Movatterモバイル変換

NdbDictionary

List

This is used for defining and retrieving data object metadata. It also includes methods for creating and dropping database objects.

The following table lists the public methods of this class and the purpose or use of each method:

Database objects such as tables and indexes created using theDictionary::create*() methods cannot be seen by the MySQL Server. This means that they cannot be accessed by MySQL clients, and that they cannot be replicated. For these reasons, it is often preferable to avoid working with them.

TheDictionary class does not have any methods for working directly with columns. You must useColumn class methods for this purpose—seeSection 2.3.1, “The Column Class”, for details.

SeeSection 2.3.10, “The List Class”, andSection 2.3.4, “The Element Structure”.

This method creates a new instance of theDictionary class.

Both the constructor and destructor for this class are protected methods, rather than public.

protected Dictionary    (      Ndb&ndb    )

AnNdb object.

ADictionary object.

The destructor takes no parameters and returns nothing:

protected ~Dictionary    (      void    )

Starts a schema transaction. An error occurs if a transaction is already active, or if the kernel metadata is locked. You can determine whether a schema transaction already exists using thehasSchemaTrans() method.

Ametadata operation occurs whenever data objects are created, altered, or dropped; such an operation can create additional suboperations in the NDB kernel.

TheNdb object and its associatedDictionary support one schema transaction at a time. By default, each metadata operation is executed separately; that is, for each operation, a schema transaction is started implicitly, the operation (including any suboperations) is executed, and the transaction is closed.

It is also possible to begin and end a schema transaction explicitly, and execute a set of user-defined operations atomically within its boundaries. In this case, all operations within the schema transaction either succeed, or are aborted and rolled back, as a unit. This is done by following the steps listed here:

Each operation is sent to the NDB kernel, which parses and saves it. A parse failure results in a rollback to the previous user operation before returning, at which point the user can either continue with or abort the entire transaction.

After all operations have been submitted,endSchemaTrans() processes and commits them. In the event of an error, the transaction is immediately aborted.

If the user exits before callingendSchemaTrans(), the NDB kernel aborts the transaction. If the user exits before the call toendSchemaTrans() returns, the kernel continues with the request, and its completion status is reported in the cluster log.

int beginSchemaTrans    (      void    )

None.

Returns 0 on success, -1 on error.

This method creates a new data file, given aDatafile object.

int createDatafile    (      const Datafile&dFile    )

A single argument—a reference to an instance ofDatafile—is required.

0 on success,-1 on failure.

Creates an event, given a reference to anEvent object.

You should keep in mind that the NDB API does not track allocated event objects, which means that the user must delete theEvent that was obtained usingcreateEvent(), after this object is no longer required.

int createEvent    (      const Event&event    )

A referenceevent to anEvent object.

0 on success,-1 on failure.

Creates aForeignKey object, given a reference to this object and anObject ID.

int createForeignKey    (      const ForeignKey&,      ObjectId* = 0,      intflags = 0    )

A reference to theForeignKey object, and anObject ID. An optional valueflags, if used, allows the creation of the foreign key without performing any foreign key checks. If set, its value must beCreateFK_NoVerify (1).

0 on success.

Creates aHashMap.

int createHashMap    (      const HashMap&hashmap,      ObjectId*id = 0    )

A reference to the hash map, and, optionally, an ID to be assigned to it.

Returns 0 on success; on failure, returns -1 and sets an error.

This method creates an index given an instance ofIndex and possibly an optional instance ofTable.

This method can be invoked with or without a reference to a table object:

Required: A reference to anIndex object.Optional: A reference to aTable object.

0 on success,-1 on failure.

This method creates a new log file group, given an instance ofLogfileGroup.

int createLogfileGroup    (      const LogfileGroup&lGroup    )

A single argument, a reference to aLogfileGroup object, is required.

0 on success,-1 on failure.

This method is used to create anNdbRecord object for use in table or index scanning operations.

The signature of this method depends on whether the resulting NdbRecord is to be used in table or index operations:

To create anNdbRecord for use in table operations, use the following:

NdbRecord* createRecord    (      const Table*table,      const RecordSpecification*recSpec,      Uint32length,      Uint32elSize    )

To create anNdbRecord for use in index operations, you can use either of the following:

NdbRecord* createRecord    (      const Index*index,      const Table*table,      const RecordSpecification*recSpec,      Uint32length,      Uint32elSize    )

or

NdbRecord* createRecord    (      const Index*index,      const RecordSpecification*recSpec,      Uint32length,      Uint32elSize    )

Dictionary::createRecord() takes the following parameters:

AnNdbRecord for use in operations involving the given table or index.

SeeSection 2.3.22, “The NdbRecord Interface”.

Creates a table given an instance ofTable.

Tables created using this method cannot be seen by the MySQL Server, cannot be updated by MySQL clients, and cannot be replicated.

int createTable    (      const Table&table    )

An instance ofTable. SeeSection 2.3.27, “The Table Class”, for more information.

0 on success,-1 on failure.

This method creates a new tablespace, given aTablespace object.

int createTablespace    (      const Tablespace&tSpace    )

This method requires a single argument—a reference to an instance ofTablespace.

0 on success,-1 on failure.

This method creates a new undo file, given anUndofile object.

int createUndofile    (      const Undofile&uFile    )

This method requires one argument: a reference to an instance ofUndofile.

0 on success,-1 on failure.

This method drops a data file, given aDatafile object.

int dropDatafile    (      const Datafile&dFile    )

A single argument—a reference to an instance ofDatafile—is required.

0 on success,-1 on failure.

This method drops an event, given a reference to anEvent object.

int dropEvent    (      const char*name,      intforce = 0    )

This method takes two parameters:

0 on success,-1 on failure.

This method drops a foreign key, given a reference to anForeignKey object to be dropped.

int dropForeignKey    (      const ForeignKey&    )

A reference to theForeignKey to be dropped.

0 on success.

This method drops an index given an instance ofIndex, and possibly an optional instance ofTable.

int dropIndex    (      const Index&index    )

int dropIndex    (      const Index&index,      const Table&table    )

This method takes two parameters, one of which is optional:

0 on success,-1 on failure.

Given an instance ofLogfileGroup, this method drops the corresponding log file group.

int dropLogfileGroup    (      const LogfileGroup&lGroup    )

A single argument, a reference to aLogfileGroup object, is required.

0 on success,-1 on failure.

Drops a table given an instance ofTable.

This method drops all foreign key constraints on thetable that is being dropped, whether the dropped table acts as a parent table, child table, or both.

Prior to NDB 8.0, anNDB table dropped using this method persisted in the MySQL data dictionary but could not be dropped usingDROP TABLE in themysql client. In NDB 8.0, such“orphan” tables can be dropped usingDROP TABLE. (Bug #29125206, Bug #93672)

int dropTable    (      const Table&table    )

An instance ofTable. SeeSection 2.3.27, “The Table Class”, for more information.

0 on success,-1 on failure.

This method drops a tablespace, given aTablespace object.

int dropTablespace    (      const Tablespace&tSpace    )

This method requires a single argument—a reference to an instance ofTablespace.

0 on success,-1 on failure.

This method drops an undo file, given anUndofile object.

int dropUndofile    (      const Undofile&uFile    )

This method requires one argument: a reference to an instance ofUndofile.

0 on success,-1 on failure.

Ends a schema transaction begun withbeginSchemaTrans(); causes operations to be processed and either committed, or aborted and rolled back. This method combines transaction execution and closing; separate methods for these tasks are not required (or implemented). This method may be called successfully even if no schema transaction is currently active.

As with many other NDB API methods, it is entirely possible forendSchemaTrans() to overwrite any current error code. For this reason, you should first check for and save any error code that may have resulted from a previous, failed operation.

int endSchemaTrans    (      Uint32flags = 0    )

The flags determines how the completed transaction is handled. The default is 0, which causes the transaction to be committed.

Dictionary::SchemaTransFlag.  You can also use withendSchemaTrans() either of theSchemaTransFlag values shown here:

Returns 0 on success; in the event of an error, returns -1 and sets anNdbError error code.

This method is used to retrieve aDatafile object, given the node ID of the data node where a data file is located and the path to the data file on that node's file system.

Datafile getDatafile    (      Uint32nodeId,      const char*path    )

This method must be invoked using two arguments, as shown here:

ADatafile object—seeSection 2.3.2, “The Datafile Class”, for details.

Get a table's default hash map.

int getDefaultHashMap    (      HashMap&dst,      Uint32fragments    )

or

int getDefaultHashMap    (      HashMap&dst,      Uint32buckets,      Uint32fragments    )

Returns 0 on success; on failure, returns -1 and sets an error.

This method is used to obtain a newEvent object representing an event, given the event's name.

getEvent() allocates memory each time it is successfully called. You should keep in mind that successive invocations of this method using the same event name return multiple, distinct objects.

The NDB API does not track allocated event objects, which means that the user must clean up eachEvent created usinggetEvent() withdelete, after the object is no longer required. Beginning with NDB 8.0.30, you can do this withreleaseEvent() instead.

const Event* getEvent    (      const char*eventName    )

TheeventName, a string (character pointer).

A pointer to anEvent object. SeeSection 2.3.5, “The Event Class”, for more information.

This method is used to obtain a newForeignKey object representing an event, given a reference to the foreign key and its name.

int getForeignKey    (      ForeignKey&dst,      const char*name    )

A reference to the foreign key and itsname, a string (character pointer).

A pointer to aForeignKey object.

Gets a hash map by name or by table.

int getHashMap    (      HashMap&dst,      const char*name    )

or

int getHashMap    (      HashMap&dst,      const Table*table    )

A reference to the hash map and either a name or aTable.

Returns 0 on success; on failure, returns -1 and sets an error.

This method retrieves a pointer to an index, given the name of the index and the name of the table to which the table belongs.

const Index* getIndex    (      const char*iName,      const char*tName    ) const

Two parameters are required:

Both of these are string values, represented by character pointers.

A pointer to anIndex. SeeSection 2.3.8, “The Index Class”, for information about this object.

This method gets aLogfileGroup object, given the name of the log file group.

LogfileGroup getLogfileGroup    (      const char*name    )

Thename of the log file group.

An instance ofLogfileGroup; seeSection 2.3.9, “The LogfileGroup Class”, for more information.

This method retrieves the most recentNDB API error.

const struct NdbError& getNdbError    (      void    ) const

None.

A reference to anNdbError object.

This method can be used to access aTable whose name is already known.

const Table* getTable    (      const char*name    ) const

Thename of the table.

A pointer to the table, orNULL if there is no table with thename supplied.

Given either the name or ID of a tablespace, this method returns the correspondingTablespace object.

This method can be invoked in either of the following two ways:

Either one of the following:

ATablespace object, as discussed inSection 2.3.28, “The Tablespace Class”.

This method gets anUndofile object, given the ID of the node where an undo file is located and the file system path to the file.

Undofile getUndofile    (      Uint32nodeId,      const char*path    )

This method requires the following two arguments:

An instance ofUndofile. For more information, seeSection 2.3.29, “The Undofile Class”.

Tells whether an NDB API schema transaction is ongoing.

bool hasSchemaTrans    (      void    ) const

None.

Returns booleanTRUE if a schema transaction is in progress, otherwiseFALSE.

Initialize a default hash map for a table.

int initDefaultHashMap    (      HashMap& dst,      Uint32 fragments    )

or

int initDefaultHashMap    (      HashMap& dst,      Uint32 buckets,      Uint32 fragments    )

A reference to the hash map and the number of fragments. Optionally the number of buckets.

Returns 0 on success; on failure, returns -1 and sets an error.

This method is used to invalidate a cached index object.

The index invalidated by this method can be referenced either as anIndex object (using a pointer), or by index name and table name, as shown here:

void invalidateIndex    (      const char*indexName,      const char*tableName    )void invalidateIndex    (      const Index*index    )

The names of the index to be removed from the cache and the table to which it belongs (indexName andtableName, respectively), or a pointer to the correspondingIndex object.

None.

This method is used to invalidate a cached table object.

void invalidateTable    (      const char*name    )

It is also possibloe to use aTable object rather than the name of the table, as shown here:

void invalidateTable    (      const Table*table    )

Thename of the table to be removed from the table cache, or a pointer to the correspondingTable object.

None.

This method returns a list of all events defined within the dictionary.

int listEvents    (      List&list    )

A reference to an emptyList. In NDB 8.0.29 and later, useclear() to empty a previously usedList for reuse.

0 on success;-1 on failure.

This method is used to obtain aList of all the indexes on a table, given the table's name.

int listIndexes    (      List&list,      const char*table) const

listIndexes() takes two arguments, both of which are required:

0 on success,-1 on failure.

This method is used to obtain a list of objects in the dictionary. It is possible to get all of the objects in the dictionary, or to restrict the list to objects of a single type.

This method has two signatures:

int listObjects    (      List&list,      Object::Typetype = Object::TypeUndefined    ) const

and

int listObjects    (      List&list,      Object::Typetype,      boolfullyQualified    ) const

A reference to an emptyList object is required—this is the list that contains the dictionary's objects afterlistObjects() is called. (SeeSection 2.3.10, “The List Class”.) An optional second argumenttype may be used to restrict the list to only those objects of the given type—that is, of the specifiedObject::Type. Iftype is not given, then the list contains all of the dictionary's objects.

You can also specify whether or not the object names in thelist are fully qualified (that is, whether the object name includes the database, schema, and possibly the table name). If you specifyfullyQualified, then you must also specify thetype.

In NDB 8.0.29 and later, you can callclear() to empty a previously usedList for reuse.

0 on success,-1 on failure.

Creates or retrieves a hash map suitable for alteration. Requires a schema transaction to be in progress; seeDictionary::beginSchemaTrans(), for more information.

Either of the following:

References to the old and new tables. Optionally, a number of buckets.

Returns 0 on success; on failure, returns -1 and sets an error.

This method is used to free anEvent after it is no longer needed. Typically this is an event returned bygetEvent().

void releaseEvent    (      const Event*event    )

TheEvent to be cleaned up.

None.

This method is used to free anNdbRecord after it is no longer needed.

void releaseRecord    (      NdbRecord*record    )

TheNdbRecord to be cleaned up.

None.

SeeSection 2.3.22, “The NdbRecord Interface”.

This method removes the table, specified by name, from the local cache.

void removeCachedTable    (      const char*table    )

The name of thetable to be removed from the cache.

None.

This method removes the specified index from the local cache, given the name of the index and that of the table in which it is contained.

void removeCachedIndex    (      const char*index,      const char*table    )

TheremoveCachedIndex() method requires two arguments:

None.