diff options
Diffstat (limited to 'sql/sql_select.cc')
| -rw-r--r-- | sql/sql_select.cc | 60 |
1 files changed, 42 insertions, 18 deletions
diff --git a/sql/sql_select.cc b/sql/sql_select.cc index 7ee1762295f..4a32ca34790 100644 --- a/sql/sql_select.cc +++ b/sql/sql_select.cc @@ -13017,6 +13017,34 @@ static int test_if_order_by_key(ORDER *order, TABLE *table, uint idx, } +/** + Find shortest key suitable for full table scan. + + @param table Table to scan + @param usable_keys Allowed keys + + @note + As far as + 1) clustered primary key entry data set is a set of all record + fields (key fields and not key fields) and + 2) secondary index entry data is a union of its key fields and + primary key fields (at least InnoDB and its derivatives don't + duplicate primary key fields there, even if the primary and + the secondary keys have a common subset of key fields), + then secondary index entry data is always a subset of primary key entry. + Unfortunately, key_info[nr].key_length doesn't show the length + of key/pointer pair but a sum of key field lengths only, thus + we can't estimate index IO volume comparing only this key_length + value of secondary keys and clustered PK. + So, try secondary keys first, and choose PK only if there are no + usable secondary covering keys or found best secondary key include + all table fields (i.e. same as PK): + + @return + MAX_KEY no suitable key found + key index otherwise +*/ + uint find_shortest_key(TABLE *table, const key_map *usable_keys) { uint best= MAX_KEY; @@ -13029,23 +13057,6 @@ uint find_shortest_key(TABLE *table, const key_map *usable_keys) uint min_length= (uint) ~0; for (uint nr=0; nr < table->s->keys ; nr++) { - /* - As far as - 1) clustered primary key entry data set is a set of all record - fields (key fields and not key fields) and - 2) secondary index entry data is a union of its key fields and - primary key fields (at least InnoDB and its derivatives don't - duplicate primary key fields there, even if the primary and - the secondary keys have a common subset of key fields), - then secondary index entry data is always a subset of primary key - entry, and the PK is always longer. - Unfortunately, key_info[nr].key_length doesn't show the length - of key/pointer pair but a sum of key field lengths only, thus - we can't estimate index IO volume comparing only this key_length - value of seconday keys and clustered PK. - So, try secondary keys first, and choose PK only if there are no - usable secondary covering keys: - */ if (nr == usable_clustered_pk) continue; if (usable_keys->is_set(nr)) @@ -13058,7 +13069,20 @@ uint find_shortest_key(TABLE *table, const key_map *usable_keys) } } } - return best != MAX_KEY ? best : usable_clustered_pk; + if (usable_clustered_pk != MAX_KEY) + { + /* + If the primary key is clustered and found shorter key covers all table + fields then primary key scan normally would be faster because amount of + data to scan is the same but PK is clustered. + It's safe to compare key parts with table fields since duplicate key + parts aren't allowed. + */ + if (best == MAX_KEY || + table->key_info[best].key_parts >= table->s->fields) + best= usable_clustered_pk; + } + return best; } /** |