dbs/mysql-server
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
mysql-server/sql/table.h
Ho Sy Tan 4a107fd10e 2025-03-05 14:31:08
2025-03-05 14:31:37 +07:00

4649 lines
161 KiB
C++
Raw Permalink Blame History

#ifndef TABLE_INCLUDED
#define TABLE_INCLUDED
/* Copyright (c) 2000, 2024, Oracle and/or its affiliates.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License, version 2.0,
as published by the Free Software Foundation.
This program is designed to work with certain software (including
but not limited to OpenSSL) that is licensed under separate terms,
as designated in a particular file or component or in included license
documentation. The authors of MySQL hereby grant you an additional
permission to link the program and your derivative works with the
separately licensed software that they have either included with
the program or referenced in the documentation.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License, version 2.0, for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */
#include <assert.h>
#include <string.h>
#include <sys/types.h>
#include <string>
#include "field_types.h"
#include "lex_string.h"
#include "map_helpers.h"
#include "mem_root_deque.h"
#include "my_alloc.h"
#include "my_base.h"
#include "my_bitmap.h"
#include "my_compiler.h"
#include "mysql/binlog/event/table_id.h" // Table_id
#include "my_inttypes.h"
#include "my_sys.h"
#include "my_table_map.h"
#include "mysql/components/services/bits/mysql_mutex_bits.h"
#include "mysql/components/services/bits/psi_table_bits.h"
#include "mysql/strings/m_ctype.h"
#include "sql/auth/auth_acls.h" // Access_bitmask
#include "sql/dd/types/foreign_key.h" // dd::Foreign_key::enum_rule
#include "sql/enum_query_type.h" // enum_query_type
#include "sql/key.h"
#include "sql/key_spec.h"
#include "sql/mdl.h" // MDL_wait_for_subgraph
#include "sql/mem_root_array.h"
#include "sql/mysqld_cs.h"
#include "sql/opt_costmodel.h" // Cost_model_table
#include "sql/partition_info.h"
#include "sql/record_buffer.h" // Record_buffer
#include "sql/sql_bitmap.h" // Bitmap
#include "sql/sql_const.h"
#include "sql/sql_list.h"
#include "sql/sql_plist.h"
#include "sql/sql_plugin_ref.h"
#include "sql/sql_sort.h" // Sort_result
#include "sql/tablesample.h"
#include "thr_lock.h"
#include "typelib.h"
class Field;
class Field_longlong;
namespace histograms {
class Histogram;
} // namespace histograms
class ACL_internal_schema_access;
class ACL_internal_table_access;
class COND_EQUAL;
class Field_json;
/* Structs that defines the TABLE */
class File_parser;
class Value_generator;
class GRANT_TABLE;
class Handler_share;
class Index_hint;
class Item;
class Item_field;
class Json_diff_vector;
class Json_seekable_path;
class Json_wrapper;
class Opt_hints_qb;
class Opt_hints_table;
class Query_result_union;
class Query_block;
class Query_expression;
class Security_context;
class SortingIterator;
class String;
class THD;
class Table_cache_element;
class Table_histograms;
class Table_histograms_collection;
class Table_ref;
class Table_trigger_dispatcher;
class Temp_table_param;
class handler;
class partition_info;
enum enum_stats_auto_recalc : int;
enum Value_generator_source : short;
enum row_type : int;
struct AccessPath;
struct HA_CREATE_INFO;
struct LEX;
struct NESTED_JOIN;
struct Partial_update_info;
struct TABLE;
struct TABLE_SHARE;
struct handlerton;
struct Name_resolution_context;
using plan_idx = int;
namespace dd {
class Table;
class View;
enum class enum_table_type;
} // namespace dd
class Common_table_expr;
class Sql_table_check_constraint;
using Sql_table_check_constraint_list =
Mem_root_array<Sql_table_check_constraint>;
class Sql_check_constraint_share;
using Sql_check_constraint_share_list =
Mem_root_array<Sql_check_constraint_share>;
typedef Mem_root_array_YY<LEX_CSTRING> Create_col_name_list;
typedef int64 query_id_t;
enum class enum_json_diff_operation;
bool assert_ref_count_is_locked(const TABLE_SHARE *);
bool assert_invalid_dict_is_locked(const TABLE *);
bool assert_invalid_stats_is_locked(const TABLE *);
#define store_record(A, B) \
memcpy((A)->B, (A)->record[0], (size_t)(A)->s->reclength)
#define restore_record(A, B) \
memcpy((A)->record[0], (A)->B, (size_t)(A)->s->reclength)
#define cmp_record(A, B) \
memcmp((A)->record[0], (A)->B, (size_t)(A)->s->reclength)
#define tmp_file_prefix "#sql" /**< Prefix for tmp tables */
#define tmp_file_prefix_length 4
#define TMP_TABLE_KEY_EXTRA 8
#define PLACEHOLDER_TABLE_ROW_ESTIMATE 2
/**
Enumerate possible types of a table from re-execution
standpoint.
Table_ref class has a member of this type.
At prepared statement prepare, this member is assigned a value
as of the current state of the database. Before (re-)execution
of a prepared statement, we check that the value recorded at
prepare matches the type of the object we obtained from the
table definition cache.
@sa check_and_update_table_version()
@sa Execute_observer
@sa Prepared_statement::reprepare()
*/
enum enum_table_ref_type {
/** Initial value set by the parser */
TABLE_REF_NULL = 0,
TABLE_REF_VIEW,
TABLE_REF_BASE_TABLE,
TABLE_REF_I_S_TABLE,
TABLE_REF_TMP_TABLE
};
/**
Enumerate possible status of a identifier name while determining
its validity
*/
enum class Ident_name_check { OK, WRONG, TOO_LONG };
/*************************************************************************/
/**
Object_creation_ctx -- interface for creation context of database objects
(views, stored routines, events, triggers). Creation context -- is a set
of attributes, that should be fixed at the creation time and then be used
each time the object is parsed or executed.
*/
class Object_creation_ctx {
public:
Object_creation_ctx *set_n_backup(THD *thd);
void restore_env(THD *thd, Object_creation_ctx *backup_ctx);
protected:
Object_creation_ctx() = default;
virtual Object_creation_ctx *create_backup_ctx(THD *thd) const = 0;
virtual void delete_backup_ctx() = 0;
virtual void change_env(THD *thd) const = 0;
public:
virtual ~Object_creation_ctx() = default;
};
/*************************************************************************/
/**
Default_object_creation_ctx -- default implementation of
Object_creation_ctx.
*/
class Default_object_creation_ctx : public Object_creation_ctx {
public:
const CHARSET_INFO *get_client_cs() { return m_client_cs; }
const CHARSET_INFO *get_connection_cl() { return m_connection_cl; }
protected:
Default_object_creation_ctx(THD *thd);
Default_object_creation_ctx(const CHARSET_INFO *client_cs,
const CHARSET_INFO *connection_cl);
protected:
Object_creation_ctx *create_backup_ctx(THD *thd) const override;
void delete_backup_ctx() override;
void change_env(THD *thd) const override;
protected:
/**
client_cs stores the value of character_set_client session variable.
The only character set attribute is used.
Client character set is included into query context, because we save
query in the original character set, which is client character set. So,
in order to parse the query properly we have to switch client character
set on parsing.
*/
const CHARSET_INFO *m_client_cs;
/**
connection_cl stores the value of collation_connection session
variable. Both character set and collation attributes are used.
Connection collation is included into query context, because it defines
the character set and collation of text literals in internal
representation of query (item-objects).
*/
const CHARSET_INFO *m_connection_cl;
};
/**
View_creation_ctx -- creation context of view objects.
*/
class View_creation_ctx : public Default_object_creation_ctx {
public:
static View_creation_ctx *create(THD *thd);
static View_creation_ctx *create(THD *thd, Table_ref *view);
private:
View_creation_ctx(THD *thd) : Default_object_creation_ctx(thd) {}
};
/*************************************************************************/
/** Order clause list element */
class Item_rollup_group_item;
struct ORDER {
/// @returns true if item pointer is same as original
bool is_item_original() const { return item[0] == item_initial; }
ORDER *next{nullptr};
/**
If the query block includes non-primitive grouping, then these modifiers are
represented as grouping sets. The variable 'grouping_set_info' functions as
a bitvector, containing the grouping set details. If the 'ith' bit of the
variable is set, then the corresponding element is included in the 'ith'
grouping set. */
MY_BITMAP *grouping_set_info{nullptr};
/**
The initial ordering expression. Usually substituted during resolving
and must not be used during optimization and execution.
*/
Item *item_initial{nullptr}; /* Storage for initial item */
public:
/**
Points at the item in the select fields. Note that this means that
after resolving, it points into a slice (see JOIN::ref_items),
even though the item is not of type Item_ref!
*/
Item **item{&item_initial};
Item_rollup_group_item *rollup_item{nullptr};
enum_order direction{
ORDER_NOT_RELEVANT}; /* Requested direction of ordering */
bool in_field_list{false}; /* true if in select field list */
/**
Tells whether this ORDER element was referenced with an alias or with an
expression in the query, and what the alias was:
SELECT a AS foo GROUP BY foo: "foo".
SELECT a AS foo GROUP BY a: nullptr.
*/
const char *used_alias{nullptr};
/**
When GROUP BY is implemented with a temporary table (i.e. the table takes
care to store only unique group rows, table->group != nullptr), each GROUP
BY expression is stored in a column of the table, which is
'field_in_tmp_table'.
Such field may point into table->record[0] (if we only use it to get its
value from a tmp table's row), or into 'buff' (if we use it to do index
lookup into the tmp table).
*/
Field *field_in_tmp_table{nullptr};
char *buff{nullptr}; /* If tmp-table group */
table_map used{0}, depend_map{0};
bool is_explicit{false}; /* Whether ASC/DESC is explicitly specified */
};
/**
State information for internal tables grants.
This structure is part of the Table_ref, and is updated
during the ACL check process.
@sa GRANT_INFO
*/
struct GRANT_INTERNAL_INFO {
/** True if the internal lookup by schema name was done. */
bool m_schema_lookup_done{false};
/** Cached internal schema access. */
const ACL_internal_schema_access *m_schema_access{nullptr};
/** True if the internal lookup by table name was done. */
bool m_table_lookup_done{false};
/** Cached internal table access. */
const ACL_internal_table_access *m_table_access{nullptr};
};
/**
@brief The current state of the privilege checking process for the current
user, SQL statement and SQL object.
@details The privilege checking process is divided into phases depending on
the level of the privilege to be checked and the type of object to be
accessed. Due to the mentioned scattering of privilege checking
functionality, it is necessary to keep track of the state of the process.
A GRANT_INFO also serves as a cache of the privilege hash tables. Relevant
members are grant_table and version.
*/
struct GRANT_INFO {
GRANT_INFO();
/**
@brief A copy of the privilege information regarding the current host,
database, object and user.
@details The version of this copy is found in GRANT_INFO::version.
*/
GRANT_TABLE *grant_table{nullptr};
/**
@brief Used for cache invalidation when caching privilege information.
@details The privilege information is stored on disk, with dedicated
caches residing in memory: table-level and column-level privileges,
respectively, have their own dedicated caches.
The GRANT_INFO works as a level 1 cache with this member updated to the
current value of the global variable @c grant_version (@c static variable
in sql_acl.cc). It is updated Whenever the GRANT_INFO is refreshed from
the level 2 cache. The level 2 cache is the @c column_priv_hash structure
(@c static variable in sql_acl.cc)
@see grant_version
*/
uint version{0};
/**
@brief The set of privileges that the current user has fulfilled for a
certain host, database, and object.
@details This field is continually updated throughout the access checking
process. In each step the "wanted privilege" is checked against the
fulfilled privileges. When/if the intersection of these sets is empty,
access is granted.
The set is implemented as a bitmap, with the bits defined in sql_acl.h.
*/
Access_bitmask privilege{0};
/** The grant state for internal tables. */
GRANT_INTERNAL_INFO m_internal;
};
enum tmp_table_type {
NO_TMP_TABLE,
NON_TRANSACTIONAL_TMP_TABLE,
TRANSACTIONAL_TMP_TABLE,
INTERNAL_TMP_TABLE,
SYSTEM_TMP_TABLE
};
/**
Category of table found in the table share.
*/
enum enum_table_category {
/**
Unknown value.
*/
TABLE_UNKNOWN_CATEGORY = 0,
/**
Temporary table.
The table is visible only in the session.
Therefore,
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
do not apply to this table.
Note that LOCK TABLE t FOR READ/WRITE
can be used on temporary tables.
Temporary tables are not part of the table cache.
2016-06-14 Contrary to what's written in these comments, the truth is:
- tables created by CREATE TEMPORARY TABLE have TABLE_CATEGORY_USER
- tables created by create_tmp_table() (internal ones) have
TABLE_CATEGORY_TEMPORARY.
ha_innodb.cc relies on this observation (so: grep it). If you clean this
up, you may also want to look at 'no_tmp_table'; its enum values' meanings
have degraded over time: INTERNAL_TMP_TABLE is not used for some internal
tmp tables (derived tables). Unification of both enums would be
great. Whatever the result, we need to be able to distinguish the two
types of temporary tables above, as usage patterns are more restricted for
the second type, and allow more optimizations.
*/
TABLE_CATEGORY_TEMPORARY = 1,
/**
User table.
These tables do honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
User tables are cached in the table cache.
*/
TABLE_CATEGORY_USER = 2,
/**
System table, maintained by the server.
These tables do honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
Typically, writes to system tables are performed by
the server implementation, not explicitly be a user.
System tables are cached in the table cache.
*/
TABLE_CATEGORY_SYSTEM = 3,
/**
Information schema tables.
These tables are an interface provided by the system
to inspect the system metadata.
These tables do *not* honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
as there is no point in locking explicitly
an INFORMATION_SCHEMA table.
Nothing is directly written to information schema tables.
Note that this value is not used currently,
since information schema tables are not shared,
but implemented as session specific temporary tables.
*/
/*
TODO: Fixing the performance issues of I_S will lead
to I_S tables in the table cache, which should use
this table type.
*/
TABLE_CATEGORY_INFORMATION = 4,
/**
Log tables.
These tables are an interface provided by the system
to inspect the system logs.
These tables do *not* honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
as there is no point in locking explicitly
a LOG table.
An example of LOG tables are:
- mysql.slow_log
- mysql.general_log,
which *are* updated even when there is either
a GLOBAL READ LOCK or a GLOBAL READ_ONLY in effect.
User queries do not write directly to these tables
(there are exceptions for log tables).
The server implementation perform writes.
Log tables are cached in the table cache.
*/
TABLE_CATEGORY_LOG = 5,
/**
Performance schema tables.
These tables are an interface provided by the system
to inspect the system performance data.
These tables do *not* honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
as there is no point in locking explicitly
a PERFORMANCE_SCHEMA table.
An example of PERFORMANCE_SCHEMA tables are:
- performance_schema.*
which *are* updated (but not using the handler interface)
even when there is either
a GLOBAL READ LOCK or a GLOBAL READ_ONLY in effect.
User queries do not write directly to these tables
(there are exceptions for SETUP_* tables).
The server implementation perform writes.
Performance tables are cached in the table cache.
*/
TABLE_CATEGORY_PERFORMANCE = 6,
/**
Replication Information Tables.
These tables are used to store replication information.
These tables do *not* honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
as there is no point in locking explicitly
a Replication Information table.
An example of replication tables are:
- mysql.slave_master_info
- mysql.slave_relay_log_info,
which *are* updated even when there is either
a GLOBAL READ LOCK or a GLOBAL READ_ONLY in effect.
User queries do not write directly to these tables.
Replication tables are cached in the table cache.
*/
TABLE_CATEGORY_RPL_INFO = 7,
/**
Gtid Table.
The table is used to store gtids.
The table does *not* honor:
- LOCK TABLE t FOR READ/WRITE
- FLUSH TABLES WITH READ LOCK
- SET GLOBAL READ_ONLY = ON
as there is no point in locking explicitly
a Gtid table.
An example of gtid_executed table is:
- mysql.gtid_executed,
which is updated even when there is either
a GLOBAL READ LOCK or a GLOBAL READ_ONLY in effect.
Gtid table is cached in the table cache.
*/
TABLE_CATEGORY_GTID = 8,
/**
A data dictionary table.
Table's with this category will skip checking the
TABLE_SHARE versions because these table structures
are fixed upon server bootstrap.
*/
TABLE_CATEGORY_DICTIONARY = 9,
/**
A ACL metadata table.
For table in this category we will skip row locks when SQL statement
reads them.
*/
TABLE_CATEGORY_ACL_TABLE = 10
};
typedef enum enum_table_category TABLE_CATEGORY;
extern ulong refresh_version;
struct TABLE_FIELD_TYPE {
LEX_CSTRING name;
LEX_CSTRING type;
LEX_CSTRING cset;
};
struct TABLE_FIELD_DEF {
uint count;
const TABLE_FIELD_TYPE *field;
};
class Table_check_intact {
protected:
bool has_keys;
virtual void report_error(uint code, const char *fmt, ...) = 0;
public:
Table_check_intact() : has_keys(false) {}
virtual ~Table_check_intact() = default;
/**
Checks whether a table is intact. Should be done *just* after the table has
been opened.
@param[in] thd Thread handle
@param[in] table The table to check
@param[in] table_def Expected structure of the table (column name
and type)
@retval false OK
@retval true There was an error.
*/
bool check(THD *thd, TABLE *table, const TABLE_FIELD_DEF *table_def);
};
/**
Class representing the fact that some thread waits for table
share to be flushed. Is used to represent information about
such waits in MDL deadlock detector.
*/
class Wait_for_flush : public MDL_wait_for_subgraph {
MDL_context *m_ctx;
TABLE_SHARE *m_share;
uint m_deadlock_weight;
public:
Wait_for_flush(MDL_context *ctx_arg, TABLE_SHARE *share_arg,
uint deadlock_weight_arg)
: m_ctx(ctx_arg),
m_share(share_arg),
m_deadlock_weight(deadlock_weight_arg) {}
MDL_context *get_ctx() const { return m_ctx; }
bool accept_visitor(MDL_wait_for_graph_visitor *dvisitor) override;
uint get_deadlock_weight() const override;
/**
Pointers for participating in the list of waiters for table share.
*/
Wait_for_flush *next_in_share;
Wait_for_flush **prev_in_share;
};
typedef I_P_List<
Wait_for_flush,
I_P_List_adapter<Wait_for_flush, &Wait_for_flush::next_in_share,
&Wait_for_flush::prev_in_share>>
Wait_for_flush_list;
typedef struct Table_share_foreign_key_info {
LEX_CSTRING referenced_table_db;
LEX_CSTRING referenced_table_name;
/**
Name of unique key matching FK in parent table, "" if there is no
unique key.
*/
LEX_CSTRING unique_constraint_name;
dd::Foreign_key::enum_rule update_rule, delete_rule;
uint columns;
/**
Arrays with names of referencing columns of the FK.
*/
LEX_CSTRING *column_name;
} TABLE_SHARE_FOREIGN_KEY_INFO;
typedef struct Table_share_foreign_key_parent_info {
LEX_CSTRING referencing_table_db;
LEX_CSTRING referencing_table_name;
dd::Foreign_key::enum_rule update_rule, delete_rule;
} TABLE_SHARE_FOREIGN_KEY_PARENT_INFO;
/**
Definition of name for generated keys, owned by TABLE_SHARE
*/
struct Key_name {
char name[NAME_CHAR_LEN];
};
/**
This structure is shared between different table objects. There is one
instance of table share per one table in the database.
*/
struct TABLE_SHARE {
TABLE_SHARE() = default;
/**
Create a new TABLE_SHARE with the given version number.
@param version the version of the TABLE_SHARE
@param secondary set to true if the TABLE_SHARE represents a table
in a secondary storage engine
*/
TABLE_SHARE(unsigned long version, bool secondary)
: m_version(version), m_secondary_engine(secondary) {}
/*
Managed collection of refererence-counted snapshots of histograms statistics
for the table. TABLE objects acquire/release pointers to histogram
statistics from this collection. A new statistics snapshot is inserted when
the share is initialized and when histograms are updated/dropped.
For temporary tables m_histograms should be nullptr since we do not support
histograms on temporary tables.
*/
Table_histograms_collection *m_histograms{nullptr};
/** Category of this table. */
TABLE_CATEGORY table_category{TABLE_UNKNOWN_CATEGORY};
MEM_ROOT mem_root;
/**
Used to allocate new handler for internal temporary table when the
size limitation of the primary storage engine is exceeded.
*/
MEM_ROOT *alloc_for_tmp_file_handler{nullptr};
TYPELIB keynames; /* Pointers to keynames */
TYPELIB *intervals{nullptr}; /* pointer to interval info */
mysql_mutex_t LOCK_ha_data; /* To protect access to ha_data */
TABLE_SHARE *next{nullptr}, **prev{nullptr}; /* Link to unused shares */
/**
Array of table_cache_instances pointers to elements of table caches
respresenting this table in each of Table_cache instances.
Allocated along with the share itself in alloc_table_share().
Each element of the array is protected by Table_cache::m_lock in the
corresponding Table_cache. False sharing should not be a problem in
this case as elements of this array are supposed to be updated rarely.
*/
Table_cache_element **cache_element{nullptr};
/* The following is copied to each TABLE on OPEN */
Field **field{nullptr};
Field **found_next_number_field{nullptr};
KEY *key_info{nullptr}; /* data of keys defined for the table */
uint *blob_field{nullptr}; /* Index to blobs in Field array */
uchar *default_values{nullptr}; /* row with default values */
LEX_STRING comment{nullptr, 0}; /* Comment about table */
LEX_STRING compress{nullptr, 0}; /* Compression algorithm */
LEX_STRING encrypt_type{nullptr, 0}; /* encryption algorithm */
/** Secondary storage engine. */
LEX_CSTRING secondary_engine{nullptr, 0};
/** Secondary engine load status */
bool secondary_load{false};
const CHARSET_INFO *table_charset{
nullptr}; /* Default charset of string fields */
MY_BITMAP all_set;
/*
Key which is used for looking-up table in table cache and in the list
of thread's temporary tables. Has the form of:
"database_name\0table_name\0" + optional part for temporary tables.
Note that all three 'table_cache_key', 'db' and 'table_name' members
must be set (and be non-zero) for tables in table cache. They also
should correspond to each other.
To ensure this one can use set_table_cache() methods.
*/
LEX_CSTRING table_cache_key{nullptr, 0};
LEX_CSTRING db{nullptr, 0}; /* Pointer to db */
LEX_CSTRING table_name{nullptr, 0}; /* Table name (for open) */
LEX_STRING path{nullptr, 0}; /* Path to .frm file (from datadir) */
LEX_CSTRING normalized_path{nullptr, 0}; /* unpack_filename(path) */
LEX_STRING connect_string{nullptr, 0};
LEX_CSTRING engine_attribute = EMPTY_CSTR;
LEX_CSTRING secondary_engine_attribute = EMPTY_CSTR;
/**
The set of indexes that are not disabled for this table. I.e. it excludes
indexes disabled by `ALTER TABLE ... DISABLE KEYS`, however it does
include invisible indexes. The data dictionary populates this bitmap.
*/
Key_map keys_in_use;
/// The set of visible and enabled indexes for this table.
Key_map visible_indexes;
Key_map keys_for_keyread;
ha_rows min_rows{0}, max_rows{0}; /* create information */
ulong avg_row_length{0}; /* create information */
ulong mysql_version{0}; /* 0 if .frm is created before 5.0 */
ulong reclength{0}; /* Recordlength */
ulong stored_rec_length{0}; /* Stored record length
(no generated-only generated fields) */
ulonglong autoextend_size{0};
plugin_ref db_plugin{nullptr}; /* storage engine plugin */
inline handlerton *db_type() const /* table_type for handler */
{
// assert(db_plugin);
return db_plugin ? plugin_data<handlerton *>(db_plugin) : nullptr;
}
/**
Value of ROW_FORMAT option for the table as provided by user.
Can be different from the real row format used by the storage
engine. ROW_TYPE_DEFAULT value indicates that no explicit
ROW_FORMAT was specified for the table. @sa real_row_type.
*/
enum row_type row_type = {}; // Zero-initialized to ROW_TYPE_DEFAULT
/** Real row format used for the table by the storage engine. */
enum row_type real_row_type = {}; // Zero-initialized to ROW_TYPE_DEFAULT
tmp_table_type tmp_table{NO_TMP_TABLE};
/**
Only for internal temporary tables.
Count of TABLEs (having this TABLE_SHARE) which have a "handler"
(table->file!=nullptr) which is open (ha_open() has been called).
*/
uint tmp_handler_count{0};
/**
Only for internal temporary tables.
Count of TABLEs (having this TABLE_SHARE) which have opened this table.
*/
uint tmp_open_count{0};
// Can only be 1,2,4,8 or 16, but use uint32_t since that how it is
// represented in InnoDB
std::uint32_t key_block_size{0}; /* create key_block_size, if used */
uint stats_sample_pages{0}; /* number of pages to sample during
stats estimation, if used, otherwise 0. */
enum_stats_auto_recalc
stats_auto_recalc{}; /* Automatic recalc of stats.
Zero-initialized to HA_STATS_AUTO_RECALC_DEFAULT
*/
uint null_bytes{0}, last_null_bit_pos{0};
uint fields{0}; /* Number of fields */
uint rec_buff_length{0}; /* Size of table->record[] buffer */
uint keys{0}; /* Number of keys defined for the table*/
uint key_parts{0}; /* Number of key parts of all keys
defined for the table
*/
uint max_key_length{0}; /* Length of the longest key */
uint max_unique_length{0}; /* Length of the longest unique key */
uint total_key_length{0};
/**
Whether this is a temporary table that already has a UNIQUE index (removing
duplicate rows on insert), so that the optimizer does not need to run
DISTINCT itself. Also used for INTERSECT and EXCEPT as a fall-back if
hashing fails (secondary overflow of in-memory hash table, in which case
we revert to de-duplication using the unique key in the output table).
*/
bool is_distinct{false};
uint null_fields{0}; /* number of null fields */
uint blob_fields{0}; /* number of blob fields */
uint varchar_fields{0}; /* number of varchar fields */
/**
For materialized derived tables; @see add_derived_key().
'first' means: having the lowest position in key_info.
*/
uint first_unused_tmp_key{0};
/**
For materialized derived tables: allocated size of key_info array.
*/
uint max_tmp_keys{0};
/**
For materialized derived tables: allocated size of base_key_parts array of
all TABLE objects. Used for generated keys.
*/
uint max_tmp_key_parts{0};
/**
Array of names for generated keys, used for materialized derived tables.
Shared among all TABLE objects referring to this table share.
*/
Key_name *key_names{nullptr};
/**
Records per key array, used for materialized derived tables.
This is a contiguous array, with size given by max_tmp_key_parts.
The array is shared with all TABLE objects referring to this table share.
*/
ulong *base_rec_per_key{nullptr};
/**
Records per key array, float rep., used for materialized derived tables.
This is a contiguous array, with size given by max_tmp_key_parts.
The array is shared with all TABLE objects referring to this table share.
*/
rec_per_key_t *base_rec_per_key_float{nullptr};
/**
Bitmap with flags representing some of table options/attributes.
@sa HA_OPTION_PACK_RECORD, HA_OPTION_PACK_KEYS, ...
@note This is basically copy of HA_CREATE_INFO::table_options bitmap
at the time of table opening/usage.
*/
uint db_create_options{0};
/**
Bitmap with flags representing some of table options/attributes which
are in use by storage engine.
@note db_options_in_use is normally copy of db_create_options but can
be overridden by SE. E.g. MyISAM does this at handler::open() and
handler::info() time.
*/
uint db_options_in_use{0};
uint rowid_field_offset{0}; /* Field_nr +1 to rowid field */
/* Primary key index number, used in TABLE::key_info[] */
uint primary_key{0};
uint next_number_index{0}; /* autoincrement key number */
uint next_number_key_offset{0}; /* autoinc keypart offset in a key */
uint next_number_keypart{0}; /* autoinc keypart number in a key */
bool error{false}; /* error during open_table_def() */
uint column_bitmap_size{0};
/// Number of generated fields
uint vfields{0};
/// Number of fields having the default value generated
uint gen_def_field_count{0};
bool system{false}; /* Set if system table (one record) */
bool db_low_byte_first{false}; /* Portable row format */
bool crashed{false};
bool is_view{false};
bool m_open_in_progress{false}; /* True: alloc'ed, false: def opened */
mysql::binlog::event::Table_id table_map_id; /* for row-based replication */
/*
Cache for row-based replication table share checks that does not
need to be repeated. Possible values are: -1 when cache value is
not calculated yet, 0 when table *shall not* be replicated, 1 when
table *may* be replicated.
*/
int cached_row_logging_check{0};
/*
Storage media to use for this table (unless another storage
media has been specified on an individual column - in versions
where that is supported)
*/
ha_storage_media default_storage_media{HA_SM_DEFAULT};
/* Name of the tablespace used for this table */
const char *tablespace{nullptr};
/**
Partition meta data. Allocated from TABLE_SHARE::mem_root,
created when reading from the dd tables,
used as template for each TABLE instance.
The reason for having it on the TABLE_SHARE is to be able to reuse the
partition_elements containing partition names, values etc. instead of
allocating them for each TABLE instance.
TODO: Currently it is filled in and then only used for generating
the partition_info_str. The plan is to clone/copy/reference each
TABLE::part_info instance from it.
What is missing before it can be completed:
1) The partition expression, currently created only during parsing which
also needs the current TABLE instance as context for name resolution etc.
2) The partition values, currently the DD stores them as text so it needs
to be converted to field images (which is now done by first parsing the
value text into an Item, then saving the Item result/value into a field
and then finally copy the field image).
*/
partition_info *m_part_info{nullptr};
// TODO: Remove these four variables:
/**
Filled in when reading from frm.
This can simply be removed when removing the .frm support,
since it is already stored in the new DD.
*/
bool auto_partitioned{false};
/**
Storing the full partitioning clause (PARTITION BY ...) which is used
when creating new partition_info object for each new TABLE object by
parsing this string.
These two will be needed until the missing parts above is fixed.
*/
char *partition_info_str{nullptr};
uint partition_info_str_len{0};
/**
Cache the checked structure of this table.
The pointer data is used to describe the structure that
a instance of the table must have. Each element of the
array specifies a field that must exist on the table.
The pointer is cached in order to perform the check only
once -- when the table is loaded from the disk.
*/
const TABLE_FIELD_DEF *table_field_def_cache{nullptr};
/** Main handler's share */
Handler_share *ha_share{nullptr};
/** Instrumentation for this table share. */
PSI_table_share *m_psi{nullptr};
/**
List of tickets representing threads waiting for the share to be flushed.
*/
Wait_for_flush_list m_flush_tickets;
/**
View object holding view definition read from DD. This object is not
cached, and is owned by the table share. We are not able to read it
on demand since we may then get a cache miss while holding LOCK_OPEN.
*/
const dd::View *view_object{nullptr};
/**
Data-dictionary object describing explicit temporary table represented
by this share. NULL for other table types (non-temporary tables, internal
temporary tables). This object is owned by TABLE_SHARE and should be
deleted along with it.
*/
dd::Table *tmp_table_def{nullptr};
/// For materialized derived tables; @see add_derived_key().
Query_block *owner_of_possible_tmp_keys{nullptr};
/**
Arrays with descriptions of foreign keys in which this table participates
as child or parent. We only cache in them information from dd::Table object
which is sufficient for use by prelocking algorithm/to check if table is
referenced by a foreign key.
*/
uint foreign_keys{0};
TABLE_SHARE_FOREIGN_KEY_INFO *foreign_key{nullptr};
uint foreign_key_parents{0};
TABLE_SHARE_FOREIGN_KEY_PARENT_INFO *foreign_key_parent{nullptr};
// List of check constraint share instances.
Sql_check_constraint_share_list *check_constraint_share_list{nullptr};
/**
Schema's read only mode - ON (true) or OFF (false). This is filled in
when the share is initialized with meta data from DD. If the schema is
altered, the tables and share are removed. This can be done since
ALTER SCHEMA acquires exclusive meta data locks on the tables in the
schema. We set this only for non-temporary tables. Otherwise, the value
of the member below is 'NOT_SET'.
*/
enum class Schema_read_only { NOT_SET, RO_OFF, RO_ON };
Schema_read_only schema_read_only{Schema_read_only::NOT_SET};
/**
Set share's table cache key and update its db and table name appropriately.
@param key_buff Buffer with already built table cache key to be
referenced from share.
@param key_length Key length.
@note
Since 'key_buff' buffer will be referenced from share it should has same
life-time as share itself.
This method automatically ensures that TABLE_SHARE::table_name/db have
appropriate values by using table cache key as their source.
*/
void set_table_cache_key(char *key_buff, size_t key_length) {
table_cache_key.str = key_buff;
table_cache_key.length = key_length;
/*
Let us use the fact that the key is "db/0/table_name/0" + optional
part for temporary tables.
*/
db.str = table_cache_key.str;
db.length = strlen(db.str);
table_name.str = db.str + db.length + 1;
table_name.length = strlen(table_name.str);
}
/**
Set share's table cache key and update its db and table name appropriately.
@param key_buff Buffer to be used as storage for table cache key
(should be at least key_length bytes).
@param key Value for table cache key.
@param key_length Key length.
NOTE
Since 'key_buff' buffer will be used as storage for table cache key
it should has same life-time as share itself.
*/
void set_table_cache_key(char *key_buff, const char *key, size_t key_length) {
memcpy(key_buff, key, key_length);
set_table_cache_key(key_buff, key_length);
}
ulonglong get_table_def_version() const { return table_map_id; }
/** Returns the version of this TABLE_SHARE. */
unsigned long version() const { return m_version; }
/**
Set the version of this TABLE_SHARE to zero. This marks the
TABLE_SHARE for automatic removal from the table definition cache
once it is no longer referenced.
*/
void clear_version();
/** Is this table share being expelled from the table definition cache? */
bool has_old_version() const { return version() != refresh_version; }
/**
Convert unrelated members of TABLE_SHARE to one enum
representing its type.
@todo perhaps we need to have a member instead of a function.
*/
enum enum_table_ref_type get_table_ref_type() const {
if (is_view) return TABLE_REF_VIEW;
switch (tmp_table) {
case NO_TMP_TABLE:
return TABLE_REF_BASE_TABLE;
case SYSTEM_TMP_TABLE:
return TABLE_REF_I_S_TABLE;
default:
return TABLE_REF_TMP_TABLE;
}
}
/**
Return a table metadata version.
* for base tables and views, we return table_map_id.
It is assigned from a global counter incremented for each
new table loaded into the table definition cache (TDC).
* for temporary tables it's table_map_id again. But for
temporary tables table_map_id is assigned from
thd->query_id. The latter is assigned from a thread local
counter incremented for every new SQL statement. Since
temporary tables are thread-local, each temporary table
gets a unique id.
* for everything else (e.g. information schema tables),
the version id is zero.
This choice of version id is a large compromise
to have a working prepared statement validation in 5.1. In
future version ids will be persistent, as described in WL#4180.
Let's try to explain why and how this limited solution allows
to validate prepared statements.
Firstly, sets (in mathematical sense) of version numbers
never intersect for different table types. Therefore,
version id of a temporary table is never compared with
a version id of a view, and vice versa.
Secondly, for base tables and views, we know that each DDL flushes
the respective share from the TDC. This ensures that whenever
a table is altered or dropped and recreated, it gets a new
version id.
Unfortunately, since elements of the TDC are also flushed on
LRU basis, this choice of version ids leads to false positives.
E.g. when the TDC size is too small, we may have a SELECT
* FROM INFORMATION_SCHEMA.TABLES flush all its elements, which
in turn will lead to a validation error and a subsequent
reprepare of all prepared statements. This is
considered acceptable, since as long as prepared statements are
automatically reprepared, spurious invalidation is only
a performance hit. Besides, no better simple solution exists.
For temporary tables, using thd->query_id ensures that if
a temporary table was altered or recreated, a new version id is
assigned. This suits validation needs very well and will perhaps
never change.
Metadata of information schema tables never changes.
Thus we can safely assume 0 for a good enough version id.
Finally, by taking into account table type, we always
track that a change has taken place when a view is replaced
with a base table, a base table is replaced with a temporary
table and so on.
@retval 0 For schema tables, DD tables and system views.
non-0 For bases tables, views and temporary tables.
@sa Table_ref::is_table_ref_id_equal()
*/
ulonglong get_table_ref_version() const;
/** Determine if the table is missing a PRIMARY KEY. */
bool is_missing_primary_key() const {
assert(primary_key <= MAX_KEY);
return primary_key == MAX_KEY;
}
uint find_first_unused_tmp_key(const Key_map &k);
bool visit_subgraph(Wait_for_flush *waiting_ticket,
MDL_wait_for_graph_visitor *gvisitor);
bool wait_for_old_version(THD *thd, struct timespec *abstime,
uint deadlock_weight);
/**
The set of indexes that the optimizer may use when creating an execution
plan.
*/
Key_map usable_indexes(const THD *thd) const;
/** Release resources and free memory occupied by the table share. */
void destroy();
/**
How many TABLE objects use this TABLE_SHARE.
@return the reference count
*/
unsigned int ref_count() const {
assert(assert_ref_count_is_locked(this));
return m_ref_count;
}
/**
Increment the reference count by one.
@return the new reference count
*/
unsigned int increment_ref_count() {
assert(assert_ref_count_is_locked(this));
assert(!m_open_in_progress);
return ++m_ref_count;
}
/**
Decrement the reference count by one.
@return the new reference count
*/
unsigned int decrement_ref_count() {
assert(assert_ref_count_is_locked(this));
assert(!m_open_in_progress);
assert(m_ref_count > 0);
return --m_ref_count;
}
/// Does this TABLE_SHARE represent a table in a primary storage engine?
bool is_primary_engine() const { return !m_secondary_engine; }
/// Does this TABLE_SHARE represent a table in a secondary storage engine?
bool is_secondary_engine() const { return m_secondary_engine; }
/**
Does this TABLE_SHARE represent a primary table that has a shadow
copy in a secondary storage engine?
*/
bool has_secondary_engine() const {
return is_primary_engine() && secondary_engine.str != nullptr;
}
/** Returns whether this table is referenced by a foreign key. */
bool is_referenced_by_foreign_key() const { return foreign_key_parents != 0; }
private:
/// How many TABLE objects use this TABLE_SHARE.
unsigned int m_ref_count{0};
/**
TABLE_SHARE version, if changed the TABLE_SHARE must be reopened.
NOTE: The TABLE_SHARE will not be reopened during LOCK TABLES in
close_thread_tables!!!
*/
unsigned long m_version{0};
protected: // To allow access from unit tests.
/// Does this TABLE_SHARE represent a table in a secondary storage engine?
bool m_secondary_engine{false};
};
/**
Class is used as a BLOB field value storage for
intermediate GROUP_CONCAT results. Used only for
GROUP_CONCAT with DISTINCT or ORDER BY options.
*/
class Blob_mem_storage {
private:
MEM_ROOT storage;
/**
Sign that some values were cut
during saving into the storage.
*/
bool truncated_value;
public:
Blob_mem_storage();
~Blob_mem_storage();
void reset() {
storage.ClearForReuse();
truncated_value = false;
}
/**
Function creates duplicate of 'from'
string in 'storage' MEM_ROOT.
@param from string to copy
@param length string length
@retval Pointer to the copied string.
@retval 0 if an error occurred.
*/
char *store(const char *from, size_t length) {
return (char *)memdup_root(&storage, from, length);
}
void set_truncated_value(bool is_truncated_value) {
truncated_value = is_truncated_value;
}
bool is_truncated_value() const { return truncated_value; }
};
/**
Class that represents a single change to a column value in partial
update of a JSON column.
*/
class Binary_diff final {
/// The offset of the start of the change.
size_t m_offset;
/// The size of the portion that is to be replaced.
size_t m_length;
public:
/**
Create a new Binary_diff object.
@param offset the offset of the beginning of the change
@param length the length of the section that is to be replaced
*/
Binary_diff(size_t offset, size_t length)
: m_offset(offset), m_length(length) {}
/// @return the offset of the changed data
size_t offset() const { return m_offset; }
/// @return the length of the changed data
size_t length() const { return m_length; }
/**
Get a pointer to the start of the replacement data.
@param field the column that is updated
@return a pointer to the start of the replacement data
*/
const char *new_data(const Field *field) const;
/**
Get a pointer to the start of the old data to be replaced.
@param field the column that is updated
@return a pointer to the start of old data to be replaced.
*/
const char *old_data(const Field *field) const;
};
/**
Vector of Binary_diff objects.
The Binary_diff objects in the vector should be ordered on offset, and none
of the diffs should be overlapping or adjacent.
*/
using Binary_diff_vector = Mem_root_array<Binary_diff>;
/**
Flags for TABLE::m_status (maximum 8 bits).
The flags define the state of the row buffer in TABLE::record[0].
*/
/**
STATUS_NOT_STARTED is set when table is not accessed yet.
Neither STATUS_NOT_FOUND nor STATUS_NULL_ROW can be set when this flag is set.
*/
#define STATUS_NOT_STARTED 1
/**
Means we were searching for a row and didn't find it. This is used by
storage engines (@see handler::index_read_map()) and the executor, both
when doing an exact row lookup and advancing a scan (no more rows in range).
*/
#define STATUS_NOT_FOUND 2
/// Reserved for use by multi-table update. Means the row has been updated.
#define STATUS_UPDATED 16
/**
Means that table->null_row is set. This is an artificial NULL-filled row
(one example: in outer join, if no match has been found in inner table).
*/
#define STATUS_NULL_ROW 32
/// Reserved for use by multi-table delete. Means the row has been deleted.
#define STATUS_DELETED 64
/* Information for one open table */
enum index_hint_type { INDEX_HINT_IGNORE, INDEX_HINT_USE, INDEX_HINT_FORCE };
/* Bitmap of table's fields */
typedef Bitmap<MAX_FIELDS> Field_map;
/*
NOTE: Despite being a struct (for historical reasons), TABLE has
a nontrivial destructor.
*/
struct TABLE {
TABLE_SHARE *s{nullptr};
handler *file{nullptr};
TABLE *next{nullptr}, *prev{nullptr};
private:
/**
Links for the lists of used/unused TABLE objects for the particular
table in the specific instance of Table_cache (in other words for
specific Table_cache_element object).
Declared as private to avoid direct manipulation with those objects.
One should use methods of I_P_List template instead.
*/
TABLE *cache_next{nullptr}, **cache_prev{nullptr};
/*
Give Table_cache_element access to the above two members to allow
using them for linking TABLE objects in a list.
*/
friend class Table_cache_element;
public:
// Pointer to the histograms available on the table.
// Protected in the same way as the pointer to the share.
const Table_histograms *histograms{nullptr};
/**
A bitmap marking the hidden generated columns that exists for functional
indexes.
*/
MY_BITMAP fields_for_functional_indexes;
/**
The current session using this table object.
Should be NULL when object is not in use.
For an internal temporary table, it is NULL when the table is closed.
Used for two purposes:
- Signal that the object is in use, and by which session.
- Pass the thread handler to storage handlers.
The field should NOT be used as a general THD reference, instead use
a passed THD reference, or, if there is no such, current_thd.
The reason for this is that we cannot guarantee the field is not NULL.
*/
THD *in_use{nullptr};
Field **field{nullptr}; /* Pointer to fields */
/// Count of hidden fields, if internal temporary table; 0 otherwise.
uint hidden_field_count{0};
uchar *record[2]{nullptr, nullptr}; /* Pointer to records */
uchar *write_row_record{nullptr}; /* Used as optimisation in
THD::write_row */
uchar *insert_values{nullptr}; /* used by INSERT ... UPDATE */
/// Buffer for use in multi-row reads. Initially empty.
Record_buffer m_record_buffer{0, 0, nullptr};
/*
Map of keys that can be used to retrieve all data from this table
needed by the query without reading the row.
*/
Key_map covering_keys;
Key_map quick_keys;
/* Merge keys are all keys that had a column referred to in the query */
Key_map merge_keys;
/*
possible_quick_keys is a superset of quick_keys to use with EXPLAIN of
JOIN-less commands (single-table UPDATE and DELETE).
When explaining regular JOINs, we use JOIN_TAB::keys to output the
"possible_keys" column value. However, it is not available for
single-table UPDATE and DELETE commands, since they don't use JOIN
optimizer at the top level. OTOH they directly use the range optimizer,
that collects all keys usable for range access here.
*/
Key_map possible_quick_keys;
/*
A set of keys that can be used in the query that references this
table.
All indexes disabled on the table's TABLE_SHARE (see TABLE::s) will be
subtracted from this set upon instantiation. Thus for any TABLE t it holds
that t.keys_in_use_for_query is a subset of t.s.keys_in_use. Generally we
must not introduce any new keys here (see setup_tables).
The set is implemented as a bitmap.
*/
Key_map keys_in_use_for_query;
/* Map of keys that can be used to calculate GROUP BY without sorting */
Key_map keys_in_use_for_group_by;
/* Map of keys that can be used to calculate ORDER BY without sorting */
Key_map keys_in_use_for_order_by;
KEY *key_info{nullptr}; /* data of keys defined for the table */
/**
Key part array for generated keys, used for materialized derived tables.
This is a contiguous array, with size given by s->max_tmp_key_parts.
*/
KEY_PART_INFO *base_key_parts{nullptr};
Field *next_number_field{nullptr}; /* Set if next_number is activated */
Field *found_next_number_field{nullptr}; /* Set on open */
/// Pointer to generated columns
Field **vfield{nullptr};
/// Pointer to fields having the default value generated
Field **gen_def_fields_ptr{nullptr};
/// Field used by unique constraint
Field *hash_field{nullptr};
// ----------------------------------------------------------------------
// The next few members are used if this (temporary) file is used solely for
// the materialization/computation of an INTERSECT or EXCEPT set operation
// (in addition to hash_field above used to detect duplicate rows). For
// INTERSECT and EXCEPT, we always use the hash field and compute the shape
// of the result set using m_set_counter. The latter is a hidden field
// located between the hash field and the row proper, only present for
// INTERSECT or EXCEPT materialized in a temporary result table. The
// materialized table has no duplicate rows, relying instead of the embedded
// counter to produce the correct number of duplicates with ALL semantics. If
// we have distinct semantics, we squash duplicates. This all happens in the
// reading step of the tmp table (TableScanIterator::Read),
// cf. m_last_operation_is_distinct. For explanation if the logic of the set
// counter, see MaterializeIterator<Profiler>::MaterializeOperand.
//
/// A priori unlimited. We pass this on to TableScanIterator at construction
/// time, q.v., to limit the number of rows out of an EXCEPT or INTERSECT.
/// For these set operations, we do not know enough to enforce the limit at
/// materialize time (as for UNION): only when reading the rows with
/// TableScanIterator do we check the counters.
/// @todo: Ideally, this limit should be communicated to TableScanIterator in
/// some other way.
ha_rows m_limit_rows{HA_POS_ERROR};
private:
/// The set counter. It points to the field in the materialized table
/// holding the counter used to compute INTERSECT and EXCEPT, in record[0].
/// For EXCEPT [DISTINCT | ALL] and INTERSECT DISTINCT this is a simple 64
/// bits counter. For INTERSECT ALL, it is subdivided into two sub counters
/// cf. class HalfCounter, cf. MaterializeOperand. See set_counter().
Field_longlong *m_set_counter{nullptr};
/// If m_set_counter is set: true if last block has DISTINCT semantics,
/// either because it is marked as such, or because we have computed this
/// to give an equivalent answer. If false, we have ALL semantics.
/// It will be true if any DISTINCT is given in the merged N-ary set
/// operation. See is_distinct().
bool m_last_operation_is_distinct{false};
/// If false, any de-duplication happens via an index on this table
/// (e.g. SELECT DISTINCT, set operation). If true, this table represents the
/// output of a set operation, and de-duplication happens via an in-memory
/// hash map, in which case we do not use any index, unless we get secondary
/// overflow.
bool m_deduplicate_with_hash_map{false};
public:
enum Set_operator_type {
SOT_NONE,
SOT_UNION_ALL,
SOT_UNION_DISTINCT,
SOT_INTERSECT_ALL,
SOT_INTERSECT_DISTINCT,
SOT_EXCEPT_ALL,
SOT_EXCEPT_DISTINCT
};
private:
/// Holds the set operation type
Set_operator_type m_set_op_type{SOT_NONE};
public:
/// Test if this tmp table stores the result of a UNION set operation or
/// a single table.
/// @return true if so, else false.
bool is_union_or_table() const { return m_set_counter == nullptr; }
void set_use_hash_map(bool use_hash_map) {
m_deduplicate_with_hash_map = use_hash_map;
}
bool uses_hash_map() const { return m_deduplicate_with_hash_map; }
/// Returns the set operation type
Set_operator_type set_op_type() {
if (m_set_op_type == SOT_NONE) {
assert(is_union_or_table()); // EXCEPT and INTERSECT are already set up
m_set_op_type = is_distinct() ? SOT_UNION_DISTINCT : SOT_UNION_ALL;
}
return m_set_op_type;
}
bool is_intersect() const {
return m_set_op_type == SOT_INTERSECT_ALL ||
m_set_op_type == SOT_INTERSECT_DISTINCT;
}
bool is_except() const {
return m_set_op_type == SOT_EXCEPT_ALL ||
m_set_op_type == SOT_EXCEPT_DISTINCT;
}
bool is_distinct() const { return m_last_operation_is_distinct; }
/**
Initialize the set counter field pointer and the type of set operation
*other than UNION*.
@param set_counter the field in the materialized table that holds the
counter we use to compute intersect or except
@param except if true, EXCEPT, else INTERSECT
@param distinct if true, the set operation is DISTINCT, else ALL
*/
void set_set_op(Field_longlong *set_counter, bool except, bool distinct) {
m_set_counter = set_counter;
m_last_operation_is_distinct = distinct;
assert(m_set_op_type == SOT_NONE);
m_set_op_type = except ? (distinct ? SOT_EXCEPT_DISTINCT : SOT_EXCEPT_ALL)
: distinct ? SOT_INTERSECT_DISTINCT
: SOT_INTERSECT_ALL;
}
Field_longlong *set_counter() { return m_set_counter; }
//
// end of INTERSECT and EXCEPT specific members
// ----------------------------------------------------------------------
Field *fts_doc_id_field{nullptr}; /* Set if FTS_DOC_ID field is present */
/* Table's triggers, 0 if there are no of them */
Table_trigger_dispatcher *triggers{nullptr};
Table_ref *pos_in_table_list{nullptr}; /* Element referring to this table */
/* Position in thd->locked_table_list under LOCK TABLES */
Table_ref *pos_in_locked_tables{nullptr};
ORDER *group{nullptr};
const char *alias{nullptr}; ///< alias or table name
uchar *null_flags{nullptr}; ///< Pointer to the null flags of record[0]
uchar *null_flags_saved{
nullptr}; ///< Saved null_flags while null_row is true
/* containers */
MY_BITMAP def_read_set, def_write_set, tmp_set, pack_row_tmp_set;
/*
Bitmap of fields that one or more query condition refers to. Only
used if optimizer_condition_fanout_filter is turned 'on'.
Currently, only the WHERE clause and ON clause of inner joins is
taken into account but not ON conditions of outer joins.
Furthermore, HAVING conditions apply to groups and are therefore
not useful as table condition filters.
*/
MY_BITMAP cond_set;
/**
Bitmap of table fields (columns), which are explicitly set in the
INSERT INTO statement. It is declared here to avoid memory allocation
on MEM_ROOT).
@sa fields_set_during_insert.
*/
MY_BITMAP def_fields_set_during_insert;
/**
The read set contains the set of columns that the execution engine needs to
process the query. In particular, it is used to tell the storage engine
which columns are needed. For virtual generated columns, the underlying base
columns are also added, since they are required in order to calculate the
virtual generated columns.
Internal operations in the execution engine that need to move rows between
buffers, such as aggregation, sorting, hash join and set operations, should
rather use read_set_internal, since the virtual generated columns have
already been calculated when the row was read from the storage engine.
Set during resolving; every field that gets resolved, sets its own bit
in the read set. In some cases, we switch the read set around during
various phases; note that it is a pointer.
In addition, for binary logging purposes, the bitmaps are set according
to the settings of @@binlog_row_image. Therefore, for logging purposes,
some additional fields, to those specified by the optimizer, may be
flagged in the read and write sets.
@c TABLE::mark_columns_per_binlog_row_image for additional details.
*/
MY_BITMAP *read_set{nullptr};
MY_BITMAP *write_set{nullptr};
/**
A bitmap of fields that are explicitly referenced by the query. This is
mostly the same as read_set, but it does not include base columns of
referenced virtual generated columns unless the base columns are referenced
explicitly in the query.
This is the read set that should be used for determining which columns to
store in join buffers, aggregation buffers, sort buffers, or similar
operations internal to the execution engine. Both because it is unnecessary
to store the implicitly read base columns in the buffer, since they won't
ever be read out of the buffer anyways, and because the base columns may not
even be possible to read, if a covering index scan is used and the index
only contains the virtual column and not all its base columns.
*/
MY_BITMAP read_set_internal;
/**
A pointer to the bitmap of table fields (columns), which are explicitly set
in the INSERT INTO statement.
fields_set_during_insert points to def_fields_set_during_insert
for base (non-temporary) tables. In other cases, it is NULL.
Triggers can not be defined for temporary tables, so this bitmap does not
matter for temporary tables.
@sa def_fields_set_during_insert.
*/
MY_BITMAP *fields_set_during_insert{nullptr};
/*
The ID of the query that opened and is using this table. Has different
meanings depending on the table type.
Temporary tables:
table->query_id is set to thd->query_id for the duration of a statement
and is reset to 0 once it is closed by the same statement. A non-zero
table->query_id means that a statement is using the table even if it's
not the current statement (table is in use by some outer statement).
Non-temporary tables:
Under pre-locked or LOCK TABLES mode: query_id is set to thd->query_id
for the duration of a statement and is reset to 0 once it is closed by
the same statement. A non-zero query_id is used to control which tables
in the list of pre-opened and locked tables are actually being used.
*/
query_id_t query_id{0};
/*
For each key that has quick_keys.is_set(key) == true: estimate of #records
and max #key parts that range access would use.
*/
ha_rows quick_rows[MAX_KEY]{0};
/* Bitmaps of key parts that =const for the entire join. */
key_part_map const_key_parts[MAX_KEY]{0};
uint quick_key_parts[MAX_KEY]{0};
uint quick_n_ranges[MAX_KEY]{0};
/*
Estimate of number of records that satisfy SARGable part of the table
condition, or table->file->records if no SARGable condition could be
constructed.
This value is used by join optimizer as an estimate of number of records
that will pass the table condition (condition that depends on fields of
this table and constants)
*/
ha_rows quick_condition_rows{0};
uint lock_position{0}; /* Position in MYSQL_LOCK.table */
uint lock_data_start{0}; /* Start pos. in MYSQL_LOCK.locks */
uint lock_count{0}; /* Number of locks */
uint db_stat{0}; /* mode of file as in handler.h */
int current_lock{0}; /* Type of lock on table */
// List of table check constraints.
Sql_table_check_constraint_list *table_check_constraint_list{nullptr};
private:
/**
If true, this table is inner w.r.t. some outer join operation, all columns
are nullable (in the query), and null_row may be true.
*/
bool nullable{false};
uint8 m_status{0}; /* What's in record[0] */
public:
/*
If true, the current table row is considered to have all columns set to
NULL, including columns declared as "not null" (see nullable).
@todo make it private, currently join buffering changes it through a pointer
*/
bool null_row{false};
bool copy_blobs{false}; /* copy_blobs when storing */
/*
TODO: Each of the following flags take up 8 bits. They can just as easily
be put into one single unsigned long and instead of taking up 18
bytes, it would take up 4.
*/
bool force_index{false};
/**
Flag set when the statement contains FORCE INDEX FOR ORDER BY
See Table_ref::process_index_hints().
*/
bool force_index_order{false};
/**
Flag set when the statement contains FORCE INDEX FOR GROUP BY
See Table_ref::process_index_hints().
*/
bool force_index_group{false};
bool const_table{false};
/// True if writes to this table should not write rows and just write keys.
bool no_rows{false};
/**
If set, the optimizer has found that row retrieval should access index
tree only.
*/
bool key_read{false};
/**
Certain statements which need the full row, set this to ban index-only
access.
*/
bool no_keyread{false};
/**
If set, indicate that the table is not replicated by the server.
*/
bool no_replicate{false};
/* To signal that the table is associated with a HANDLER statement */
bool open_by_handler{false};
/**
To indicate that value of the auto_increment field was provided
explicitly by the user or from some other source (e.g. in case of
INSERT ... SELECT, ALTER TABLE or LOAD DATA) and not as default
or result of conversion from NULL value.
@note Since auto_increment fields are always non-NULL we can't find
out using methods of Field class if 0 value stored in such field
was provided explicitly or is result of applying default/conversion
from NULL value. In the former case no new auto_increment value
needs to be generated in MODE_NO_AUTO_VALUE_ON_ZERO mode, while
the latter cases require new value generation. Hence the need
for this flag.
@note Used only in the MODE_NO_AUTO_VALUE_ON_ZERO mode and only
by handler::write_row().
*/
bool autoinc_field_has_explicit_non_null_value{false};
bool alias_name_used{false}; /* true if table_name is alias */
bool get_fields_in_item_tree{false}; /* Signal to fix_field */
private:
/**
This TABLE object is invalid and cannot be reused. TABLE object might have
inconsistent info or handler might not allow some operations.
For example, TABLE might have inconsistent info about partitioning.
We also use this flag to avoid calling handler::reset() for partitioned
InnoDB tables after in-place ALTER TABLE API commit phase and to force
closing table after REPAIR TABLE has failed during its prepare phase as
well.
@note This member can be set only by thread that owns/has opened the
table and while holding its THD::LOCK_thd_data lock.
It can be read without locking by this owner thread, or by some other
thread concurrently after acquiring owner's THD::LOCK_thd_data.
@note The TABLE will not be reopened under LOCK TABLES in
close_thread_tables().
*/
bool m_invalid_dict{false};
/**
This TABLE object is invalid and cannot be reused as it has outdated
rec_per_key and handler stats.
@note This member is protected from concurrent access to it by lock of
Table Cache's partition to which this TABLE object belongs,
*/
bool m_invalid_stats{false};
/**
For tmp tables. true <=> tmp table has been instantiated.
Also indicates that table was successfully opened since
we immediately delete tmp tables which we fail to open.
*/
bool created{false};
public:
/// For a materializable derived or SJ table: true if has been materialized
bool materialized{false};
struct /* field connections */
{
class JOIN_TAB *join_tab{nullptr};
class QEP_TAB *qep_tab{nullptr};
thr_lock_type lock_type{TL_UNLOCK}; /* How table is used */
bool not_exists_optimize{false};
/*
true <=> range optimizer found that there is no rows satisfying
table conditions.
*/
bool impossible_range{false};
} reginfo;
/**
@todo This member should not be declared in-line. That makes it
impossible for any function that does memory allocation to take a const
reference to a TABLE object.
*/
MEM_ROOT mem_root;
/**
Initialized in Item_func_group_concat::setup for appropriate
temporary table if GROUP_CONCAT is used with ORDER BY | DISTINCT
and BLOB field count > 0.
*/
Blob_mem_storage *blob_storage{nullptr};
/**
Not owned by the TABLE; used only from filesort_free_buffers().
See comments on SortingIterator::CleanupAfterQuery().
*/
SortingIterator *sorting_iterator{nullptr};
SortingIterator *duplicate_removal_iterator{nullptr};
/**
The result of applying a unique operation (by row ID) to the table, if done.
In particular, this is done in some forms of index merge.
*/
Sort_result unique_result;
partition_info *part_info{nullptr}; /* Partition related information */
/* If true, all partitions have been pruned away */
bool all_partitions_pruned_away{false};
MDL_ticket *mdl_ticket{nullptr};
private:
/// Cost model object for operations on this table
Cost_model_table m_cost_model;
#ifndef NDEBUG
/**
Internal tmp table sequential number. Increased in the order of
creation. Used for debugging purposes when many tmp tables are used
during execution (e.g several windows with window functions)
*/
uint tmp_table_seq_id{0};
#endif
public:
void reset();
void init(THD *thd, Table_ref *tl);
bool init_tmp_table(THD *thd, TABLE_SHARE *share, MEM_ROOT *m_root,
CHARSET_INFO *charset, const char *alias, Field **fld,
uint *blob_fld, bool is_virtual);
bool fill_item_list(mem_root_deque<Item *> *item_list) const;
void clear_column_bitmaps(void);
void prepare_for_position(void);
void mark_column_used(Field *field, enum enum_mark_columns mark);
void mark_columns_used_by_index_no_reset(uint index, MY_BITMAP *map,
uint key_parts = 0) const;
void mark_columns_used_by_index(uint index);
void mark_auto_increment_column(void);
void mark_columns_needed_for_update(THD *thd, bool mark_binlog_columns);
void mark_columns_needed_for_delete(THD *thd);
void mark_columns_needed_for_insert(THD *thd);
void mark_columns_per_binlog_row_image(THD *thd);
void mark_generated_columns(bool is_update);
void mark_gcol_in_maps(const Field *field);
void mark_check_constraint_columns(bool is_update);
void column_bitmaps_set(MY_BITMAP *read_set_arg, MY_BITMAP *write_set_arg);
inline void column_bitmaps_set_no_signal(MY_BITMAP *read_set_arg,
MY_BITMAP *write_set_arg) {
read_set = read_set_arg;
write_set = write_set_arg;
}
inline void use_all_columns() {
column_bitmaps_set(&s->all_set, &s->all_set);
}
inline void default_column_bitmaps() {
read_set = &def_read_set;
write_set = &def_write_set;
}
void invalidate_dict();
void invalidate_stats();
/**
@note Can be called by thread owning table without additional locking, and
by any other thread which has acquired owner's THD::LOCK_thd_data lock.
*/
inline bool has_invalid_dict() const {
assert(assert_invalid_dict_is_locked(this));
return !db_stat || m_invalid_dict;
}
/// @note Can be called by thread owning Table_cache::m_lock
inline bool has_invalid_stats() {
assert(assert_invalid_stats_is_locked(this));
return m_invalid_stats;
}
/// @returns first non-hidden column
Field **visible_field_ptr() const { return field + hidden_field_count; }
/// @returns count of visible fields
uint visible_field_count() const { return s->fields - hidden_field_count; }
bool alloc_tmp_keys(uint new_key_count, uint new_key_part_count,
bool modify_share);
bool add_tmp_key(Field_map *key_parts, bool invisible, bool modify_share);
void move_tmp_key(int old_idx, bool modify_share);
void drop_unused_tmp_keys(bool modify_share);
void set_keyread(bool flag);
/**
Check whether the given index has a virtual generated columns.
@param index_no the given index to check
@returns true if if index is defined over at least one virtual generated
column
*/
inline bool index_contains_some_virtual_gcol(uint index_no) const {
assert(index_no < s->keys);
return key_info[index_no].flags & HA_VIRTUAL_GEN_KEY;
}
void update_const_key_parts(Item *conds);
bool check_read_removal(uint index);
ptrdiff_t default_values_offset() const {
return (ptrdiff_t)(s->default_values - record[0]);
}
/// @returns true if a storage engine handler object is assigned to table
bool has_storage_handler() const { return file != nullptr; }
/// Set storage handler for temporary table
void set_storage_handler(handler *file_arg) {
// Ensure consistent call order
assert((file == nullptr && file_arg != nullptr) ||
(file != nullptr && file_arg == nullptr));
assert(!is_created());
assert(file_arg->inited == handler::NONE);
file = file_arg;
}
/// Return true if table is instantiated, and false otherwise.
bool is_created() const { return created; }
/**
Set the table as "created", and enable flags in storage engine
that could not be enabled without an instantiated table.
*/
void set_created();
/**
Set the contents of table to be "deleted", ie "not created", after having
deleted the contents.
*/
void set_deleted() { created = materialized = false; }
/// Set table as nullable, ie it is inner wrt some outer join
void set_nullable() { nullable = true; }
/// Return whether table is nullable
bool is_nullable() const { return nullable; }
/// @return true if table contains one or more generated columns
bool has_gcol() const { return vfield; }
/**
Life cycle of the row buffer is as follows:
- The initial state is "not started".
- When reading a row through the storage engine handler, the status is set
as "has row" or "no row", depending on whether a row was found or not.
The "not started" state is cleared, as well as the "null row" state,
the updated state and the deleted state.
- When making a row available in record[0], make sure to update row status
similarly to how the storage engine handler does it.
- If a NULL-extended row is needed in join execution, the "null row" state
is set. Note that this can be combined with "has row" if a row was read
but condition on it was evaluated to false (happens for single-row
lookup), or "no row" if no more rows could be read.
Note also that for the "null row" state, the NULL bits inside the
row are set to one, so the row inside the row buffer is no longer usable,
unless the NULL bits are saved in a separate buffer.
- The "is updated" and "is deleted" states are set when row is updated or
deleted, respectively.
*/
/// Set status for row buffer as "not started"
void set_not_started() {
m_status = STATUS_NOT_STARTED | STATUS_NOT_FOUND;
null_row = false;
}
/// @return true if a row operation has been done
bool is_started() const { return !(m_status & STATUS_NOT_STARTED); }
/// Set status for row buffer: contains row
void set_found_row() {
m_status = 0;
null_row = false;
}
/**
Set status for row buffer: contains no row. This is set when
- A lookup operation finds no row
- A scan operation scans past the last row of the range.
- An error in generating key values before calling storage engine.
*/
void set_no_row() {
m_status = STATUS_NOT_FOUND;
null_row = false;
}
/**
Set "row found" status from handler result
@param status 0 if row was found, <> 0 if row was not found
*/
void set_row_status_from_handler(int status) {
m_status = status ? STATUS_NOT_FOUND : 0;
null_row = false;
}
/**
Set current row as "null row", for use in null-complemented outer join.
The row buffer may or may not contain a valid row.
set_null_row() and reset_null_row() are used by the join executor to
signal the presence or absence of a NULL-extended row for an outer joined
table. Null rows may also be used to specify rows that are all NULL in
grouing operations.
@note this is a destructive operation since the NULL value bit vector
is overwritten. Caching operations must be aware of this.
*/
void set_null_row() {
null_row = true;
m_status |= STATUS_NULL_ROW;
if (s->null_bytes > 0) memset(null_flags, 255, s->null_bytes);
}
/// Clear "null row" status for the current row
void reset_null_row() {
null_row = false;
m_status &= ~STATUS_NULL_ROW;
}
/// Set "updated" property for the current row
void set_updated_row() {
assert(is_started() && has_row());
m_status |= STATUS_UPDATED;
}
/// Set "deleted" property for the current row
void set_deleted_row() {
assert(is_started() && has_row());
m_status |= STATUS_DELETED;
}
/// @return true if there is a row in row buffer
bool has_row() const { return !(m_status & STATUS_NOT_FOUND); }
/// @return true if current row is null-extended
bool has_null_row() const { return null_row; }
/// @return true if current row has been updated (multi-table update)
bool has_updated_row() const { return m_status & STATUS_UPDATED; }
/// @return true if current row has been deleted (multi-table delete)
bool has_deleted_row() const { return m_status & STATUS_DELETED; }
/// Save the NULL flags of the current row into the designated buffer.
/// This should be done before null-complementing a table accessed
/// with EQRefIterator or a const table, as they need to be able to
/// restore the original contents of the record buffer before
/// reading the next row. This is necessary because of their special
/// code for avoiding table access if the same row should be
/// accessed by the next read.
void save_null_flags() {
if (s->null_bytes > 0) memcpy(null_flags_saved, null_flags, s->null_bytes);
}
/// Restore the NULL flags of the current row from the designated buffer
void restore_null_flags() {
if (s->null_bytes > 0) memcpy(null_flags, null_flags_saved, s->null_bytes);
}
/// Empties internal temporary table (deletes rows, closes scan)
bool empty_result_table();
/**
Initialize the optimizer cost model.
This function should be called each time a new query is started.
@param cost_model_server the main cost model object for the query
*/
void init_cost_model(const Cost_model_server *cost_model_server) {
m_cost_model.init(cost_model_server, this);
}
/**
Return the cost model object for this table.
*/
const Cost_model_table *cost_model() const { return &m_cost_model; }
/**
Bind all the table's value generator columns in all the forms:
stored/virtual GC, default expressions and checked constraints.
@details When a table is opened from the dictionary, the Value Generator
expressions are bound during opening (see fix_value_generator_fields()).
After query execution, Item::cleanup() is called on them
(see cleanup_value_generator_items()). When the table is opened from the
table cache, the Value Generetor(s) need to be bound again and this
function does that.
*/
void bind_value_generators_to_fields();
/**
Clean any state in items associated with generated columns to be ready for
the next statement.
*/
void cleanup_value_generator_items();
#ifndef NDEBUG
void set_tmp_table_seq_id(uint arg) { tmp_table_seq_id = arg; }
#endif
/**
Update covering keys depending on max read key length.
Update available covering keys for the table, based on a constrained field
and the identified covering prefix keys: If the matched part of field is
longer than the index prefix,
the prefix index cannot be used as a covering index.
@param[in] field Pointer to field object
@param[in] key_read_length Max read key length
@param[in] covering_prefix_keys Covering prefix keys
*/
void update_covering_prefix_keys(Field *field, uint16 key_read_length,
Key_map *covering_prefix_keys);
/**
Returns the primary engine handler for the table.
If none exist, nullptr is returned.
*/
handler *get_primary_handler() const;
private:
/**
Bitmap that tells which columns are eligible for partial update in an
update statement.
The bitmap is lazily allocated in the TABLE's mem_root when
#mark_column_for_partial_update() is called.
*/
MY_BITMAP *m_partial_update_columns{nullptr};
/**
Object which contains execution time state used for partial update
of JSON columns.
It is allocated in the execution mem_root by #setup_partial_update() if
there are columns that have been marked as eligible for partial update.
*/
Partial_update_info *m_partial_update_info{nullptr};
/**
This flag decides whether or not we should log the drop temporary table
command.
*/
bool should_binlog_drop_if_temp_flag{false};
public:
/**
Does this table have any columns that can be updated using partial update
in the current row?
@return whether any columns in the current row can be updated using partial
update
*/
bool has_binary_diff_columns() const;
/**
Get the list of binary diffs that have been collected for a given column in
the current row, or `nullptr` if partial update cannot be used for that
column.
@param field the column to get binary diffs for
@return the list of binary diffs for the column, or `nullptr` if the column
cannot be updated using partial update
*/
const Binary_diff_vector *get_binary_diffs(const Field *field) const;
/**
Mark a given column as one that can potentially be updated using
partial update during execution of an update statement.
Whether it is actually updated using partial update, is not
determined until execution time, since that depends both on the
data that is in the column and the new data that is written to the
column.
This function should be called during preparation of an update
statement.
@param field a column which is eligible for partial update
@retval false on success
@retval true on out-of-memory
*/
bool mark_column_for_partial_update(const Field *field);
/**
Has this column been marked for partial update?
Note that this only tells if the column satisfies the syntactical
requirements for being partially updated. Use #is_binary_diff_enabled() or
#is_logical_diff_enabled() instead to see if partial update should be used
on the column.
@param field the column to check
@return whether the column has been marked for partial update
*/
bool is_marked_for_partial_update(const Field *field) const;
/**
Does this table have any columns that were marked with
#mark_column_for_partial_update()?
Note that this only tells if any of the columns satisfy the syntactical
requirements for being partially updated. Use
#has_binary_diff_columns(), #is_binary_diff_enabled() or
#is_logical_diff_enabled() instead to see if partial update should be used
on a column.
*/
bool has_columns_marked_for_partial_update() const;
/**
Enable partial update of JSON columns in this table. It is only
enabled for the columns that have previously been marked for
partial update using #mark_column_for_partial_update().
@param logical_diffs should logical JSON diffs be collected in addition
to the physical binary diffs?
This function should be called once per statement execution, when
the update statement is optimized.
@retval false on success
@retval true on out-of-memory
*/
bool setup_partial_update(bool logical_diffs);
/**
@see setup_partial_update(bool)
This is a wrapper that auto-computes the value of the parameter
logical_diffs.
@retval false on success
@retval true on out-of-memory
*/
bool setup_partial_update();
/**
Add a binary diff for a column that is updated using partial update.
@param field the column that is being updated
@param offset the offset of the changed portion
@param length the length of the changed portion
@retval false on success
@retval true on out-of-memory
*/
bool add_binary_diff(const Field *field, size_t offset, size_t length);
/**
Clear the diffs that have been collected for partial update of
JSON columns, and re-enable partial update for any columns where
partial update was temporarily disabled for the current row.
Should be called between each row that is updated.
*/
void clear_partial_update_diffs();
/**
Clean up state used for partial update of JSON columns.
This function should be called at the end of each statement
execution.
*/
void cleanup_partial_update();
/**
Temporarily disable collection of binary diffs for a column in the current
row.
This function is called during execution to disable partial update of a
column that was previously marked as eligible for partial update with
#mark_column_for_partial_update() during preparation.
Partial update of this column will be re-enabled when we go to the next
row.
@param field the column to stop collecting binary diffs for
*/
void disable_binary_diffs_for_current_row(const Field *field);
/**
Temporarily disable collection of Json_diff objects describing the
logical changes of a JSON column in the current row.
Collection of logical JSON diffs is re-enabled when we go to the next row.
@param field the column to stop collecting logical JSON diffs for
*/
void disable_logical_diffs_for_current_row(const Field *field) const;
/**
Get a buffer that can be used to hold the partially updated column value
while performing partial update.
*/
String *get_partial_update_buffer();
/**
Add a logical JSON diff describing a logical change to a JSON column in
partial update.
@param field the column that is updated
@param path the JSON path that is changed
@param operation the operation to perform
@param new_value the new value in the path
@throws std::bad_alloc if memory cannot be allocated
*/
void add_logical_diff(const Field_json *field, const Json_seekable_path &path,
enum_json_diff_operation operation,
const Json_wrapper *new_value);
/**
Get the list of JSON diffs that have been collected for a given column in
the current row, or `nullptr` if partial update cannot be used for that
column.
@param field the column to get JSON diffs for
@return the list of JSON diffs for the column, or `nullptr` if the column
cannot be updated using partial update
*/
const Json_diff_vector *get_logical_diffs(const Field_json *field) const;
/**
Is partial update using binary diffs enabled on this JSON column?
@param field the column to check
@return whether the column can be updated with binary diffs
*/
bool is_binary_diff_enabled(const Field *field) const;
/**
Is partial update using logical diffs enabled on this JSON column?
@param field the column to check
@return whether the column can be updated with JSON diffs
*/
bool is_logical_diff_enabled(const Field *field) const;
/**
Virtual fields of type BLOB have a flag m_keep_old_value. This flag is set
to false for all such fields in this table.
*/
void blobs_need_not_keep_old_value();
/**
Set the variable should_binlog_drop_if_temp_flag, so that
the logging of temporary tables can be decided.
@param should_binlog the value to set flag should_binlog_drop_if_temp_flag
*/
void set_binlog_drop_if_temp(bool should_binlog);
/**
@return whether should_binlog_drop_if_temp_flag flag is
set or not
*/
bool should_binlog_drop_if_temp(void) const;
/**
Find the histogram for the given field index.
@note If this is called on a TABLE object that belongs to a secondary
engine, it will take a round-trip through the handler in order to obtain the
histogram from the TABLE object associated with the primary engine. This is
done to avoid storing histograms on both the primary and secondary
TABLE_SHARE.
@param field_index The index of the field we want to find a histogram for.
@retval nullptr if no histogram is found.
@retval Pointer to a histogram if one is found.
*/
const histograms::Histogram *find_histogram(uint field_index) const;
};
static inline void empty_record(TABLE *table) {
restore_record(table, s->default_values);
if (table->s->null_bytes > 0)
memset(table->null_flags, 255, table->s->null_bytes);
}
#define MY_I_S_MAYBE_NULL 1
#define MY_I_S_UNSIGNED 2
struct ST_FIELD_INFO {
/**
This is used as column name.
*/
const char *field_name;
/**
For string-type columns, this is the maximum number of
characters. Otherwise, it is the 'display-length' for the column.
For the data type MYSQL_TYPE_DATETIME this field specifies the
number of digits in the fractional part of time value.
*/
uint field_length;
/**
This denotes data type for the column. For the most part, there seems to
be one entry in the enum for each SQL data type, although there seem to
be a number of additional entries in the enum.
*/
enum_field_types field_type;
int value;
/**
This is used to set column attributes. By default, columns are @c NOT
@c NULL and @c SIGNED, and you can deviate from the default
by setting the appropriate flags. You can use either one of the flags
@c MY_I_S_MAYBE_NULL and @c MY_I_S_UNSIGNED or
combine them using the bitwise or operator @c |. Both flags are
defined in table.h.
*/
uint field_flags; // Field attributes (maybe_null, signed, unsigned etc.)
const char *old_name;
uint open_method; // Not used
};
struct ST_SCHEMA_TABLE {
const char *table_name;
ST_FIELD_INFO *fields_info;
/* Fill table with data */
int (*fill_table)(THD *thd, Table_ref *tables, Item *cond);
/* Handle fields for old SHOW */
int (*old_format)(THD *thd, ST_SCHEMA_TABLE *schema_table);
int (*process_table)(THD *thd, Table_ref *tables, TABLE *table, bool res,
LEX_CSTRING db_name, LEX_CSTRING table_name);
bool hidden;
};
/**
Strategy for how to process a view or derived table (merge or materialization)
*/
enum enum_view_algorithm {
VIEW_ALGORITHM_UNDEFINED = 0,
VIEW_ALGORITHM_TEMPTABLE = 1,
VIEW_ALGORITHM_MERGE = 2
};
#define VIEW_SUID_INVOKER 0
#define VIEW_SUID_DEFINER 1
#define VIEW_SUID_DEFAULT 2
/* view WITH CHECK OPTION parameter options */
#define VIEW_CHECK_NONE 0
#define VIEW_CHECK_LOCAL 1
#define VIEW_CHECK_CASCADED 2
/* result of view WITH CHECK OPTION parameter check */
#define VIEW_CHECK_OK 0
#define VIEW_CHECK_ERROR 1
#define VIEW_CHECK_SKIP 2
/** The threshold size a blob field buffer before it is freed */
#define MAX_TDC_BLOB_SIZE 65536
/**
Struct that describes an expression selected from a derived table or view.
*/
struct Field_translator {
/**
Points to an item that represents the expression.
If the item is determined to be unused, the pointer is set to NULL.
*/
Item *item;
/// Name of selected expression
const char *name;
};
/*
Column reference of a NATURAL/USING join. Since column references in
joins can be both from views and stored tables, may point to either a
Field (for tables), or a Field_translator (for views).
*/
class Natural_join_column {
public:
Field_translator *view_field; /* Column reference of merge view. */
Item_field *table_field; /* Column reference of table or temp view. */
Table_ref *table_ref; /* Original base table/view reference. */
/*
True if a common join column of two NATURAL/USING join operands. Notice
that when we have a hierarchy of nested NATURAL/USING joins, a column can
be common at some level of nesting but it may not be common at higher
levels of nesting. Thus this flag may change depending on at which level
we are looking at some column.
*/
bool is_common;
public:
Natural_join_column(Field_translator *field_param, Table_ref *tab);
Natural_join_column(Item_field *field_param, Table_ref *tab);
const char *name();
Item *create_item(THD *thd);
Field *field();
const char *table_name();
const char *db_name();
GRANT_INFO *grant();
};
/**
This is generic enum. It may be reused in the ACL statements
for clauses that can map to the values defined in this enum.
*/
enum class Lex_acl_attrib_udyn {
UNCHANGED, /* The clause is not specified */
DEFAULT, /* Default value of clause is specified */
YES, /* Value that maps to True is specified */
NO /* Value that maps to False is specified */
};
struct LEX_MFA {
LEX_CSTRING plugin;
LEX_CSTRING auth;
LEX_CSTRING generated_password;
LEX_CSTRING challenge_response;
LEX_CSTRING client_plugin;
uint nth_factor;
/*
The following flags are indicators for the SQL syntax used while
parsing CREATE/ALTER user. While other members are self-explanatory,
'uses_authentication_string_clause' signifies if the password is in
hash form (if the var was set to true) or not.
*/
bool uses_identified_by_clause;
bool uses_authentication_string_clause;
bool uses_identified_with_clause;
bool has_password_generator;
/* flag set during CREATE USER .. INITIAL AUTHENTICATION BY */
bool passwordless;
/* flag set during ALTER USER .. ADD nth FACTOR */
bool add_factor;
/* flag set during ALTER USER .. MODIFY nth FACTOR */
bool modify_factor;
/* flag set during ALTER USER .. DROP nth FACTOR */
bool drop_factor;
/*
flag used during authentication and to decide if server should
be in sandbox mode or not
*/
bool requires_registration;
/* flag set during ALTER USER .. nth FACTOR UNREGISTER */
bool unregister;
/* flag set during ALTER USER .. INITIATE REGISTRATION */
bool init_registration;
/* flag set during ALTER USER .. FINISH REGISTRATION */
bool finish_registration;
LEX_MFA() { reset(); }
void reset() {
plugin = EMPTY_CSTR;
auth = NULL_CSTR;
generated_password = NULL_CSTR;
challenge_response = NULL_CSTR;
client_plugin = NULL_CSTR;
nth_factor = 1;
uses_identified_by_clause = false;
uses_authentication_string_clause = false;
uses_identified_with_clause = false;
has_password_generator = false;
passwordless = false;
add_factor = false;
drop_factor = false;
modify_factor = false;
requires_registration = false;
unregister = false;
init_registration = false;
finish_registration = false;
}
void copy(LEX_MFA *m, MEM_ROOT *alloc);
};
/*
This structure holds the specifications relating to
ALTER user ... PASSWORD EXPIRE ...
*/
struct LEX_ALTER {
bool update_password_expired_fields;
bool update_password_expired_column;
bool use_default_password_lifetime;
uint16 expire_after_days;
bool update_account_locked_column;
bool account_locked;
uint32 password_history_length;
bool use_default_password_history;
bool update_password_history;
uint32 password_reuse_interval;
bool use_default_password_reuse_interval;
bool update_password_reuse_interval;
uint failed_login_attempts;
bool update_failed_login_attempts;
int password_lock_time;
bool update_password_lock_time;
/* Holds the specification of 'PASSWORD REQUIRE CURRENT' clause. */
Lex_acl_attrib_udyn update_password_require_current;
void cleanup() {
update_password_expired_fields = false;
update_password_expired_column = false;
use_default_password_lifetime = true;
expire_after_days = 0;
update_account_locked_column = false;
account_locked = false;
use_default_password_history = true;
update_password_history = false;
use_default_password_reuse_interval = true;
update_password_reuse_interval = false;
update_password_require_current = Lex_acl_attrib_udyn::UNCHANGED;
password_history_length = 0;
password_reuse_interval = 0;
update_password_lock_time = false;
update_failed_login_attempts = false;
failed_login_attempts = 0;
password_lock_time = 0;
}
};
/*
This structure holds the specifications related to
mysql user and the associated auth details.
*/
struct LEX_USER {
LEX_CSTRING user;
LEX_CSTRING host;
LEX_CSTRING current_auth;
bool uses_replace_clause;
bool retain_current_password;
bool discard_old_password;
LEX_ALTER alter_status;
/* restrict MFA methods to atmost 3 authentication plugins */
LEX_MFA first_factor_auth_info;
List<LEX_MFA> mfa_list;
bool with_initial_auth;
void init() {
user = NULL_CSTR;
host = NULL_CSTR;
current_auth = NULL_CSTR;
uses_replace_clause = false;
retain_current_password = false;
discard_old_password = false;
alter_status.account_locked = false;
alter_status.expire_after_days = 0;
alter_status.update_account_locked_column = false;
alter_status.update_password_expired_column = false;
alter_status.update_password_expired_fields = false;
alter_status.use_default_password_lifetime = true;
alter_status.use_default_password_history = true;
alter_status.update_password_require_current =
Lex_acl_attrib_udyn::UNCHANGED;
alter_status.password_history_length = 0;
alter_status.password_reuse_interval = 0;
alter_status.failed_login_attempts = 0;
alter_status.password_lock_time = 0;
alter_status.update_failed_login_attempts = false;
alter_status.update_password_lock_time = false;
first_factor_auth_info.reset();
mfa_list.clear();
with_initial_auth = false;
}
LEX_USER() { init(); }
bool add_mfa_identifications(LEX_MFA *factor2, LEX_MFA *factor3 = nullptr);
/*
Allocates the memory in the THD mem pool and initialize the members of
this struct. It is preferable to use this method to create a LEX_USER
rather allocating the memory in the THD and initializing the members
explicitly.
*/
static LEX_USER *alloc(THD *thd);
static LEX_USER *alloc(THD *thd, LEX_STRING *user, LEX_STRING *host);
/*
Initialize the members of this struct. It is preferable to use this method
to initialize a LEX_USER rather initializing the members explicitly.
*/
static LEX_USER *init(LEX_USER *to_init, THD *thd, LEX_STRING *user,
LEX_STRING *host);
};
/**
Derive type of metadata lock to be requested for table used by a DML
statement from the type of THR_LOCK lock requested for this table.
*/
inline enum enum_mdl_type mdl_type_for_dml(enum thr_lock_type lock_type) {
return lock_type >= TL_WRITE_ALLOW_WRITE
? (lock_type == TL_WRITE_LOW_PRIORITY ? MDL_SHARED_WRITE_LOW_PRIO
: MDL_SHARED_WRITE)
: MDL_SHARED_READ;
}
/**
Type of table which can be open for an element of table list.
*/
enum enum_open_type {
OT_TEMPORARY_OR_BASE = 0,
OT_TEMPORARY_ONLY,
OT_BASE_ONLY
};
/**
This structure is used to keep info about possible key for the result table
of a derived table/view.
The 'referenced_by' is the table map of tables to which this possible
key corresponds.
The 'used_field' is a map of fields of which this key consists of.
See also the comment for the Table_ref::update_derived_keys function.
*/
class Derived_key {
public:
table_map referenced_by;
Field_map used_fields;
uint key_part_count{0};
};
class Table_function;
/*
Table reference in the FROM clause.
These table references can be of several types that correspond to
different SQL elements. Below we list all types of TABLE_LISTs with
the necessary conditions to determine when a Table_ref instance
belongs to a certain type.
1) table (Table_ref::view == NULL)
- base table
(Table_ref::derived == NULL)
- subquery - Table_ref::table is a temp table
(Table_ref::derived != NULL)
- information schema table
(Table_ref::schema_table != NULL)
NOTICE: for schema tables Table_ref::field_translation may be != NULL
2) view (Table_ref::view != NULL)
- merge (Table_ref::effective_algorithm == VIEW_ALGORITHM_MERGE)
also (Table_ref::field_translation != NULL)
- temptable(Table_ref::effective_algorithm == VIEW_ALGORITHM_TEMPTABLE)
also (Table_ref::field_translation == NULL)
3) nested table reference (Table_ref::nested_join != NULL)
- table sequence - e.g. (t1, t2, t3)
TODO: how to distinguish from a JOIN?
- general JOIN
TODO: how to distinguish from a table sequence?
- NATURAL JOIN
(Table_ref::natural_join != NULL)
- JOIN ... USING
(Table_ref::join_using_fields != NULL)
- semi-join
;
*/
class Table_ref {
public:
Table_ref() = default;
/**
Only to be used by legacy code that temporarily needs a Table_ref,
more specifically: Query_result_create::binlog_show_create_table().
*/
explicit Table_ref(TABLE *table_arg) : table(table_arg) {}
/// Constructor that can be used when the strings are null terminated.
Table_ref(const char *db_name, const char *table_name,
enum thr_lock_type lock_type)
: Table_ref(db_name, strlen(db_name), table_name, strlen(table_name),
table_name, lock_type) {}
/**
Creates a Table_ref object with pre-allocated strings for database,
table and alias.
*/
Table_ref(TABLE *table_arg, const char *db_name_arg, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
const char *alias_arg, enum thr_lock_type lock_type_arg)
: db(db_name_arg),
table_name(table_name_arg),
alias(alias_arg),
m_map(1),
table(table_arg),
m_lock_descriptor{lock_type_arg},
db_length(db_length_arg),
table_name_length(table_name_length_arg) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
}
/// Constructor that can be used when the strings are null terminated.
Table_ref(const char *db_name, const char *table_name, const char *alias,
enum thr_lock_type lock_type)
: Table_ref(db_name, strlen(db_name), table_name, strlen(table_name),
alias, lock_type) {}
/**
This constructor can be used when a Table_ref is needed for an
existing temporary table. These typically have very long table names, since
it is a fully qualified path. For this reason, the table is set to the
alias. The database name is left blank. The lock descriptor is set to
TL_READ.
*/
Table_ref(TABLE *table_arg, const char *alias_arg)
: db(""),
table_name(alias_arg),
alias(alias_arg),
m_map(1),
table(table_arg),
m_lock_descriptor{TL_READ},
db_length(0),
table_name_length(strlen(alias_arg)) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
}
/**
Sets an explicit enum_mdl_type value, without initializing
m_lock_descriptor.
*/
Table_ref(TABLE *table_arg, const char *alias_arg, enum_mdl_type mdl_type)
: db(table_arg->s->db.str),
table_name(table_arg->s->table_name.str),
alias(alias_arg),
m_map(1),
table(table_arg),
db_length(table_arg->s->db.length),
table_name_length(table_arg->s->table_name.length) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name, mdl_type,
MDL_TRANSACTION);
}
Table_ref(const char *db_name, const char *table_name_arg,
enum thr_lock_type lock_type_arg,
enum enum_mdl_type mdl_request_type)
: db(db_name),
table_name(table_name_arg),
alias(table_name_arg),
m_map(1),
m_lock_descriptor{lock_type_arg},
db_length(strlen(db_name)),
table_name_length(strlen(table_name_arg)) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
mdl_request.set_type(mdl_request_type);
}
Table_ref(const char *db_name, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
enum thr_lock_type lock_type_arg,
enum enum_mdl_type mdl_request_type)
: db(db_name),
table_name(table_name_arg),
alias(table_name_arg),
m_map(1),
m_lock_descriptor{lock_type_arg},
db_length(db_length_arg),
table_name_length(table_name_length_arg) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
mdl_request.set_type(mdl_request_type);
}
Table_ref(const char *db_name, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
enum thr_lock_type lock_type_arg)
: db(db_name),
table_name(table_name_arg),
alias(table_name_arg),
m_map(1),
m_lock_descriptor{lock_type_arg},
db_length(db_length_arg),
table_name_length(table_name_length_arg) {}
/**
Sets an explicit enum_mdl_type value, without initializing
m_lock_descriptor.
*/
Table_ref(const char *db_name, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
const char *alias_arg, enum enum_mdl_type mdl_request_type)
: db(db_name),
table_name(table_name_arg),
alias(alias_arg),
m_map(1),
db_length(db_length_arg),
table_name_length(table_name_length_arg) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
mdl_request.set_type(mdl_request_type);
}
Table_ref(const char *db_name, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
const char *alias_arg, enum thr_lock_type lock_type_arg,
enum enum_mdl_type mdl_request_type)
: db(db_name),
table_name(table_name_arg),
alias(alias_arg),
m_map(1),
m_lock_descriptor{lock_type_arg},
db_length(db_length_arg),
table_name_length(table_name_length_arg) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
mdl_request.set_type(mdl_request_type);
}
Table_ref(const char *db_name_arg, size_t db_length_arg,
const char *table_name_arg, size_t table_name_length_arg,
const char *alias_arg, enum thr_lock_type lock_type_arg)
: db(db_name_arg),
table_name(table_name_arg),
alias(alias_arg),
m_map(1),
m_lock_descriptor{lock_type_arg},
db_length(db_length_arg),
table_name_length(table_name_length_arg) {
MDL_REQUEST_INIT(&mdl_request, MDL_key::TABLE, db, table_name,
mdl_type_for_dml(m_lock_descriptor.type), MDL_TRANSACTION);
}
/// Create a Table_ref object representing a nested join
static Table_ref *new_nested_join(MEM_ROOT *allocator, const char *alias,
Table_ref *embedding,
mem_root_deque<Table_ref *> *belongs_to,
Query_block *select);
Item **join_cond_ref() { return &m_join_cond; }
Item *join_cond() const { return m_join_cond; }
void set_join_cond(Item *val) {
// If optimization has started, it's too late to change m_join_cond.
assert(m_join_cond_optim == nullptr || m_join_cond_optim == (Item *)1);
m_join_cond = val;
}
Item *join_cond_optim() const { return m_join_cond_optim; }
void set_join_cond_optim(Item *cond) {
/*
Either we are setting to "empty", or there must pre-exist a
permanent condition.
*/
assert(cond == nullptr || cond == (Item *)1 || m_join_cond != nullptr);
m_join_cond_optim = cond;
}
Item **join_cond_optim_ref() { return &m_join_cond_optim; }
/// @returns true if semi-join nest
bool is_sj_nest() const { return m_is_sj_or_aj_nest && !m_join_cond; }
/// @returns true if anti-join nest
bool is_aj_nest() const { return m_is_sj_or_aj_nest && m_join_cond; }
/// @returns true if anti/semi-join nest
bool is_sj_or_aj_nest() const { return m_is_sj_or_aj_nest; }
/// Makes the next a semi/antijoin nest
void set_sj_or_aj_nest() {
assert(!m_is_sj_or_aj_nest);
m_is_sj_or_aj_nest = true;
}
/// Merge tables from a query block into a nested join structure
bool merge_underlying_tables(Query_block *select);
/// Reset table
void reset();
/// Evaluate the check option of a view
int view_check_option(THD *thd) const;
/// Produce a textual identification of this object
void print(const THD *thd, String *str, enum_query_type query_type) const;
/// Check which single table inside a view that matches a table map
bool check_single_table(Table_ref **table_ref, table_map map);
/// Allocate a buffer for inserted column values
bool set_insert_values(MEM_ROOT *mem_root);
Table_ref *first_leaf_for_name_resolution();
/**
Retrieve the last (right-most) leaf in a nested join tree with
respect to name resolution.
Given that 'this' is a nested table reference, recursively walk
down the right-most children of 'this' until we reach a leaf
table reference with respect to name resolution.
The right-most child of a nested table reference is the first
element in the list of children because the children are inserted
in reverse order.
@return
- If 'this' is a nested table reference - the right-most child
of the tree rooted in 'this',
- else - 'this'
*/
Table_ref *last_leaf_for_name_resolution();
bool is_leaf_for_name_resolution() const;
/// Return the outermost view this table belongs to, or itself
inline const Table_ref *top_table() const {
return belong_to_view ? belong_to_view : this;
}
inline Table_ref *top_table() {
return const_cast<Table_ref *>(
const_cast<const Table_ref *>(this)->top_table());
}
/// Prepare check option for a view
bool prepare_check_option(THD *thd, bool is_cascaded = false);
/// Merge WHERE condition of view or derived table into outer query
bool merge_where(THD *thd);
/// Prepare replace filter for a view (used for REPLACE command)
bool prepare_replace_filter(THD *thd);
/// Return true if this represents a named view
bool is_view() const { return view != nullptr; }
/// Return true if this represents a derived table (an unnamed view)
bool is_derived() const { return derived != nullptr && view == nullptr; }
/// Return true if this represents a named view or a derived table
bool is_view_or_derived() const { return derived != nullptr; }
/// Return true if this represents a table function
bool is_table_function() const { return table_function != nullptr; }
/**
@returns true if this is a recursive reference inside the definition of a
recursive CTE.
@note that it starts its existence as a dummy derived table, until the
end of resolution when it's not a derived table anymore, just a reference
to the materialized temporary table. Whereas a non-recursive
reference to the recursive CTE is a derived table.
*/
bool is_recursive_reference() const { return m_is_recursive_reference; }
/// @returns true if this is a base table (permanent or temporary)
bool is_base_table() const {
return !(is_view_or_derived() || is_table_function() ||
is_recursive_reference());
}
/**
@see is_recursive_reference().
@returns true if error
*/
bool set_recursive_reference();
/**
@returns true for a table that represents an optimizer internal table,
is a derived table, a recursive reference, a table function.
Internal tables are only visible inside a query expression, and is hence
not visible in any schema, or need any kind of privilege checking.
*/
bool is_internal() const {
return is_derived() || is_recursive_reference() || is_table_function();
}
/**
@returns true for a table that is a placeholder, ie a derived table,
a view, a recursive reference, a table function or a schema table.
A table is also considered to be a placeholder if it does not have a
TABLE object for some other reason.
*/
bool is_placeholder() const {
return is_view_or_derived() || is_recursive_reference() ||
is_table_function() || schema_table || table == nullptr;
}
/// Return true if view or derived table and can be merged
bool is_mergeable() const;
/**
Checks if this is a table that contains zero rows or one row, and that can
be materialized during optimization.
@returns true if materializable table contains one or zero rows, and
materialization during optimization is permitted
Returning true, if the hypergraph optimizer is not active, implies that the
table is materialized during optimization, so it need not be optimized
during execution. The hypergraph optimizer does not care about const tables,
so such tables are not executed during optimization time when it is active.
*/
bool materializable_is_const() const;
/// Return true if this is a derived table or view that is merged
bool is_merged() const { return effective_algorithm == VIEW_ALGORITHM_MERGE; }
/// Set table to be merged
void set_merged() {
assert(effective_algorithm == VIEW_ALGORITHM_UNDEFINED);
effective_algorithm = VIEW_ALGORITHM_MERGE;
}
/// Return true if this is a materializable derived table/view
bool uses_materialization() const {
return effective_algorithm == VIEW_ALGORITHM_TEMPTABLE;
}
/// Set table to be materialized
void set_uses_materialization() {
// @todo We should do this only once, but currently we cannot:
// assert(effective_algorithm == VIEW_ALGORITHM_UNDEFINED);
assert(effective_algorithm != VIEW_ALGORITHM_MERGE);
effective_algorithm = VIEW_ALGORITHM_TEMPTABLE;
}
/// Return true if table is updatable
bool is_updatable() const { return m_updatable; }
/// Set table as updatable. (per default, a table is non-updatable)
void set_updatable() { m_updatable = true; }
/// Return true if table is insertable-into
bool is_insertable() const { return m_insertable; }
/// Set table as insertable-into. (per default, a table is not insertable)
void set_insertable() { m_insertable = true; }
/// Return true if table is being updated
bool is_updated() const { return m_updated; }
/// Set table and all referencing views as being updated
void set_updated() {
for (Table_ref *tr = this; tr != nullptr; tr = tr->referencing_view)
tr->m_updated = true;
}
/// Return true if table is being inserted into
bool is_inserted() const { return m_inserted; }
/// Set table and all referencing views as being inserted into
void set_inserted() {
for (Table_ref *tr = this; tr != nullptr; tr = tr->referencing_view)
tr->m_inserted = true;
}
/// Return true if table is being deleted from
bool is_deleted() const { return m_deleted; }
/// Set table and all referencing views as being deleted from
void set_deleted() {
for (Table_ref *tr = this; tr != nullptr; tr = tr->referencing_view)
tr->m_deleted = true;
}
/// Set table as full-text search (default is not fulltext searched)
void set_fulltext_searched() { m_fulltext_searched = true; }
/// Returns true if a MATCH function references this table.
bool is_fulltext_searched() const { return m_fulltext_searched; }
/// Is this table only available in an external storage engine?
bool is_external() const;
/**
Set table as readonly, ie it is neither updatable, insertable nor
deletable during this statement.
*/
void set_readonly() {
m_updatable = false;
m_insertable = false;
}
/**
Return true if this is a view or derived table that is defined over
more than one base table, and false otherwise.
*/
bool is_multiple_tables() const {
if (is_view_or_derived()) {
assert(is_merged()); // Cannot be a materialized view
return leaf_tables_count() > 1;
} else {
assert(nested_join == nullptr); // Must be a base table
return false;
}
}
/// Return no. of base tables a merged view or derived table is defined over.
uint leaf_tables_count() const;
/// Return first leaf table of a base table or a view/derived table
Table_ref *first_leaf_table() {
Table_ref *tr = this;
while (tr->merge_underlying_list) tr = tr->merge_underlying_list;
return tr;
}
/// Return any leaf table that is not an inner table of an outer join
/// @todo WL#6570 with prepare-once, replace with first_leaf_table()
/// when WL#6059 is merged in (it really converts RIGHT JOIN to
/// LEFT JOIN so the first leaf is part of a LEFT JOIN,
/// guaranteed).
Table_ref *any_outer_leaf_table() {
Table_ref *tr = this;
while (tr->merge_underlying_list) {
tr = tr->merge_underlying_list;
/*
"while" is used, however, an "if" might be sufficient since there is
no more than one inner table in a join nest (with outer_join true).
*/
while (tr->outer_join) tr = tr->next_local;
}
return tr;
}
/**
Set the LEX object of a view (will also define this as a view).
@note: The value 1 is used to indicate a view but without a valid
query object. Use only if the LEX object is not going to
be used in later processing.
*/
void set_view_query(LEX *lex) { view = lex; }
/// Return the valid LEX object for a view.
LEX *view_query() const {
assert(view != nullptr && view != (LEX *)1);
return view;
}
/**
Set the query expression of a derived table or view.
(Will also define this as a derived table, unless it is a named view.)
*/
void set_derived_query_expression(Query_expression *query_expr) {
derived = query_expr;
}
/// Return the query expression of a derived table or view.
Query_expression *derived_query_expression() const {
assert(derived);
return derived;
}
/// Resolve a derived table or view reference
bool resolve_derived(THD *thd, bool apply_semijoin);
/// Optimize the query expression representing a derived table/view
bool optimize_derived(THD *thd);
/// Create result table for a materialized derived table/view
bool create_materialized_table(THD *thd);
/// Materialize derived table
bool materialize_derived(THD *thd);
/// Check if we can push outer where condition to this derived table
bool can_push_condition_to_derived(THD *thd);
/// Return the number of hidden fields added for the temporary table
/// created for this derived table.
uint get_hidden_field_count_for_derived() const;
/// Prepare security context for a view
bool prepare_security(THD *thd);
Security_context *find_view_security_context(THD *thd);
bool prepare_view_security_context(THD *thd);
/**
Compiles the tagged hints list and fills up TABLE::keys_in_use_for_query,
TABLE::keys_in_use_for_group_by, TABLE::keys_in_use_for_order_by,
TABLE::force_index and TABLE::covering_keys.
*/
bool process_index_hints(const THD *thd, TABLE *table);
/**
Compare the version of metadata from the previous execution
(if any) with values obtained from the current table
definition cache element.
@sa check_and_update_table_version()
*/
bool is_table_ref_id_equal(TABLE_SHARE *s) const {
return (m_table_ref_type == s->get_table_ref_type() &&
m_table_ref_version == s->get_table_ref_version());
}
/**
Record the value of metadata version of the corresponding
table definition cache element in this parse tree node.
@sa check_and_update_table_version()
*/
void set_table_ref_id(TABLE_SHARE *s) {
set_table_ref_id(s->get_table_ref_type(), s->get_table_ref_version());
}
void set_table_ref_id(enum_table_ref_type table_ref_type_arg,
ulonglong table_ref_version_arg) {
m_table_ref_type = table_ref_type_arg;
m_table_ref_version = table_ref_version_arg;
}
/**
If a derived table, returns query block id of first underlying query block.
Zero if not derived.
*/
uint query_block_id() const;
/**
This is for showing in EXPLAIN.
If a derived table, returns query block id of first underlying query block
of first materialized Table_ref instance. Zero if not derived.
*/
uint query_block_id_for_explain() const;
/**
@brief Returns the name of the database that the referenced table belongs
to.
*/
const char *get_db_name() const { return db; }
/**
@brief Returns the name of the table that this Table_ref represents.
@details The unqualified table name or view name for a table or view,
respectively.
*/
const char *get_table_name() const { return table_name; }
int fetch_number_of_rows(
ha_rows fallback_estimate = PLACEHOLDER_TABLE_ROW_ESTIMATE);
bool update_derived_keys(THD *, Field *, Item **, uint, bool *);
bool generate_keys();
/// Setup a derived table to use materialization
bool setup_materialized_derived(THD *thd);
bool setup_materialized_derived_tmp_table(THD *thd);
/// Setup a table function to use materialization
bool setup_table_function(THD *thd);
bool create_field_translation(THD *thd);
/**
@brief Returns the outer join nest that this Table_ref belongs to, if
any.
@details There are two kinds of join nests, outer-join nests and semi-join
nests. This function returns non-NULL in the following cases:
@li 1. If this table/nest is embedded in a nest and this nest IS NOT a
semi-join nest. (In other words, it is an outer-join nest.)
@li 2. If this table/nest is embedded in a nest and this nest IS a
semi-join nest, but this semi-join nest is embedded in another
nest. (This other nest will be an outer-join nest, since all inner
joined nested semi-join nests have been merged in
@c simplify_joins() ).
Note: This function assumes that @c simplify_joins() has been performed.
Before that, join nests will be present for all types of join.
@return outer join nest, or NULL if none.
*/
Table_ref *outer_join_nest() const {
if (!embedding) return nullptr;
if (embedding->is_sj_nest()) return embedding->embedding;
return embedding;
}
/**
Return true if this table is an inner table of some outer join.
Examine all the embedding join nests of the table.
@note This function works also before redundant join nests have been
eliminated.
@return true if table is an inner table of some outer join, false otherwise.
*/
bool is_inner_table_of_outer_join() const {
if (outer_join) return true;
for (Table_ref *emb = embedding; emb; emb = emb->embedding) {
if (emb->outer_join) return true;
}
return false;
}
/**
Return the base table entry of an updatable table.
In DELETE and UPDATE, a view used as a target table must be mergeable,
updatable and defined over a single table.
*/
const Table_ref *updatable_base_table() const {
const Table_ref *tbl = this;
assert(tbl->is_updatable() && !tbl->is_multiple_tables());
while (tbl->is_view_or_derived()) {
tbl = tbl->merge_underlying_list;
assert(tbl->is_updatable() && !tbl->is_multiple_tables());
}
return tbl;
}
Table_ref *updatable_base_table() {
return const_cast<Table_ref *>(
static_cast<const Table_ref *>(this)->updatable_base_table());
}
/**
Mark that there is a NATURAL JOIN or JOIN ... USING between two tables.
This function marks that table b should be joined with a either via
a NATURAL JOIN or via JOIN ... USING. Both join types are special
cases of each other, so we treat them together. The function
setup_conds() creates a list of equal condition between all fields
of the same name for NATURAL JOIN or the fields in
Table_ref::join_using_fields for JOIN ... USING.
The list of equality conditions is stored
either in b->join_cond(), or in JOIN::conds, depending on whether there
was an outer join.
EXAMPLE
@verbatim
SELECT * FROM t1 NATURAL LEFT JOIN t2
<=>
SELECT * FROM t1 LEFT JOIN t2 ON (t1.i=t2.i and t1.j=t2.j ... )
SELECT * FROM t1 NATURAL JOIN t2 WHERE <some_cond>
<=>
SELECT * FROM t1, t2 WHERE (t1.i=t2.i and t1.j=t2.j and <some_cond>)
SELECT * FROM t1 JOIN t2 USING(j) WHERE <some_cond>
<=>
SELECT * FROM t1, t2 WHERE (t1.j=t2.j and <some_cond>)
@endverbatim
@param b Right join argument.
*/
void add_join_natural(Table_ref *b) { b->natural_join = this; }
/**
Set granted privileges for a table.
Can be used when generating temporary tables that are also used in
resolver process, such as when generating a UNION table
@param privilege Privileges granted for this table.
*/
void set_privileges(Access_bitmask privilege) {
grant.privilege |= privilege;
}
bool save_properties();
void restore_properties();
/*
List of tables local to a subquery or the top-level SELECT (used by
SQL_I_List). Considers views as leaves (unlike 'next_leaf' below).
Created at parse time in Query_block::add_table_to_list() ->
table_list.link_in_list().
*/
Table_ref *next_local{nullptr};
/* link in a global list of all queries tables */
Table_ref *next_global{nullptr}, **prev_global{nullptr};
const char *db{nullptr}, *table_name{nullptr}, *alias{nullptr};
/*
Target tablespace name: When creating or altering tables, this
member points to the tablespace_name in the HA_CREATE_INFO struct.
*/
LEX_CSTRING target_tablespace_name{nullptr, 0};
char *option{nullptr}; /* Used by cache index */
/** Table level optimizer hints for this table. */
Opt_hints_table *opt_hints_table{nullptr};
/* Hints for query block of this table. */
Opt_hints_qb *opt_hints_qb{nullptr};
void set_lock(const Lock_descriptor &descriptor) {
m_lock_descriptor = descriptor;
}
const Lock_descriptor &lock_descriptor() const { return m_lock_descriptor; }
bool is_derived_unfinished_materialization() const;
private:
/**
The members below must be kept aligned so that (1 << m_tableno) == m_map.
A table that takes part in a join operation must be assigned a unique
table number.
*/
uint m_tableno{0}; ///< Table number within query block
table_map m_map{0}; ///< Table map, derived from m_tableno
/**
If this table or join nest is the Y in "X [LEFT] JOIN Y ON C", this
member points to C. May also be generated from JOIN ... USING clause.
It may be modified only by permanent transformations (permanent = done
once for all executions of a prepared statement).
*/
Item *m_join_cond{nullptr};
bool m_is_sj_or_aj_nest{false};
public:
/*
(Valid only for semi-join nests) Bitmap of tables that are within the
semi-join (this is different from bitmap of all nest's children because
tables that were pulled out of the semi-join nest remain listed as
nest's children).
*/
table_map sj_inner_tables{0};
/*
During parsing - left operand of NATURAL/USING join where 'this' is
the right operand. After parsing (this->natural_join == this) iff
'this' represents a NATURAL or USING join operation. Thus after
parsing 'this' is a NATURAL/USING join iff (natural_join != NULL).
*/
Table_ref *natural_join{nullptr};
/*
True if 'this' represents a nested join that is a NATURAL JOIN.
For one of the operands of 'this', the member 'natural_join' points
to the other operand of 'this'.
*/
bool is_natural_join{false};
/* Field names in a USING clause for JOIN ... USING. */
List<String> *join_using_fields{nullptr};
/*
Explicitly store the result columns of either a NATURAL/USING join or
an operand of such a join.
*/
List<Natural_join_column> *join_columns{nullptr};
/* true if join_columns contains all columns of this table reference. */
bool is_join_columns_complete{false};
/*
List of nodes in a nested join tree, that should be considered as
leaves with respect to name resolution. The leaves are: views,
top-most nodes representing NATURAL/USING joins, subqueries, and
base tables. All of these Table_ref instances contain a
materialized list of columns. The list is local to a subquery.
*/
Table_ref *next_name_resolution_table{nullptr};
/* Index names in a "... JOIN ... USE/IGNORE INDEX ..." clause. */
List<Index_hint> *index_hints{nullptr};
TABLE *table{nullptr}; /* opened table */
mysql::binlog::event::Table_id
table_id{}; /* table id (from binlog) for opened table */
/*
Query_result for derived table to pass it from table creation to table
filling procedure
*/
Query_result_union *derived_result{nullptr};
/*
Reference from aux_tables to local list entry of main select of
multi-delete statement:
delete t1 from t2,t1 where t1.a<'B' and t2.b=t1.b;
here it will be reference of first occurrence of t1 to second (as you
can see this lists can't be merged)
*/
Table_ref *correspondent_table{nullptr};
/*
Holds the function used as the table function
*/
Table_function *table_function{nullptr};
/**
If we've previously made an access path for “derived”, it is cached here.
This is useful if we need to plan the query block twice (the hypergraph
optimizer can do so, with and without in2exists predicates), both saving
work and avoiding issues when we try to throw away the old items_to_copy
for a new (identical) one.
*/
AccessPath *access_path_for_derived{nullptr};
Item *sampling_percentage{nullptr};
private:
/// Sampling information.
tablesample_type sampling_type{
tablesample_type::UNSPECIFIED_TABLESAMPLE_TYPE};
double sampling_percentage_val{0};
/**
This field is set to non-null for derived tables and views. It points
to the Query_expression representing the derived table/view.
E.g. for a query
@verbatim SELECT * FROM (SELECT a FROM t1) b @endverbatim
*/
Query_expression *derived{nullptr}; /* Query_expression of derived table */
/// If non-NULL, the CTE which this table is derived from.
Common_table_expr *m_common_table_expr{nullptr};
/**
If the user has specified column names with the syntaxes "table name
parenthesis column names":
WITH qn(column names) AS (select...)
or
FROM (select...) dt(column names)
or
CREATE VIEW v(column_names) AS ...
then this points to the list of column names. NULL otherwise.
*/
const Create_col_name_list *m_derived_column_names{nullptr};
public:
ST_SCHEMA_TABLE *schema_table{nullptr}; /* Information_schema table */
Query_block *schema_query_block{nullptr};
/*
True when the view field translation table is used to convert
schema table fields for backwards compatibility with SHOW command.
*/
bool schema_table_reformed{false};
/* link to query_block where this table was used */
Query_block *query_block{nullptr};
private:
LEX *view{nullptr}; /* link on VIEW lex for merging */
public:
/// Array of selected expressions from a derived table or view.
Field_translator *field_translation{nullptr};
/// pointer to element after last one in translation table above
Field_translator *field_translation_end{nullptr};
/*
List (based on next_local) of underlying tables of this view. I.e. it
does not include the tables of subqueries used in the view. Is set only
for merged views.
*/
Table_ref *merge_underlying_list{nullptr};
/*
- 0 for base tables
- in case of the view it is the list of all (not only underlying
tables but also used in subquery ones) tables of the view.
*/
mem_root_deque<Table_ref *> *view_tables{nullptr};
/* most upper view this table belongs to */
Table_ref *belong_to_view{nullptr};
/*
The view directly referencing this table
(non-zero only for merged underlying tables of a view).
*/
Table_ref *referencing_view{nullptr};
/* Ptr to parent MERGE table list item. See top comment in ha_myisammrg.cc */
Table_ref *parent_l{nullptr};
/*
Security context (non-zero only for tables which belong
to view with SQL SECURITY DEFINER)
*/
Security_context *security_ctx{nullptr};
/*
This view security context (non-zero only for views with
SQL SECURITY DEFINER)
*/
Security_context *view_sctx{nullptr};
/*
List of all base tables local to a subquery including all view
tables. Unlike 'next_local', this in this list views are *not*
leaves. Created in setup_tables() -> make_leaf_tables().
*/
Table_ref *next_leaf{nullptr};
Item *derived_where_cond{nullptr}; ///< WHERE condition from derived table
Item *check_option{nullptr}; ///< WITH CHECK OPTION condition
Item *replace_filter{nullptr}; ///< Filter for REPLACE command
LEX_STRING select_stmt{nullptr, 0}; ///< text of (CREATE/SELECT) statement
LEX_STRING source{nullptr, 0}; ///< source of CREATE VIEW
LEX_STRING timestamp{nullptr, 0}; ///< GMT time stamp of last operation
LEX_USER definer; ///< definer of view
void set_tablesample(tablesample_type sampling_type_arg,
Item *sampling_percentage_arg) {
sampling_type = sampling_type_arg;
sampling_percentage = sampling_percentage_arg;
}
bool has_tablesample() const {
return sampling_type != tablesample_type::UNSPECIFIED_TABLESAMPLE_TYPE;
}
bool update_sampling_percentage();
double get_sampling_percentage() const;
bool validate_tablesample_clause(THD *thd);
tablesample_type get_sampling_type() const { return sampling_type; }
/**
@note: This field is currently not reliable when read from dictionary:
If an underlying view is changed, updatable_view is not changed,
due to lack of dependency checking in dictionary implementation.
Prefer to use is_updatable() during preparation and optimization.
*/
ulonglong updatable_view{0}; ///< VIEW can be updated
/**
@brief The declared algorithm, if this is a view.
@details One of
- VIEW_ALGORITHM_UNDEFINED
- VIEW_ALGORITHM_TEMPTABLE
- VIEW_ALGORITHM_MERGE
@todo Replace with an enum
*/
ulonglong algorithm{0};
ulonglong view_suid{0}; ///< view is suid (true by default)
ulonglong with_check{0}; ///< WITH CHECK OPTION
private:
/// The view algorithm that is actually used, if this is a view.
enum_view_algorithm effective_algorithm{VIEW_ALGORITHM_UNDEFINED};
Lock_descriptor m_lock_descriptor;
public:
GRANT_INFO grant;
public:
/// True if right argument of LEFT JOIN; false in other cases (i.e. if left
/// argument of LEFT JOIN, if argument of INNER JOIN; RIGHT JOINs are
/// converted to LEFT JOIN during contextualization).
bool outer_join{false};
/// True if was originally the left argument of a RIGHT JOIN, before we
/// made it the right argument of a LEFT JOIN.
bool join_order_swapped{false};
uint shared{0}; /* Used in multi-upd */
size_t db_length{0};
size_t table_name_length{0};
private:
/// True if VIEW/TABLE is updatable, based on analysis of query (SQL rules).
bool m_updatable{false};
/// True if VIEW/TABLE is insertable, based on analysis of query (SQL rules).
bool m_insertable{false};
/// True if table is target of UPDATE statement, or updated in IODKU stmt.
bool m_updated{false};
/// True if table is target of INSERT statement.
bool m_inserted{false};
/// True if table is target of DELETE statement, or deleted in REPLACE stmt.
bool m_deleted{false};
bool m_fulltext_searched{false}; ///< True if fulltext searched
public:
bool straight{false}; /* optimize with prev table */
/**
True for tables and views being changed in a data change statement.
Also true for tables subject to a SELECT ... FOR UPDATE.
Also used by replication to filter out statements that can be ignored,
especially important for multi-table UPDATE and DELETE.
*/
bool updating{false};
/// preload only non-leaf nodes (IS THIS USED???)
bool ignore_leaves{false};
/**
The set of tables in the query block that this table depends on.
Can be set due to outer join, join order hints or NOT EXISTS relationship.
*/
table_map dep_tables{0};
/// The outer tables that an outer join's join condition depends on
table_map join_cond_dep_tables{0};
/**
Is non-NULL if this table reference is a nested join, ie it represents
the inner tables of an outer join, the tables contained in the
parentheses of an inner join (eliminated during resolving), the tables
referenced in a derived table or view, in a semi-join nest, the tables
from the subquery.
*/
NESTED_JOIN *nested_join{nullptr};
/// The nested join containing this table reference.
Table_ref *embedding{nullptr};
/// The join list immediately containing this table reference
mem_root_deque<Table_ref *> *join_list{nullptr};
/// stop PS caching
bool cacheable_table{false};
/**
Specifies which kind of table should be open for this element
of table list.
*/
enum_open_type open_type{OT_TEMPORARY_OR_BASE};
/* true if this merged view contain auto_increment field */
bool contain_auto_increment{false};
/// true <=> VIEW CHECK OPTION condition is processed (also for prep. stmts)
bool check_option_processed{false};
/// true <=> Filter condition is processed
bool replace_filter_processed{false};
dd::enum_table_type required_type{};
char timestamp_buffer[20]{0}; /* buffer for timestamp (19+1) */
/*
This Table_ref object is just placeholder for prelocking, it will be
used for implicit LOCK TABLES only and won't be used in real statement.
*/
bool prelocking_placeholder{false};
/**
Indicates that if Table_ref object corresponds to the table/view
which requires special handling.
*/
enum {
/* Normal open. */
OPEN_NORMAL = 0,
/* Associate a table share only if the the table exists. */
OPEN_IF_EXISTS,
/*
Associate a table share only if the the table exists.
Also upgrade metadata lock to exclusive if table doesn't exist.
*/
OPEN_FOR_CREATE,
/* Don't associate a table share. */
OPEN_STUB
} open_strategy{OPEN_NORMAL};
bool internal_tmp_table{false};
/** true if an alias for this table was specified in the SQL. */
bool is_alias{false};
/** true if the table is referred to in the statement using a fully
qualified name (@<db_name@>.@<table_name@>).
*/
bool is_fqtn{false};
/**
If true, this table is a derived (materialized) table which was created
from a scalar subquery, cf.
Query_block::transform_scalar_subqueries_to_join_with_derived
*/
bool m_was_scalar_subquery{false};
/* View creation context. */
View_creation_ctx *view_creation_ctx{nullptr};
/*
Attributes to save/load view creation context in/from frm-file.
They are required only to be able to use existing parser to load
view-definition file. As soon as the parser parsed the file, view
creation context is initialized and the attributes become redundant.
These attributes MUST NOT be used for any purposes but the parsing.
*/
LEX_CSTRING view_client_cs_name{nullptr, 0};
LEX_CSTRING view_connection_cl_name{nullptr, 0};
/*
View definition (SELECT-statement) in the UTF-form.
*/
LEX_STRING view_body_utf8{nullptr, 0};
// True, If this is a system view
bool is_system_view{false};
/*
Set to 'true' if this is a DD table being opened in the context of a
dictionary operation. Note that when 'false', this may still be a DD
table when opened in a non-DD context, e.g. as part of an I_S view
query.
*/
bool is_dd_ctx_table{false};
/* End of view definition context. */
/* List of possible keys. Valid only for materialized derived tables/views. */
List<Derived_key> derived_key_list;
/**
Indicates what triggers we need to pre-load for this Table_ref
when opening an associated TABLE. This is filled after
the parsed tree is created.
*/
uint8 trg_event_map{0};
bool schema_table_filled{false};
MDL_request mdl_request;
/// if true, EXPLAIN can't explain view due to insufficient rights.
bool view_no_explain{false};
/* List to carry partition names from PARTITION (...) clause in statement */
List<String> *partition_names{nullptr};
/// Set table number
void set_tableno(uint tableno) {
assert(tableno < MAX_TABLES);
m_tableno = tableno;
m_map = (table_map)1 << tableno;
}
/// Return table number
uint tableno() const { return m_tableno; }
/// Return table map derived from table number
table_map map() const {
assert(((table_map)1 << m_tableno) == m_map);
return m_map;
}
/// If non-NULL, the CTE which this table is derived from.
Common_table_expr *common_table_expr() const { return m_common_table_expr; }
void set_common_table_expr(Common_table_expr *c) { m_common_table_expr = c; }
/// @see m_derived_column_names
const Create_col_name_list *derived_column_names() const {
return m_derived_column_names;
}
void set_derived_column_names(const Create_col_name_list *d) {
m_derived_column_names = d;
}
private:
/*
A group of members set and used only during JOIN::optimize().
*/
/**
Optimized copy of m_join_cond (valid for one single
execution). Initialized by Query_block::get_optimizable_conditions().
*/
Item *m_join_cond_optim{nullptr};
public:
COND_EQUAL *cond_equal{nullptr}; ///< Used with outer join
/// true <=> this table is a const one and was optimized away.
bool optimized_away{false};
/**
true <=> all possible keys for a derived table were collected and
could be re-used while statement re-execution.
*/
bool derived_keys_ready{false};
private:
/// If a recursive reference inside the definition of a CTE.
bool m_is_recursive_reference{false};
// End of group for optimization
/** See comments for set_metadata_id() */
enum_table_ref_type m_table_ref_type{TABLE_REF_NULL};
/** See comments for TABLE_SHARE::get_table_ref_version() */
ulonglong m_table_ref_version{0};
/*
All members whose names are suffixed with "_saved" are duplicated in
class TABLE but actually belong in this class. They are saved from class
TABLE when preparing a statement and restored when executing the statement.
They are not required for a regular (non-prepared) statement.
*/
Key_map covering_keys_saved;
Key_map merge_keys_saved;
Key_map keys_in_use_for_query_saved;
Key_map keys_in_use_for_group_by_saved;
Key_map keys_in_use_for_order_by_saved;
bool nullable_saved{false};
bool force_index_saved{false};
bool force_index_order_saved{false};
bool force_index_group_saved{false};
MY_BITMAP lock_partitions_saved;
MY_BITMAP read_set_saved;
MY_BITMAP write_set_saved;
MY_BITMAP read_set_internal_saved;
};
/*
Iterator over the fields of a generic table reference.
*/
class Field_iterator {
public:
virtual ~Field_iterator() = default;
virtual void set(Table_ref *) = 0;
virtual void next() = 0;
virtual bool end_of_fields() = 0; /* Return 1 at end of list */
virtual const char *name() = 0;
virtual Item *create_item(THD *) = 0;
virtual Field *field() = 0;
};
/*
Iterator over the fields of a base table, view with temporary
table, or subquery.
*/
class Field_iterator_table : public Field_iterator {
Field **ptr;
public:
Field_iterator_table() : ptr(nullptr) {}
void set(Table_ref *table) override { ptr = table->table->field; }
void set_table(TABLE *table) { ptr = table->field; }
void next() override { ptr++; }
bool end_of_fields() override { return *ptr == nullptr; }
const char *name() override;
Item *create_item(THD *thd) override;
Field *field() override { return *ptr; }
};
/**
Iterator over the fields of a merged derived table or view.
*/
class Field_iterator_view : public Field_iterator {
Field_translator *ptr, *array_end;
Table_ref *view;
public:
Field_iterator_view() : ptr(nullptr), array_end(nullptr) {}
void set(Table_ref *table) override;
void next() override { ptr++; }
bool end_of_fields() override { return ptr == array_end; }
const char *name() override;
Item *create_item(THD *thd) override;
Item **item_ptr() { return &ptr->item; }
Field *field() override { return nullptr; }
inline Item *item() { return ptr->item; }
Field_translator *field_translator() { return ptr; }
};
/*
Field_iterator interface to the list of materialized fields of a
NATURAL/USING join.
*/
class Field_iterator_natural_join : public Field_iterator {
List_iterator_fast<Natural_join_column> column_ref_it;
Natural_join_column *cur_column_ref;
public:
Field_iterator_natural_join() : cur_column_ref(nullptr) {}
~Field_iterator_natural_join() override = default;
void set(Table_ref *table) override;
void next() override;
bool end_of_fields() override { return !cur_column_ref; }
const char *name() override { return cur_column_ref->name(); }
Item *create_item(THD *thd) override {
return cur_column_ref->create_item(thd);
}
Field *field() override { return cur_column_ref->field(); }
Natural_join_column *column_ref() { return cur_column_ref; }
};
/**
Generic iterator over the fields of an arbitrary table reference.
This class unifies the various ways of iterating over the columns
of a table reference depending on the type of SQL entity it
represents. If such an entity represents a nested table reference,
this iterator encapsulates the iteration over the columns of the
members of the table reference.
The implementation assumes that all underlying NATURAL/USING table
references already contain their result columns and are linked into
the list Table_ref::next_name_resolution_table.
*/
class Field_iterator_table_ref : public Field_iterator {
Table_ref *table_ref, *first_leaf, *last_leaf;
Field_iterator_table table_field_it;
Field_iterator_view view_field_it;
Field_iterator_natural_join natural_join_it;
Field_iterator *field_it;
void set_field_iterator();
public:
Field_iterator_table_ref() : field_it(nullptr) {}
void set(Table_ref *table) override;
void next() override;
bool end_of_fields() override {
return (table_ref == last_leaf && field_it->end_of_fields());
}
const char *name() override { return field_it->name(); }
const char *get_table_name();
const char *get_db_name();
GRANT_INFO *grant();
Item *create_item(THD *thd) override { return field_it->create_item(thd); }
Field *field() override { return field_it->field(); }
Natural_join_column *get_or_create_column_ref(THD *thd,
Table_ref *parent_table_ref);
Natural_join_column *get_natural_column_ref();
};
struct OPEN_TABLE_LIST {
OPEN_TABLE_LIST *next;
char *db, *table;
uint32 in_use, locked;
};
static inline my_bitmap_map *tmp_use_all_columns(TABLE *table,
MY_BITMAP *bitmap) {
my_bitmap_map *old = bitmap->bitmap;
bitmap->bitmap = table->s->all_set.bitmap; // does not repoint last_word_ptr
return old;
}
static inline void tmp_restore_column_map(MY_BITMAP *bitmap,
my_bitmap_map *old) {
bitmap->bitmap = old;
}
/* The following is only needed for debugging */
static inline my_bitmap_map *dbug_tmp_use_all_columns(TABLE *table
[[maybe_unused]],
MY_BITMAP *bitmap
[[maybe_unused]]) {
#ifndef NDEBUG
return tmp_use_all_columns(table, bitmap);
#else
return nullptr;
#endif
}
static inline void dbug_tmp_restore_column_map(MY_BITMAP *bitmap
[[maybe_unused]],
my_bitmap_map *old
[[maybe_unused]]) {
#ifndef NDEBUG
tmp_restore_column_map(bitmap, old);
#endif
}
/*
Variant of the above : handle both read and write sets.
Provide for the possibility of the read set being the same as the write set
*/
static inline void dbug_tmp_use_all_columns(
TABLE *table [[maybe_unused]], my_bitmap_map **save [[maybe_unused]],
MY_BITMAP *read_set [[maybe_unused]],
MY_BITMAP *write_set [[maybe_unused]]) {
#ifndef NDEBUG
save[0] = read_set->bitmap;
save[1] = write_set->bitmap;
(void)tmp_use_all_columns(table, read_set);
(void)tmp_use_all_columns(table, write_set);
#endif
}
static inline void dbug_tmp_restore_column_maps(
MY_BITMAP *read_set [[maybe_unused]], MY_BITMAP *write_set [[maybe_unused]],
my_bitmap_map **old [[maybe_unused]]) {
#ifndef NDEBUG
tmp_restore_column_map(read_set, old[0]);
tmp_restore_column_map(write_set, old[1]);
#endif
}
void init_mdl_requests(Table_ref *table_list);
/**
Unpacks the definition of a value generator in all its forms: generated
column, default expression or checked constraint.
The function parses the text definition of this expression, resolves its
items and runs validation and calculates the base_columns_map which is used
for tracking the columns the expression depends on.
@param[in] thd Thread handler
@param[in] table Table having the value generator to be unpacked
@param[in,out] val_generator Contains the expression in string format, and,
if successful will be replaced by the parser
with a new one having the unpacked expression.
@param[in] source Source of value generator(a generated column,
a regular column with generated default value or
a check constraint).
@param[in] source_name Name of the source (generated column, a regular
column with generated default value or a check
constraint).
@param[in] field The column the value generator depends on. Can
be null for checked constraints which do not
depend on a single column.
@param[in] is_create_table Indicates that table is opened as part
of CREATE or ALTER and does not yet exist in SE
@param[out] error_reported updated flag for the caller that no other error
messages are to be generated.
@retval true Failure.
@retval false Success.
*/
bool unpack_value_generator(THD *thd, TABLE *table,
Value_generator **val_generator,
Value_generator_source source,
const char *source_name, Field *field,
bool is_create_table, bool *error_reported);
/**
Unpack the partition expression. Parse the partition expression
to produce an Item.
@param[in] thd Thread handler
@param[in] outparam Table object
@param[in] share TABLE_SHARE object
@param[in] engine_type Engine type of the partitions.
@param[in] is_create_table Indicates that table is opened as part of
CREATE or ALTER and does not yet exist in SE
@retval true Failure.
@retval false Success.
*/
bool unpack_partition_info(THD *thd, TABLE *outparam, TABLE_SHARE *share,
handlerton *engine_type, bool is_create_table);
int open_table_from_share(THD *thd, TABLE_SHARE *share, const char *alias,
uint db_stat, uint prgflag, uint ha_open_flags,
TABLE *outparam, bool is_create_table,
const dd::Table *table_def_param);
TABLE_SHARE *alloc_table_share(const char *db, const char *table_name,
const char *key, size_t key_length,
bool open_secondary);
void init_tmp_table_share(THD *thd, TABLE_SHARE *share, const char *key,
size_t key_length, const char *table_name,
const char *path, MEM_ROOT *mem_root);
void free_table_share(TABLE_SHARE *share);
void update_create_info_from_table(HA_CREATE_INFO *info, TABLE *form);
Ident_name_check check_db_name(const char *name, size_t length);
Ident_name_check check_and_convert_db_name(LEX_STRING *db,
bool preserve_lettercase);
bool check_column_name(const char *name);
Ident_name_check check_table_name(const char *name, size_t length);
int rename_file_ext(const char *from, const char *to, const char *ext);
char *get_field(MEM_ROOT *mem, Field *field);
bool get_field(MEM_ROOT *mem, Field *field, class String *res);
int closefrm(TABLE *table, bool free_share);
void free_blobs(TABLE *table);
void free_blob_buffers_and_reset(TABLE *table, uint32 size);
int set_zone(int nr, int min_zone, int max_zone);
void append_unescaped(String *res, const char *pos, size_t length);
char *fn_rext(char *name);
TABLE_CATEGORY get_table_category(const LEX_CSTRING &db,
const LEX_CSTRING &name);
/* performance schema */
extern LEX_CSTRING PERFORMANCE_SCHEMA_DB_NAME;
extern LEX_CSTRING GENERAL_LOG_NAME;
extern LEX_CSTRING SLOW_LOG_NAME;
/* information schema */
extern LEX_CSTRING INFORMATION_SCHEMA_NAME;
/* mysql schema name and DD ID */
extern LEX_CSTRING MYSQL_SCHEMA_NAME;
static const uint MYSQL_SCHEMA_DD_ID = 1;
/* mysql tablespace name and DD ID */
extern LEX_CSTRING MYSQL_TABLESPACE_NAME;
static const uint MYSQL_TABLESPACE_DD_ID = 1;
/* replication's tables */
extern LEX_CSTRING RLI_INFO_NAME;
extern LEX_CSTRING MI_INFO_NAME;
extern LEX_CSTRING WORKER_INFO_NAME;
inline bool is_infoschema_db(const char *name, size_t len) {
return (
INFORMATION_SCHEMA_NAME.length == len &&
!my_strcasecmp(system_charset_info, INFORMATION_SCHEMA_NAME.str, name));
}
inline bool is_infoschema_db(const char *name) {
return !my_strcasecmp(system_charset_info, INFORMATION_SCHEMA_NAME.str, name);
}
inline bool is_perfschema_db(const char *name, size_t len) {
return (PERFORMANCE_SCHEMA_DB_NAME.length == len &&
!my_strcasecmp(system_charset_info, PERFORMANCE_SCHEMA_DB_NAME.str,
name));
}
inline bool is_perfschema_db(const char *name) {
return !my_strcasecmp(system_charset_info, PERFORMANCE_SCHEMA_DB_NAME.str,
name);
}
/**
Check if the table belongs to the P_S, excluding setup and threads tables.
@note Performance Schema tables must be accessible independently of the
LOCK TABLE mode. This function is needed to handle the special case
of P_S tables being used under LOCK TABLE mode.
*/
inline bool belongs_to_p_s(Table_ref *tl) {
return (!strcmp("performance_schema", tl->db) &&
strcmp(tl->table_name, "threads") &&
strstr(tl->table_name, "setup_") == nullptr);
}
/**
return true if the table was created explicitly.
*/
inline bool is_user_table(TABLE *table) {
const char *name = table->s->table_name.str;
return strncmp(name, tmp_file_prefix, tmp_file_prefix_length);
}
bool is_simple_order(ORDER *order);
uint add_pk_parts_to_sk(KEY *sk, uint sk_n, KEY *pk, uint pk_n,
TABLE_SHARE *share, handler *handler_file,
uint *usable_parts, bool use_extended_sk);
void setup_key_part_field(TABLE_SHARE *share, handler *handler_file,
uint primary_key_n, KEY *keyinfo, uint key_n,
uint key_part_n, uint *usable_parts,
bool part_of_key_not_extended);
const uchar *get_field_name(const uchar *arg, size_t *length);
void repoint_field_to_record(TABLE *table, uchar *old_rec, uchar *new_rec);
bool update_generated_write_fields(const MY_BITMAP *bitmap, TABLE *table);
bool update_generated_read_fields(uchar *buf, TABLE *table,
uint active_index = MAX_KEY);
/**
Check if a Table_ref instance represents a pre-opened temporary table.
*/
inline bool is_temporary_table(const Table_ref *tl) {
if (tl->is_view() || tl->schema_table) return false;
if (!tl->table) return false;
/*
NOTE: 'table->s' might be NULL for specially constructed TABLE
instances. See SHOW TRIGGERS for example.
*/
if (!tl->table->s) return false;
return tl->table->s->tmp_table != NO_TMP_TABLE;
}
/**
After parsing, a Common Table Expression is accessed through a
Table_ref. This class contains all information about the CTE which the
Table_ref needs.
@note that before and during parsing, the CTE is described by a
PT_common_table_expr.
*/
class Common_table_expr {
public:
Common_table_expr(MEM_ROOT *mem_root)
: references(mem_root), recursive(false), tmp_tables(mem_root) {}
TABLE *clone_tmp_table(THD *thd, Table_ref *tl);
bool substitute_recursive_reference(THD *thd, Query_block *sl);
/// Remove one table reference.
void remove_table(Table_ref *tr);
/// Empties the materialized CTE and informs all of its clones.
bool clear_all_references();
/**
All references to this CTE in the statement, except those inside the
query expression defining this CTE.
In other words, all non-recursive references.
*/
Mem_root_array<Table_ref *> references;
/// True if it's a recursive CTE
bool recursive;
/**
List of all TABLE_LISTSs reading/writing to the tmp table created to
materialize this CTE. Due to shared materialization, only the first one
has a TABLE generated by create_tmp_table(); other ones have a TABLE
generated by open_table_from_share().
*/
Mem_root_array<Table_ref *> tmp_tables;
/// Name of the WITH block. Used only for EXPLAIN FORMAT=tree.
LEX_STRING name;
};
/**
This iterates on those references to a derived table / view / CTE which are
materialized. If a recursive CTE, this includes recursive references.
Upon construction it is passed a non-recursive materialized reference
to the derived table (Table_ref*).
For a CTE it may return more than one reference; for a derived table or a
view, there is only one (as references to a same view are treated as
independent objects).
References are returned as TABLE*.
*/
class Derived_refs_iterator {
Table_ref *const start; ///< The reference provided in construction.
size_t ref_idx{0}; ///< Current index in cte->tmp_tables
bool m_is_first{true}; ///< True when at first reference in list
public:
explicit Derived_refs_iterator(Table_ref *start_arg) : start(start_arg) {}
TABLE *get_next() {
const Common_table_expr *cte = start->common_table_expr();
m_is_first = ref_idx == 0;
// Derived tables and views have a single reference.
if (cte == nullptr) {
return ref_idx++ == 0 ? start->table : nullptr;
}
/*
CTEs may have multiple references. Return the next one, but notice that
some references may have been deleted.
*/
while (ref_idx < cte->tmp_tables.size()) {
TABLE *table = cte->tmp_tables[ref_idx++]->table;
if (table != nullptr) return table;
}
return nullptr;
}
void rewind() {
ref_idx = 0;
m_is_first = true;
}
/// @returns true if the last get_next() returned the first element.
bool is_first() const {
// Call after get_next() has been called:
assert(ref_idx > 0);
return m_is_first;
}
};
/**
RAII class to reset TABLE::autoinc_field_has_explicit_non_null_value after
processing individual row in INSERT or LOAD DATA statements.
*/
class Autoinc_field_has_explicit_non_null_value_reset_guard {
public:
Autoinc_field_has_explicit_non_null_value_reset_guard(TABLE *table)
: m_table(table) {}
~Autoinc_field_has_explicit_non_null_value_reset_guard() {
m_table->autoinc_field_has_explicit_non_null_value = false;
}
private:
TABLE *m_table;
};
// Whether we can ask the storage engine for the row ID of the last row read.
//
// Some operations needs a row ID to operate correctly (i.e. weedout). Normally,
// the row ID is provided by the storage engine by calling handler::position().
// But there are cases when position() should not be called:
//
// 1. If we have a const table (rows are fetched during optimization), we
// should not call position().
// 2. If we have a NULL-complemented row, calling position() would give a
// random row ID back, as there has not been any row read.
//
// Operations that needs the row ID must also check the value of
// QEP_TAB::rowid_status to see whether they actually need a row ID.
// See QEP_TAB::rowid_status for more details.
inline bool can_call_position(const TABLE *table) {
return !table->const_table && !(table->is_nullable() && table->null_row);
}
//////////////////////////////////////////////////////////////////////////
/*
NOTE:
These structures are added to read .frm file in upgrade scenario.
They should not be used any where else in the code.
They will be removed in future release.
Any new code should not be added in this section.
*/
/**
These members were removed from TABLE_SHARE as they are not used in
in the code. open_binary_frm() uses these members while reading
.frm files.
*/
class FRM_context {
public:
FRM_context()
: default_part_db_type(nullptr),
null_field_first(false),
stored_fields(0),
view_def(nullptr),
frm_version(0),
fieldnames() {}
handlerton *default_part_db_type;
bool null_field_first;
uint stored_fields; /* Number of stored fields
(i.e. without generated-only ones) */
enum utype {
NONE,
DATE,
SHIELD,
NOEMPTY,
CASEUP,
PNR,
BGNR,
PGNR,
YES,
NO,
REL,
CHECK,
EMPTY_VAL, // EMPTY_VAL rather than EMPTY since EMPTY can conflict with
// system headers.
UNKNOWN_FIELD,
CASEDN,
NEXT_NUMBER,
INTERVAL_FIELD,
BIT_FIELD,
TIMESTAMP_OLD_FIELD,
CAPITALIZE,
BLOB_FIELD,
TIMESTAMP_DN_FIELD,
TIMESTAMP_UN_FIELD,
TIMESTAMP_DNUN_FIELD,
GENERATED_FIELD = 128
};
/**
For shares representing views File_parser object with view
definition read from .FRM file.
*/
const File_parser *view_def;
uchar frm_version;
TYPELIB fieldnames; /* Pointer to fieldnames */
};
/**
Create TABLE_SHARE from .frm file.
FRM_context object is used to store the value removed from
TABLE_SHARE. These values are used only for .frm file parsing.
@param[in] thd Thread handle.
@param[in] path Path of the frm file.
@param[out] share TABLE_SHARE to be populated.
@param[out] frm_context FRM_context object.
@param[in] db Database name.
@param[in] table Table name.
@param[in] is_fix_view_cols_and_deps Fix view column data, table
and routine dependency.
@retval 0 ON SUCCESS
@retval -1 ON FAILURE
@retval -2 ON LESS SEVER FAILURE (see read_frm_file)
*/
int create_table_share_for_upgrade(THD *thd, const char *path,
TABLE_SHARE *share, FRM_context *frm_context,
const char *db, const char *table,
bool is_fix_view_cols_and_deps);
//////////////////////////////////////////////////////////////////////////
/**
Create a copy of the key_info from TABLE_SHARE object to TABLE object.
Wherever prefix key is present, allocate a new Field object, having its
field_length set to the prefix key length, and point the table's matching
key_part->field to this new Field object.
This ensures that unpack_partition_info() reads the correct prefix length of
partitioned fields
@param table Table for which key_info is to be allocated
@param root MEM_ROOT in which to allocate key_info
@retval false Success
@retval true Failed to allocate memory for table.key_info in root
*/
bool create_key_part_field_with_prefix_length(TABLE *table, MEM_ROOT *root);
#endif /* TABLE_INCLUDED */
Reference in a new issue View git blame Copy permalink