MDEV-38992 TABLESAMPLE Clause Implementation

mysql-test/main/tablesample.result

@@ -0,0 +1,79 @@
+DROP TABLE IF EXISTS t1, t2;
+DROP VIEW IF EXISTS v1;
+CREATE TABLE t1 (a INT, b INT, KEY idx_b(b));
+CREATE TABLE t2 (a INT, c INT);
+CREATE VIEW v1 AS SELECT * FROM t1;
+SELECT * FROM t1 TABLESAMPLE SYSTEM (10);
+a	b
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (50);
+a	b
+SELECT * FROM t1 TABLESAMPLE SYSTEM (12.5);
+a	b
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (0.5);
+a	b
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (NULL);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'NULL)' at line 1
+SELECT * FROM t1 TABLESAMPLE NONEXISTENTMETHOD (12.5);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'NONEXISTENTMETHOD (12.5)' at line 1
+SELECT * FROM t1 TABLESAMPLE SYSTEM (-12.5);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near '-12.5)' at line 1
+SELECT * FROM t1 TABLESAMPLE SYSTEM (10) JOIN t2 ON t1.a = t2.a;
+a	b	a	c
+SELECT * FROM v1 TABLESAMPLE SYSTEM (20);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+PREPARE stmt2 FROM 'SELECT * FROM t1 TABLESAMPLE BERNOULLI (?)';
+SET @pct = 30;
+EXECUTE stmt2 USING @pct;
+a	b
+SET @pct = 300;
+EXECUTE stmt2 USING @pct;
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+DEALLOCATE PREPARE stmt2;
+DROP PROCEDURE IF EXISTS p1;
+Warnings:
+Note	1305	PROCEDURE test.p1 does not exist
+CREATE PROCEDURE p1(IN sample_pct INT)
+BEGIN
+SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+CALL p1(40);
+a	b
+CALL p1(101);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+DROP PROCEDURE p1;
+CREATE PROCEDURE p1(IN sample_pct DECIMAL)
+BEGIN
+SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+CALL p1(40.11);
+a	b
+Warnings:
+Note	1265	Data truncated for column 'sample_pct' at row 0
+DROP PROCEDURE p1;
+DROP PROCEDURE IF EXISTS p_char;
+Warnings:
+Note	1305	PROCEDURE test.p_char does not exist
+CREATE PROCEDURE p_char(IN sample_pct CHAR(10))
+BEGIN
+SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+ERROR HY000: A variable of a non-numeric based type in TABLESAMPLE clause
+DROP PROCEDURE IF EXISTS p_date;
+Warnings:
+Note	1305	PROCEDURE test.p_date does not exist
+CREATE PROCEDURE p_date(IN sample_pct DATE)
+BEGIN
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (sample_pct);
+END//
+ERROR HY000: A variable of a non-numeric based type in TABLESAMPLE clause
+SELECT * FROM information_schema.tables TABLESAMPLE BERNOULLI (5);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+SELECT * FROM mysql.user TABLESAMPLE SYSTEM (10);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+SELECT * FROM (SELECT * FROM t1) AS derived_tbl TABLESAMPLE SYSTEM (50);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'TABLESAMPLE SYSTEM (50)' at line 1
+WITH cte_tbl AS (SELECT * FROM t1)
+SELECT * FROM cte_tbl TABLESAMPLE BERNOULLI (10);
+ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use
+DROP TABLE t1, t2;
+DROP VIEW v1;

mysql-test/main/tablesample.test

@@ -0,0 +1,133 @@
+#
+#   MDEV-38992 SQL Standard TABLESAMPLE clause
+#
+#   For now, there are tests just for checking syntax
+#
+
+--disable_warnings
+DROP TABLE IF EXISTS t1, t2;
+DROP VIEW IF EXISTS v1;
+--enable_warnings
+
+CREATE TABLE t1 (a INT, b INT, KEY idx_b(b));
+CREATE TABLE t2 (a INT, c INT);
+CREATE VIEW v1 AS SELECT * FROM t1;
+
+#
+# Basic Syntax Validation
+#
+SELECT * FROM t1 TABLESAMPLE SYSTEM (10);
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (50);
+
+SELECT * FROM t1 TABLESAMPLE SYSTEM (12.5);
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (0.5);
+
+--error ER_PARSE_ERROR
+SELECT * FROM t1 TABLESAMPLE BERNOULLI (NULL);
+
+--error ER_PARSE_ERROR
+SELECT * FROM t1 TABLESAMPLE NONEXISTENTMETHOD (12.5);
+
+--error ER_PARSE_ERROR
+SELECT * FROM t1 TABLESAMPLE SYSTEM (-12.5);
+
+SELECT * FROM t1 TABLESAMPLE SYSTEM (10) JOIN t2 ON t1.a = t2.a;
+
+--error ER_SYNTAX_ERROR
+SELECT * FROM v1 TABLESAMPLE SYSTEM (20);
+
+#
+# Testing Prepared statements
+#
+PREPARE stmt2 FROM 'SELECT * FROM t1 TABLESAMPLE BERNOULLI (?)';
+
+SET @pct = 30;
+EXECUTE stmt2 USING @pct;
+
+SET @pct = 300;
+--error ER_SYNTAX_ERROR
+EXECUTE stmt2 USING @pct;
+
+DEALLOCATE PREPARE stmt2;
+
+#
+# Testing stored procedures
+#
+DROP PROCEDURE IF EXISTS p1;
+
+DELIMITER //;
+CREATE PROCEDURE p1(IN sample_pct INT)
+BEGIN
+  SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+DELIMITER ;//
+
+CALL p1(40);
+
+--error ER_SYNTAX_ERROR
+CALL p1(101);
+
+DROP PROCEDURE p1;
+
+DELIMITER //;
+CREATE PROCEDURE p1(IN sample_pct DECIMAL)
+BEGIN
+  SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+DELIMITER ;//
+
+CALL p1(40.11);
+
+DROP PROCEDURE p1;
+
+DROP PROCEDURE IF EXISTS p_char;
+
+DELIMITER //;
+--error ER_WRONG_SPVAR_TYPE_IN_TABLESAMPLE
+CREATE PROCEDURE p_char(IN sample_pct CHAR(10))
+BEGIN
+  SELECT * FROM t1 TABLESAMPLE SYSTEM (sample_pct);
+END//
+DELIMITER ;//
+
+DROP PROCEDURE IF EXISTS p_date;
+
+DELIMITER //;
+--error ER_WRONG_SPVAR_TYPE_IN_TABLESAMPLE
+CREATE PROCEDURE p_date(IN sample_pct DATE)
+BEGIN
+  SELECT * FROM t1 TABLESAMPLE BERNOULLI (sample_pct);
+END//
+DELIMITER ;//
+
+#
+# TABLESAMPLE should not work on system tables
+#
+--error ER_SYNTAX_ERROR
+SELECT * FROM information_schema.tables TABLESAMPLE BERNOULLI (5);
+
+--error ER_SYNTAX_ERROR
+SELECT * FROM mysql.user TABLESAMPLE SYSTEM (10);
+
+#
+# TABLESAMPLE should not work on derived tables
+#
+--error ER_PARSE_ERROR
+SELECT * FROM (SELECT * FROM t1) AS derived_tbl TABLESAMPLE SYSTEM (50);
+
+--error ER_SYNTAX_ERROR
+WITH cte_tbl AS (SELECT * FROM t1)
+SELECT * FROM cte_tbl TABLESAMPLE BERNOULLI (10);
+
+# INSERT INTO t1 VALUES (1, 1), (1, 2), (1, 3), (1, 4), (1, 6), (1, 7), (1, 8), (1, 9), (1, 10);
+# INSERT INTO t2 VALUES (1, 1), (1, 2), (1, 3), (1, 4), (1, 6), (1, 7), (1, 8), (1, 9), (1, 10);
+
+# SELECT * FROM t1 TABLESAMPLE BERNOULLI (50);
+
+# SELECT * FROM t1 TABLESAMPLE BERNOULLI (50) JOIN t2 TABLESAMPLE BERNOULLI (50) ON t1.a = t2.a;
+
+#
+# Cleanup
+#
+DROP TABLE t1, t2;
+DROP VIEW v1;

sql/CMakeLists.txt

@@ -185,6 +185,7 @@ SET (SQL_SOURCE
                sql_type_string.cc
                sql_type_geom.cc sql_type_vector.cc
                item_windowfunc.cc sql_window.cc
+               sql_tablesample.cc
 	       sql_cte.cc
                item_vers.cc
                sql_sequence.cc sql_sequence.h ha_sequence.h

sql/item.cc

@@ -4708,6 +4708,22 @@ bool Item_param::set_from_item(THD *thd, Item *item)
       DBUG_RETURN(set_limit_clause_param(val));
     }
   }
+  if (tablesample_clause_param)
+  {
+    double val= item->val_real();
+    if (item->null_value)
+    {
+      set_null(DTCollation_numeric());
+      set_handler(&type_handler_null);
+      DBUG_RETURN(false);
+    }
+    else
+    {
+      unsigned_flag= item->unsigned_flag;
+      set_handler(item->type_handler());
+      DBUG_RETURN(set_tablesample_clause_param(val));
+    }
+  }
   st_value tmp;
   item->save_in_value(thd, &tmp);
   DBUG_RETURN(set_from_value(thd, tmp, item->type_handler(), *item));

sql/item.h

@@ -526,10 +526,11 @@ class Rewritable_query_parameter
   uint len_in_query;
 
   bool limit_clause_param;
+  bool tablesample_clause_param;
 
   Rewritable_query_parameter(uint pos_in_q= 0, uint len_in_q= 0)
     : pos_in_query(pos_in_q), len_in_query(len_in_q),
-      limit_clause_param(false)
+      limit_clause_param(false), tablesample_clause_param(false)
   { }
 
   virtual ~Rewritable_query_parameter() = default;
@@ -3405,20 +3406,28 @@ class Item_splocal :public Item_sp_variable,
   Field *create_field_for_create_select(MEM_ROOT *root, TABLE *table) override
   { return create_table_field_from_handler(root, table); }
 
-  bool is_valid_limit_clause_variable_with_error() const
+  bool is_valid_numeric_clause_variable_with_error() const
   {
     /*
       In case if the variable has an anchored data type, e.g.:
         DECLARE a TYPE OF t1.a;
       type_handler() is set to &type_handler_null and this
       function detects such variable as not valid in LIMIT.
     */
-    if (type_handler()->is_limit_clause_valid_type())
+    if (type_handler()->is_numeric_clause_valid_type())
       return true;
     my_error(ER_WRONG_SPVAR_TYPE_IN_LIMIT, MYF(0));
     return false;
   }
 
+  bool is_valid_tablesample_clause_variable_with_error() const
+  {
+    if (type_handler()->is_tablesample_clause_valid_type())
+      return true;
+    my_error(ER_WRONG_SPVAR_TYPE_IN_TABLESAMPLE, MYF(0));
+    return false;
+  }
+
 protected:
   Item *shallow_copy(THD *thd) const override
   { return get_item_copy<Item_splocal>(thd, this); }
@@ -4746,6 +4755,12 @@ class Item_param final :public Item_basic_value,
     set_int(nr, MY_INT64_NUM_DECIMAL_DIGITS);
     return !unsigned_flag && value.integer < 0;
   }
+  bool set_tablesample_clause_param(double d)
+  {
+    value.set_handler(&type_handler_double);
+    set_double(d);
+    return !unsigned_flag && value.real < 0;
+  }
   const String *query_val_str(THD *thd, String *str) const;
 
   bool convert_str_value(THD *thd);
@@ -4780,6 +4795,11 @@ class Item_param final :public Item_basic_value,
     return state == SHORT_DATA_VALUE &&
            value.type_handler()->cmp_type() == INT_RESULT;
   }
+  bool has_double_value() const
+  {
+    return state == SHORT_DATA_VALUE &&
+           value.type_handler()->cmp_type() == REAL_RESULT;
+  }
   bool is_stored_routine_parameter() const override { return true; }
   /*
     This method is used to make a copy of a basic constant item when

sql/lex.h

@@ -86,6 +86,7 @@ SYMBOL symbols[] = {
   { "BACKUP",	        SYM(BACKUP_SYM)},
   { "BEFORE",	        SYM(BEFORE_SYM)},
   { "BEGIN",	        SYM(BEGIN_MARIADB_SYM)},
+  { "BERNOULLI",		SYM(BERNOULLI)},
   { "BETWEEN",		SYM(BETWEEN_SYM)},
   { "BIGINT",		SYM(BIGINT)},
   { "BINARY",		SYM(BINARY)},
@@ -663,6 +664,7 @@ SYMBOL symbols[] = {
   { "TABLE",		SYM(TABLE_SYM)},
   { "TABLE_NAME",       SYM(TABLE_NAME_SYM)},
   { "TABLES",		SYM(TABLES)},
+  { "TABLESAMPLE",		SYM(TABLESAMPLE_SYM)},
   { "TABLESPACE",	        SYM(TABLESPACE)},
   { "TABLE_CHECKSUM",	SYM(TABLE_CHECKSUM_SYM)},
   { "TEMPORARY",	SYM(TEMPORARY)},

sql/opt_hints.cc

@@ -958,6 +958,24 @@ void Opt_hints_table::update_index_hint_map(Key_map *keys_to_use,
 
 bool Opt_hints_table::update_index_hint_maps(THD *thd, TABLE *tbl)
 {
+  /*
+    A TABLESAMPLE clause forces a sampling scan of the table and
+    index-based access can bias the result. Ignore any index hints
+    (old- or new-style) entirely and make sure no key is considered
+    usable, regardless of what the hints say.
+  */
+  if (tbl->pos_in_table_list && tbl->pos_in_table_list->tablesample_clause)
+  {
+    tbl->keys_in_use_for_query.clear_all();
+    tbl->keys_in_use_for_group_by.clear_all();
+    tbl->keys_in_use_for_order_by.clear_all();
+    tbl->keys_in_use_for_rowid_filter.clear_all();
+    tbl->covering_keys.clear_all();
+    tbl->force_index= tbl->force_index_join= tbl->force_index_group=
+        tbl->force_index_order= false;
+    return true;  // handled: caller must not also run process_index_hints()
+  }
+
   if (!is_fixed(INDEX_HINT_ENUM) && !is_fixed(JOIN_INDEX_HINT_ENUM) &&
       !is_fixed(GROUP_INDEX_HINT_ENUM) && !is_fixed(ORDER_INDEX_HINT_ENUM) &&
       !is_fixed(ROWID_FILTER_HINT_ENUM))