23:21:08 tgl Exp $
*
*-------------------------------------------------------------------------
*/
@ -86,8 +86,8 @@ make_one_rel(Query *root, List *rels)
* set_base_rel_pathlist
* Finds all paths available for scanning each relation entry in
* 'rels'. Sequential scan and any available indices are considered
* if possible(indices are not considered for lower nesting levels).
* All unique paths are attached to the relation's 'pathlist' field.
* if possible (indices are not considered for lower nesting levels).
* All useful paths are attached to the relation's 'pathlist' field.
*
* MODIFIES: rels
*/
@ -98,21 +98,32 @@ set_base_rel_pathlist(Query *root, List *rels)
foreach(temp, rels)
{
RelOptInfo *rel = (RelOptInfo *) lfirst(temp);
List *indices = find_relation_indices(root, rel);
List *sequential_scan_list;
List *rel_index_scan_list;
List *or_index_scan_list;
RelOptInfo *rel = (RelOptInfo *) lfirst(temp);
sequential_scan_list = lcons(create_seqscan_path(rel), NIL);
rel_index_scan_list = create_index_paths(root,
rel,
find_relation_indices(root, rel),
indices,
rel->restrictinfo,
rel->joininfo);
or_index_scan_list = create_or_index_paths(root, rel, rel->restrictinfo);
/* Note: create_or_index_paths depends on create_index_paths
* to have marked OR restriction clauses with relevant indices;
* this is why it doesn't need to be given the full list of indices.
*/
or_index_scan_list = create_or_index_paths(root, rel,
rel->restrictinfo);
/* add_pathlist will discard any paths that are dominated by
* another available path, keeping only those paths that are
* superior along at least one dimension of cost or sortedness.
*/
rel->pathlist = add_pathlist(rel,
sequential_scan_list,
nconc(rel_index_scan_list,
@ -128,7 +139,6 @@ set_base_rel_pathlist(Query *root, List *rels)
rel->size = compute_rel_size(rel);
rel->width = compute_rel_width(rel);
}
return;
}
/*