You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
koffice/kexi/kexidb/indexschema.h

210 lines
7.8 KiB

/* This file is part of the KDE project
Copyright (C) 2003-2004 Jaroslaw Staniek <js@iidea.pl>
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Library General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Library General Public License for more details.
You should have received a copy of the GNU Library General Public License
along with this library; see the file COPYING.LIB. If not, write to
the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
* Boston, MA 02110-1301, USA.
*/
#ifndef KEXIDB_INDEX_H
#define KEXIDB_INDEX_H
#include <qvaluelist.h>
#include <qstring.h>
#include <kexidb/fieldlist.h>
#include <kexidb/schemadata.h>
#include <kexidb/relationship.h>
namespace KexiDB {
class Connection;
class TableSchema;
class QuerySchema;
class Relationship;
/*! @short Provides information about database index that can be created for a database table.
IndexSchema object stores information about table fields that
defines this index and additional properties like: whether index is unique
or primary key (requires unique). Single-field index can be also auto generated.
*/
class KEXI_DB_EXPORT IndexSchema : public FieldList, public SchemaData
{
public:
typedef QPtrList<IndexSchema> List;
typedef QPtrListIterator<IndexSchema> ListIterator;
/*! Constructs empty index schema object
that is assigned to \a table, and will be owned by this table.
Any fields added with addField() won't be owned by index,
but by its table. Do not forget to add these fields to table,
because adding these to IndexSchema is not enough.
*/
IndexSchema(TableSchema *tableSchema);
/*! Copy constructor. Copies all attributes from index \a idx, and
fields assigned with it but the fields are taken (by name) from
\a parentTable, not from \a idx itself, so it's possible to copy of index
for a copy of table.
To copy an index within the same table it's enough to call:
\code
new IndexSchema(idx, *idx.table());
\endcode
@todo All relationships should be also copied
*/
IndexSchema(const IndexSchema& idx, TableSchema& parentTable);
/*! Destroys the index. Field objects are not deleted.
All Relationship objects listed in masterRelationships() list
are destroyed (these are also detached from
detail-side indices before destruction).
Relationship objects listed in detailsRelationships() are not touched. */
virtual ~IndexSchema();
/*! Adds field at the end of field list.
Field will not be owned by index. Field must belong to a table
the index is bulit on, otherwise field couldn't be added. */
virtual FieldList& addField(Field *field);
/*! \return table that index is defined for. */
TableSchema* table() const;
/*! \return list of relationships from the table (of this index),
i.e. any such relationship in which this table is at 'master' side.
See Relationship class documentation for more information.
All objects listed here will be automatically destroyed on this IndexSchema object destruction. */
Relationship::List* masterRelationships() { return &m_master_rels; }
/*! \return list of relationships to the table (of this index),
i.e. any such relationship in which this table is at 'details' side.
See Relationship class documentation for more information. */
Relationship::List* detailsRelationships() { return &m_details_rels; }
/*! Attaches relationship definition \a rel to this IndexSchema object.
If \a rel relationship has this IndexSchema defined at the master-side,
\a rel is added to the list of master relationships (available with masterRelationships()).
If \a rel relationship has this IndexSchema defined at the details-side,
\a rel is added to the list of details relationships (available with detailsRelationships()).
For the former case, attached \a rel object is now owned by this IndexSchema object.
Note: call detachRelationship() for IndexSchema object that \a rel
was previously attached to, if any. */
void attachRelationship(Relationship *rel);
/*! Detaches relationship definition \a rel for this IndexSchema object
from the list of master relationships (available with masterRelationships()),
or from details relationships list, depending for which side of the relationship
is this IndexSchem object assigned.
Note: If \a rel was detached from masterRelationships() list,
this object now has no parent, so you need to attach it to somewhere or destruct it.
*/
void detachRelationship(Relationship *rel);
/*! \return true if index is auto-generated.
Auto-generated index is one-field index
that was automatically generated
for CREATE TABLE statement when the field has
UNIQUE or PRIMARY KEY constraint enabled.
Any newly created IndexSchema object
has this flag set to false.
This flag is handled internally by TableSchema.
It can be usable for GUI application if we do not
want display implicity/auto generated indices
on the indices list or we if want to show these
indices to the user in a special way.
*/
bool isAutoGenerated() const;
/*! \return true if this index is primary key of its table.
This can be one or multifield. */
bool isPrimaryKey() const;
/*! Sets PRIMARY KEY flag. \sa isPrimary().
Note: Setting PRIMARY KEY on (true),
UNIQUE flag will be also implicity set. */
void setPrimaryKey(bool set);
/*! \return true if this is unique index.
This can be one or multifield. */
bool isUnique() const;
/*! Sets UNIQUE flag. \sa isUnique().
Note: Setting UNIQUE off (false), PRIMARY KEY flag will
be also implicity set off, because this UNIQUE
is the requirement for PRIMARY KEYS. */
void setUnique(bool set);
/*! \return String for debugging purposes. */
virtual QString debugString();
protected:
/*! Internal constructor for convenience.
Constructs a new index schema object
that is assigned, and adds one \a field to it.
The field must be assigned to a table.
This results in the implicit index that is usually used interanlly
to declare foreign keys, used in relationships.
*/
// IndexSchema(Field* singleField);
/*! Sets auto-generated flag. This method should be called only
from TableSchema code
\sa isAutoGenerated(). */
void setAutoGenerated(bool set);
/*! If \a set is true, declares that the index defines a foreign key,
created implicity for Relationship object. Setting this to true, implies
clearing 'primary key', 'unique' and 'auto generated' flags.
If this index contains just single field, it's 'foreign field'
flag will be set to true as well. */
void setForeignKey(bool set);
/*! Internal version of attachRelationship(). If \a ownedByMaster is true,
attached \a rel object will be owned by this index. */
void attachRelationship(Relationship *rel, bool ownedByMaster);
TableSchema *m_tableSchema; //! table on that index is built
/*! a list of master relationships for the table (of this index),
this index is a master key for these relationships
and therefore - owner of these */
Relationship::List m_master_owned_rels;
/*! a list of master relationships that are not owned by this schema */
Relationship::List m_master_rels;
/*! a list of relationships to table (of this index) */
Relationship::List m_details_rels;
bool m_primary : 1;
bool m_unique : 1;
bool m_isAutoGenerated : 1;
bool m_isForeignKey : 1;
friend class Connection;
friend class TableSchema;
friend class QuerySchema;
friend class Relationship;
};
} //namespace KexiDB
#endif