LEFT JOIN optimization: Change LEFT JOIN to normal join if possible

mysql-test/r/select.result:
  Added test for LEFT JOIN optimization
mysql-test/t/select.test:
  Added test for LEFT JOIN optimization
sql/item.h:
  LEFT JOIN optimization
sql/item_cmpfunc.cc:
  LEFT JOIN optimization
sql/item_cmpfunc.h:
  LEFT JOIN optimization
sql/item_func.cc:
  LEFT JOIN optimization
sql/item_func.h:
  LEFT JOIN optimization
sql/item_strfunc.cc:
  LEFT JOIN optimization
sql/sql_base.cc:
  Heart of LEFT JOIN optimization

mysql-test/r/select.result

@@ -2569,16 +2569,46 @@ fld1	fld1
 250503	250505
 250504	250505
 250505	250505
+insert into t2 (fld1, companynr) values (999999,99);
 select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null;
 companynr	companyname
+99	NULL
+select count(*) from t2 left join t4 using (companynr) where t4.companynr is not null;
+count(*)
+1199
 explain select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null;
 table	type	possible_keys	key	key_len	ref	rows	Extra
-t2	ALL	NULL	NULL	NULL	NULL	1199	
+t2	ALL	NULL	NULL	NULL	NULL	1200	
 t4	eq_ref	PRIMARY	PRIMARY	1	test.t2.companynr	1	Using where; Not exists
 explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr is null;
 table	type	possible_keys	key	key_len	ref	rows	Extra
 t4	ALL	NULL	NULL	NULL	NULL	12	
-t2	ALL	NULL	NULL	NULL	NULL	1199	Using where; Not exists
+t2	ALL	NULL	NULL	NULL	NULL	1200	Using where; Not exists
+delete from t2 where fld1=999999;
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
+t4	eq_ref	PRIMARY	PRIMARY	1	test.t2.companynr	1	
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
+t4	eq_ref	PRIMARY	PRIMARY	1	test.t2.companynr	1	
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 and t4.companynr > 0;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
+t4	eq_ref	PRIMARY	PRIMARY	1	test.t2.companynr	1	Using where
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr is null;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t4	ALL	NULL	NULL	NULL	NULL	12	
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0 or t4.companynr > 0;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t4	ALL	PRIMARY	NULL	NULL	NULL	12	
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where ifnull(t2.companynr,1)>0;
+table	type	possible_keys	key	key_len	ref	rows	Extra
+t4	ALL	NULL	NULL	NULL	NULL	12	
+t2	ALL	NULL	NULL	NULL	NULL	1199	Using where
 select distinct t2.companynr,t4.companynr from t2,t4 where t2.companynr=t4.companynr+1;
 companynr	companynr
 37	36

mysql-test/t/select.test

@@ -1527,10 +1527,24 @@ select t2.fld1,t22.fld1 from t2,t2 t22 where t2.fld1 >= 250501 and t2.fld1 <= 25
 #
 # Test of left join.
 #
+insert into t2 (fld1, companynr) values (999999,99);
 
 select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null;
+select count(*) from t2 left join t4 using (companynr) where t4.companynr is not null;
 explain select t2.companynr,companyname from t2 left join t4 using (companynr) where t4.companynr is null;
 explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr is null;
+delete from t2 where fld1=999999;
+
+#
+# Test left join optimization
+
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0;
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0;
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 and t4.companynr > 0;
+# Following can't be optimized
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr is null;
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where t2.companynr > 0 or t2.companynr < 0 or t4.companynr > 0;
+explain select t2.companynr,companyname from t4 left join t2 using (companynr) where ifnull(t2.companynr,1)>0;
 
 #
 # Joins with forms.

sql/item.h

@@ -71,7 +71,24 @@ public:
   virtual double  val_result() { return val(); }
   virtual longlong val_int_result() { return val_int(); }
   virtual String *str_result(String* tmp) { return val_str(tmp); }
+  /* bit map of tables used by item */
   virtual table_map used_tables() const { return (table_map) 0L; }
+  /*
+    Return table map of tables that can't be NULL tables (tables that are
+    used in a context where if they would contain a NULL row generated
+    by a LEFT or RIGHT join, the item would not be true).
+    This expression is used on WHERE item to determinate if a LEFT JOIN can be
+    converted to a normal join.
+    Generally this function should return used_tables() if the function
+    would return null if any of the arguments are null
+    As this is only used in the beginning of optimization, the value don't
+    have to be updated in update_used_tables()
+  */
+  virtual table_map not_null_tables() const { return used_tables(); }
+  /*
+    Returns true if this is a simple constant item like an integer, not
+    a constant expression
+  */
   virtual bool basic_const_item() const { return 0; }
   virtual Item *new_item() { return 0; }	/* Only for const items */
   virtual cond_result eq_cmp_result() const { return COND_OK; }

sql/item_cmpfunc.cc

@@ -292,10 +292,12 @@ void Item_func_interval::fix_length_and_dec()
     }
   }
   maybe_null=0; max_length=2;
-  used_tables_cache|=item->used_tables();
+  used_tables_cache|=     item->used_tables();
+  not_null_tables_cache&= item->not_null_tables();
   with_sum_func= with_sum_func || item->with_sum_func;
 }
 
+
 void Item_func_interval::split_sum_func(List<Item> &fields)
 {
   if (item->with_sum_func && item->type() != SUM_FUNC_ITEM)
@@ -1073,8 +1075,9 @@ void Item_func_in::fix_length_and_dec()
   }
   maybe_null= item->maybe_null;
   max_length=2;
-  used_tables_cache|=item->used_tables();
-  const_item_cache&=item->const_item();
+  used_tables_cache|=     item->used_tables();
+  not_null_tables_cache&= item->not_null_tables();
+  const_item_cache&=      item->const_item();
 }
 
 
@@ -1174,14 +1177,21 @@ Item_cond::fix_fields(THD *thd,TABLE_LIST *tables)
   char buff[sizeof(char*)];			// Max local vars in function
   used_tables_cache=0;
   const_item_cache=0;
+  /*
+    and_table_cache is the value that Item_cond_or() returns for
+    not_null_tables()
+  */
+  and_tables_cache= ~(table_map) 0;
 
   if (thd && check_stack_overrun(thd,buff))
     return 0;					// Fatal error flag is set!
   while ((item=li++))
   {
+    table_map tmp_table_map;
     while (item->type() == Item::COND_ITEM &&
 	   ((Item_cond*) item)->functype() == functype())
     {						// Identical function
+
       li.replace(((Item_cond*) item)->list);
       ((Item_cond*) item)->list.empty();
 #ifdef DELETE_ITEMS
@@ -1193,9 +1203,12 @@ Item_cond::fix_fields(THD *thd,TABLE_LIST *tables)
       item->top_level_item();
     if (item->fix_fields(thd,tables))
       return 1; /* purecov: inspected */
-    used_tables_cache|=item->used_tables();
-    with_sum_func= with_sum_func || item->with_sum_func;
-    const_item_cache&=item->const_item();
+    used_tables_cache|=     item->used_tables();
+    tmp_table_map=	    item->not_null_tables();
+    not_null_tables_cache|= tmp_table_map;
+    and_tables_cache&=      tmp_table_map;
+    const_item_cache&=	    item->const_item();
+    with_sum_func=	    with_sum_func || item->with_sum_func;
     if (item->maybe_null)
       maybe_null=1;
   }
@@ -1234,17 +1247,19 @@ Item_cond::used_tables() const
   return used_tables_cache;
 }
 
+
 void Item_cond::update_used_tables()
 {
-  used_tables_cache=0;
-  const_item_cache=1;
   List_iterator_fast<Item> li(list);
   Item *item;
+
+  used_tables_cache=0;
+  const_item_cache=1;
   while ((item=li++))
   {
     item->update_used_tables();
-    used_tables_cache|=item->used_tables();
-    const_item_cache&= item->const_item();
+    used_tables_cache|= item->used_tables();
+    const_item_cache&=  item->const_item();
   }
 }
 
@@ -1348,12 +1363,16 @@ Item *and_expressions(Item *a, Item *b, Item **org_item)
   {
     Item_cond *res;
     if ((res= new Item_cond_and(a, (Item*) b)))
+    {
       res->used_tables_cache= a->used_tables() | b->used_tables();
+      res->not_null_tables_cache= a->not_null_tables() | b->not_null_tables();
+    }
     return res;
   }
   if (((Item_cond_and*) a)->add((Item*) b))
     return 0;
   ((Item_cond_and*) a)->used_tables_cache|= b->used_tables();
+  ((Item_cond_and*) a)->not_null_tables_cache|= b->not_null_tables();
   return a;
 }
 
@@ -1489,6 +1508,8 @@ Item_func_regex::fix_fields(THD *thd,TABLE_LIST *tables)
   max_length=1; decimals=0;
   binary=args[0]->binary || args[1]->binary;
   used_tables_cache=args[0]->used_tables() | args[1]->used_tables();
+  not_null_tables_cache= (args[0]->not_null_tables() |
+			  args[1]->not_null_tables());
   const_item_cache=args[0]->const_item() && args[1]->const_item();
   if (!regex_compiled && args[1]->const_item())
   {

sql/item_cmpfunc.h

@@ -204,7 +204,7 @@ public:
   enum Item_result result_type () const { return cached_result_type; }
   void fix_length_and_dec();
   const char *func_name() const { return "ifnull"; }
-  unsigned int size_of() { return sizeof(*this);}  
+  table_map not_null_tables() const { return 0; }
 };
 
 
@@ -224,7 +224,7 @@ public:
   }
   void fix_length_and_dec();
   const char *func_name() const { return "if"; }
-  unsigned int size_of() { return sizeof(*this);}  
+  table_map not_null_tables() const { return 0; }
 };
 
 
@@ -239,7 +239,7 @@ public:
   enum Item_result result_type () const { return cached_result_type; }
   void fix_length_and_dec();
   const char *func_name() const { return "nullif"; }
-  unsigned int size_of() { return sizeof(*this);}  
+  table_map not_null_tables() const { return 0; }
 };
 
 
@@ -254,9 +254,10 @@ public:
   void fix_length_and_dec();
   enum Item_result result_type () const { return cached_result_type; }
   const char *func_name() const { return "coalesce"; }
-  unsigned int size_of() { return sizeof(*this);}  
+  table_map not_null_tables() const { return 0; }
 };
 
+
 class Item_func_case :public Item_func
 {
   Item * first_expr, *else_expr;
@@ -270,6 +271,7 @@ public:
   String *val_str(String *);
   void fix_length_and_dec();
   void update_used_tables();
+  table_map not_null_tables() const { return 0; }
   enum Item_result result_type () const { return cached_result_type; }
   const char *func_name() const { return "case"; }
   void print(String *str);
@@ -479,10 +481,12 @@ public:
       }
     }
   }
+  table_map not_null_tables() const { return 0; }
   optimize_type select_optimize() const { return OPTIMIZE_NULL; }
   unsigned int size_of() { return sizeof(*this);}  
 };
 
+
 class Item_func_isnotnull :public Item_bool_func
 {
 public:
@@ -495,9 +499,10 @@ public:
   }
   const char *func_name() const { return "isnotnull"; }
   optimize_type select_optimize() const { return OPTIMIZE_NULL; }
-  unsigned int size_of() { return sizeof(*this);}  
+  table_map not_null_tables() const { return 0; }
 };
 
+
 class Item_func_like :public Item_bool_func2
 {
   char escape;
@@ -572,6 +577,8 @@ class Item_cond :public Item_bool_func
 protected:
   List<Item> list;
   bool abort_on_null;
+  table_map and_tables_cache;
+
 public:
   /* Item_cond() is only used to create top level items */
   Item_cond() : Item_bool_func(), abort_on_null(1) { const_item_cache=0; }
@@ -611,6 +618,7 @@ public:
   enum Functype functype() const { return COND_OR_FUNC; }
   longlong val_int();
   const char *func_name() const { return "or"; }
+  table_map not_null_tables() const { return and_tables_cache; }
 };
 
 

sql/item_func.cc

@@ -61,7 +61,7 @@ Item_func::fix_fields(THD *thd,TABLE_LIST *tables)
   Item **arg,**arg_end;
   char buff[STACK_BUFF_ALLOC];			// Max argument in function
   binary=0;
-  used_tables_cache=0;
+  used_tables_cache= not_null_tables_cache= 0;
   const_item_cache=1;
 
   if (thd && check_stack_overrun(thd,buff))
@@ -78,8 +78,9 @@ Item_func::fix_fields(THD *thd,TABLE_LIST *tables)
       if (item->binary)
 	binary=1;
       with_sum_func= with_sum_func || item->with_sum_func;
-      used_tables_cache|=item->used_tables();
-      const_item_cache&= item->const_item();
+      used_tables_cache|=     item->used_tables();
+      not_null_tables_cache|= item->not_null_tables();
+      const_item_cache&=      item->const_item();
     }
   }
   fix_length_and_dec();
@@ -122,6 +123,13 @@ table_map Item_func::used_tables() const
   return used_tables_cache;
 }
 
+
+table_map Item_func::not_null_tables() const
+{
+  return not_null_tables_cache;
+}
+
+
 void Item_func::print(String *str)
 {
   str->append(func_name());

sql/item_func.h

@@ -34,7 +34,7 @@ protected:
   Item **args,*tmp_arg[2];
 public:
   uint arg_count;
-  table_map used_tables_cache;
+  table_map used_tables_cache, not_null_tables_cache;
   bool const_item_cache;
   enum Functype { UNKNOWN_FUNC,EQ_FUNC,EQUAL_FUNC,NE_FUNC,LT_FUNC,LE_FUNC,
 		  GE_FUNC,GT_FUNC,FT_FUNC,
@@ -97,6 +97,7 @@ public:
   bool fix_fields(THD *,struct st_table_list *);
   void make_field(Send_field *field);
   table_map used_tables() const;
+  table_map not_null_tables() const;
   void update_used_tables();
   bool eq(const Item *item, bool binary_cmp) const;
   virtual optimize_type select_optimize() const { return OPTIMIZE_NONE; }
@@ -588,7 +589,8 @@ public:
   void split_sum_func(List<Item> &fields);
   void update_used_tables()
   {
-    item->update_used_tables() ; Item_func::update_used_tables();
+    item->update_used_tables();
+    Item_func::update_used_tables();
     used_tables_cache|= item->used_tables();
     const_item_cache&=  item->const_item();
   }
@@ -597,6 +599,7 @@ public:
   {
     maybe_null=0; max_length=3;
     used_tables_cache|= item->used_tables();
+    not_null_tables_cache&= item->not_null_tables();
     const_item_cache&=  item->const_item();
     with_sum_func= with_sum_func || item->with_sum_func;
   }
@@ -736,6 +739,7 @@ public:
     return res;
   }
   Item_result result_type () const { return udf.result_type(); }
+  table_map not_null_tables() const { return 0; }
   unsigned int size_of() { return sizeof(*this);}  
 };
 
@@ -969,6 +973,7 @@ public:
   }
   enum Functype functype() const { return FT_FUNC; }
   void update_used_tables() {}
+  table_map not_null_tables() const { return 0; }
   bool fix_fields(THD *thd,struct st_table_list *tlist);
   bool eq(const Item *, bool binary_cmp) const;
   longlong val_int() { return val()!=0.0; }

sql/item_strfunc.cc

@@ -606,9 +606,10 @@ void Item_func_concat_ws::fix_length_and_dec()
     max_length=MAX_BLOB_WIDTH;
     maybe_null=1;
   }
-  used_tables_cache|=separator->used_tables();
-  const_item_cache&=separator->const_item();
-  with_sum_func= with_sum_func || separator->with_sum_func;
+  used_tables_cache|=     separator->used_tables();
+  not_null_tables_cache&= separator->not_null_tables();
+  const_item_cache&=	  separator->const_item();
+  with_sum_func=	  with_sum_func || separator->with_sum_func;
 }
 
 void Item_func_concat_ws::update_used_tables()
@@ -1509,8 +1510,9 @@ void Item_func_elt::fix_length_and_dec()
   }
   maybe_null=1;					// NULL if wrong first arg
   with_sum_func= with_sum_func || item->with_sum_func;
-  used_tables_cache|=item->used_tables();
-  const_item_cache&=item->const_item();
+  used_tables_cache|=	  item->used_tables();
+  not_null_tables_cache&= item->not_null_tables();
+  const_item_cache&=	  item->const_item();
 }
 
 
@@ -1591,8 +1593,9 @@ void Item_func_make_set::fix_length_and_dec()
   max_length=arg_count-1;
   for (uint i=1 ; i < arg_count ; i++)
     max_length+=args[i]->max_length;
-  used_tables_cache|=item->used_tables();
-  const_item_cache&=item->const_item();
+  used_tables_cache|=	  item->used_tables();
+  not_null_tables_cache&= item->not_null_tables();
+  const_item_cache&=	  item->const_item();
   with_sum_func= with_sum_func || item->with_sum_func;
 }
 

sql/sql_base.cc

@@ -1903,11 +1903,11 @@ bool setup_tables(TABLE_LIST *tables)
 
     table->used_fields=0;
     table->const_table=0;
-    table->outer_join=table->null_row=0;
+    table->null_row=0;
     table->status=STATUS_NO_RECORD;
     table->keys_in_use_for_query= table->keys_in_use;
     table->used_keys= table->keys_for_keyread;
-    table->maybe_null=test(table->outer_join=table_list->outer_join);
+    table->maybe_null=test(table->outer_join= table_list->outer_join);
     table->tablenr=tablenr;
     table->map= (table_map) 1 << tablenr;
     table->force_index= table_list->force_index;
@@ -2027,6 +2027,7 @@ insert_fields(THD *thd,TABLE_LIST *tables, const char *db_name,
 
 int setup_conds(THD *thd,TABLE_LIST *tables,COND **conds)
 {
+  table_map not_null_tables= 0;
   DBUG_ENTER("setup_conds");
   thd->set_query_id=1;
   thd->cond_count=0;
@@ -2036,6 +2037,7 @@ int setup_conds(THD *thd,TABLE_LIST *tables,COND **conds)
     thd->where="where clause";
     if ((*conds)->fix_fields(thd,tables))
       DBUG_RETURN(1);
+    not_null_tables= (*conds)->not_null_tables();
   }
 
   /* Check if we are using outer joins */
@@ -2049,9 +2051,15 @@ int setup_conds(THD *thd,TABLE_LIST *tables,COND **conds)
 	DBUG_RETURN(1);
       thd->cond_count++;
 
-      /* If it's a normal join, add the ON/USING expression to the WHERE */
-      if (!table->outer_join)
+      /*
+	If it's a normal join or a LEFT JOIN which can be optimized away
+	add the ON/USING expression to the WHERE
+      */
+      if (!table->outer_join ||
+	  ((table->table->map & not_null_tables) &&
+	   !(specialflag & SPECIAL_NO_NEW_FUNC)))
       {
+	table->outer_join= 0;
 	if (!(*conds=and_conds(*conds, table->on_expr)))
 	  DBUG_RETURN(1);
 	table->on_expr=0;