net.akaish.kitty.orm.annotations.column

Annotation Type KITTY_COLUMN

@Retention(value=RUNTIME)
 @Target(value=FIELD)
 @Inherited
public @interface KITTY_COLUMN

This interface designed for describing modelClass fields to be used in this lib automatically with SQLite DB. There are some things that not obvious and while writing it it was kind of revelation.

How to set synthetic primary key here, through field types and affinities:

If you want to use this field as primary key associated, use TypeAffinities.INTEGER as affinity in columnAffinity() without PRIMARY_KEY.autoincrement() flag and without AscDesc.DESCENDING as value for PRIMARY_KEY.orderAscDesc(), cause:
1) there some issues with aliasing primary key with oid if key type is not integer, however other integer types such as INT and BIGINT not presented in this code, they can be presented in future ( Other integer type names like "INT" or "BIGINT" or "SHORT INTEGER" or "UNSIGNED INTEGER" causes the primary key column to behave as an ordinary table column with integer affinity and a unique tableIndex, not as an alias for the rowid.)
2) CREATE TABLE t(x INTEGER PRIMARY KEY DESCENDING, y, z) doesn't result in "x" being an alias for the rowid
3) the main purpose of using AUTOINCREMENT attribute is to prevent SQLite to reuse value that has not been used or from the previously deleted row. In case if integer primary key is associated with oid (row id) it would be just overkill.
So, use for synthetic primary key something like "x INTEGER NOT NULL PRIMARY KEY"
You can achieve this in two different ways:
1) In KITTY_COLUMN set isIPK() to true and columnAffinity() to TypeAffinities.INTEGER or
2) Set in KITTY_COLUMN @link KITTY_COLUMN#columnAffinity()} to TypeAffinities.INTEGER and annotate model's field with PRIMARY_KEY without setting PRIMARY_KEY.autoincrement() and PRIMARY_KEY.orderAscDesc() to value AscDesc.DESCENDING. Also do not forget to annotate field with NOT_NULL (SQLite supports NULL values for IPK but they not supported by KittyORM).

Why not implemented such things as additional types (such as SMALLINT, VARCHAR etc) and field size constraints (such as VARCHAR(20))
Ok, it not implemented cause it just doesn't work. As I understood, there is no difference between, for example INT and BIGINT in SQLite, cause anyway it would be treated not by container but with more abstract storage class and type affinity and as for me, it is better to use only type affinities, cause if you would put 8-bit value in TINYINT it would work. I suppose all this types were added only for implementation of SQL language. And anyway all would be stored as text in db. Also, length restrictions just ignored in SQLite. You can set check constraint for checking type of value between insertion it db table, but I don't think it is needed.

Author:

akaish (Denis Bogomolov)

See Also:

SQL As Understood By SQLite#CREATE TABLE, SQLite AUTOINCREMENT : Why You Should Avoid Using It

Required Element Summary

Required Elements

Modifier and Type	Required Element and Description
int	columnOrder Returns field order

Optional Element Summary

Optional Elements

Modifier and Type	Optional Element and Description
TypeAffinities	columnAffinity Returns associated with this variable column type of sql table
java.lang.String	columnName Returns associated with this variable column name of sql table
boolean	isIPK Returns if this field should be used as simple tableIndex column for this table If true than resulting column would automatically get two constraints: NOT_NULL @NOT_NULL PRIMARY_KEY @PRIMARY_KEY{orderAscDesc = ASCENDING} This would result in something like 'COLNAME INTEGER NOT NULL PRIMARY KEY ASC' Also, notice that this would rewrite columnAffinity() to INTEGER and any NOT_NULL and PRIMARY_KEY would be skipped.
boolean	isValueGeneratedOnInsert Returns true if this field is part of primary key and this value should be generated on databaseClass insert What is idea of this field?

Element Detail

columnOrder

public abstract int columnOrder

Returns field order

Returns:

columnName

public abstract java.lang.String columnName

Returns associated with this variable column name of sql table

Returns:

Default:

""

columnAffinity

public abstract TypeAffinities columnAffinity

Returns associated with this variable column type of sql table

Returns:

Default:

net.akaish.kitty.orm.enums.TypeAffinities.NOT_SET_USE_DEFAULT_MAPPING

isValueGeneratedOnInsert

public abstract boolean isValueGeneratedOnInsert

Returns true if this field is part of primary key and this value should be generated on databaseClass insert
What is idea of this field? KittyMapper.save(KittyModel) method should understand what operation on model should be performed, insert or update. Firstly this method checks availability of rowid, if mapper has rowid support turned off or table was created with WITHOUT ROWID, than method would try to check all columns of primary key and fields of model that represents table on NULL values.
So, if all of fields of model that are representations of columns of PK that are marked as generated on insert are NULL than insert operation would be performed, otherwise update operation would be run.
To enable this feature you have to implement triggers for those values manually at create\migrate scripts

Default:

false

isIPK

public abstract boolean isIPK

Returns if this field should be used as simple tableIndex column for this table
If true than resulting column would automatically get two constraints:
NOT_NULL @NOT_NULL
PRIMARY_KEY @PRIMARY_KEY{orderAscDesc = ASCENDING}
This would result in something like 'COLNAME INTEGER NOT NULL PRIMARY KEY ASC'
Also, notice that this would rewrite columnAffinity() to INTEGER and any NOT_NULL and PRIMARY_KEY would be skipped.
And make sure that field has Long type in order that KittyMapper would process KittyMapper.save(KittyModel) operation effectively

Returns:

Default:

false