WL#2985 "Partition pruning", postreview fixes: Small code fixes and better comments

mysql-test/r/partition.result:
  Added testcase for BUG#15819
mysql-test/t/partition.test:
  Added testcase for BUG#15819
sql/item.h:
  WL#2985 "Partition pruning", postreview fixes: better comments
sql/item_timefunc.cc:
  WL#2985 "Partition pruning", postreview fixes: better comments
sql/opt_range.cc:
  WL#2985 "Partition pruning", postreview fixes: 
  - better comments, local function renames
  - Made SEL_ARG::is_singlepoint() to correctly handle NULL edge values.
  - fix uninitialized variable access: s/res |=/res =/
sql/sql_class.cc:
  WL#2985 "Partition pruning", postreview fixes: 
  Set correct max. length of "partitions" column in EXPLAIN output
sql/sql_lex.h:
  WL#2985 "Partition pruning", postreview fixes: better comments
sql/sql_partition.cc:
  WL#2985 "Partition pruning", postreview fixes: better comments
This commit is contained in:
unknown 2005-12-27 15:04:35 +03:00
commit e6bff9a606
8 changed files with 163 additions and 34 deletions

View file

@ -381,13 +381,20 @@ public:
put values of field_i into table record buffer;
return item->val_int();
}
NOTE
At the moment function monotonicity is not well defined (and so may be
incorrect) for Item trees with parameters/return types that are different
from INT_RESULT, may be NULL, or are unsigned.
It will be possible to address this issue once the related partitioning bugs
(BUG#16002, BUG#15447, BUG#13436) are fixed.
*/
typedef enum monotonicity_info
{
NON_MONOTONIC, /* none of the below holds */
MONOTONIC_INCREASING, /* F() is unary and "x < y" => "F(x) < F(y)" */
MONOTONIC_STRICT_INCREASING /* F() is unary and "x < y" => "F(x) <= F(y)" */
MONOTONIC_INCREASING, /* F() is unary and (x < y) => (F(x) <= F(y)) */
MONOTONIC_STRICT_INCREASING /* F() is unary and (x < y) => (F(x) < F(y)) */
} enum_monotonicity_info;
/*************************************************************************/