DEF-VIEW-CLASS — Defines CLOS classes with mapping to SQL database.Macro
nameThe class name.
superclassesThe superclasses for the defined class.
slotsThe class slot definitions.
class optionsThe class options.
classThe defined class.
:db-kind - specifies the kind of
database mapping which is performed for this slot and
defaults to :base which indicates
that the slot maps to an ordinary column of the database
table. A :db-kind value of
:key indicates that this slot is a
special kind of :base slot which
maps onto a column which is one of the unique keys for the
database table, the value :join
indicates this slot represents a join onto another
View Class
which contains View Class objects, and the value
:virtual indicates a standard CLOS
slot which does not map onto columns of the database
table.
:db-info - if a slot is specified
with :db-kind
:join, the slot option
:db-info contains a property list
which specifies the nature of the join. The valid members
of the list are:
:join-class
class-name - the name of the
class to join on.
:home-key
slot-name - the name of the slot
of this class for joining
:foreign-key
slot-name - the name of the slot
of the :join-class for joining
:target-slot
target-slot - this is an optional
parameter. If specified, then the join slot of the
defining class will contain instances of this target
slot rather than of the join class. This can be useful
when the :join-class is an
intermediate class in a
many-to-many relationship and the
application is actually interested in the
:target-slot.
:retrieval
time - The default value is
:deferred, which defers filling
this slot until the value is accessed. The other valid
value is :immediate which
performs the SQL query when the instance of the class
is created. In this case, the
:set is automatically set to
NIL
:set set -
This controls what is stored in the join slot. The
default value is T. When set is
T and target-slot is undefined,
the join slot will contain a list of instances of the
join class. Whereas, if
target-slot is defined, then the
join slot will contain a list of pairs of
(target-value join-instance).
When set is NIL, the join slot
will contain a single instances.
:type - for slots of
:db-kind :base or
:key, the :type slot
option has a special interpretation such that Lisp
types, such as string, integer and float are
automatically converted into appropriate SQL types for
the column onto which the slot maps. This behaviour may
be overridden using the :db-type slot
option. The valid values are:
string - a variable length
character field up to *default-string-length*
characters.
|
(string n) - a fixed length
character field n characters
long.
|
varchar - a variable length
character field up to *default-string-length*
characters.
|
(varchar n) - a variable length
character field up to n
characters in length.
|
char - a single character field
|
integer - signed integer
at least 32-bits wide |
(integer n) |
float |
(float n) |
long-float |
number |
(number n) |
(number n p) |
tinyint - An integer column 8-bits
wide. [not supported by all database backends]
|
smallint - An integer column 16-bits
wide. [not supported by all database backends]
|
bigint - An integer column
64-bits wide. [not supported by all database backends]
|
universal-time - an integer
field sufficiently wide to store a
universal-time. On most databases, a slot of this
type assigned a SQL type of
BIGINT
|
wall-time - a slot which stores
a date and time in a SQL timestamp column. CLSQL
provides a number of time manipulation functions to
support objects of type wall-time.
|
date - a slot which stores the
date (without any time of day resolution) in a
column. CLSQL provides a number of time
manipulation functions that operate on date values.
|
duration - stores a
duration structure. CLSQL provides
routines for wall-time and
duration processing.
|
boolean - stores a T or
NIL value. |
generalized-boolean - similar
to a boolean in that either a
T or NIL value is stored in the SQL
database. However, any Lisp object can be stored in
the Lisp object. A Lisp value of NIL is stored as
FALSE in the database, any
other Lisp value is stored as
TRUE.
|
keyword - stores a keyword
|
symbol - stores a symbol |
list - stores a list by writing
it to a string. The items in the list must be able to
be readable written.
|
vector - stores a vector
similarly to list |
array - stores a array
similarly to list |
:column - specifies the name of
the SQL column which the slot maps onto, if
:db-kind is not
:virtual, and defaults to the
slot name. If the slot name is used for the SQL column
name, any hypens in the slot name are converted
to underscore characters.
:void-value - specifies the value
to store in the Lisp instance if the SQL value is NULL and
defaults to NIL.
:db-constraints - is a keyword
symbol representing an SQL column constraint expression or
a list of such symbols. The following column constraints
are supported: :not-null,
:primary-key, :unique,
:unsigned (MySQL specific),
:zerofill (MySQL specific) and
:auto-increment (MySQL specific).
:db-type - a string to specify the SQL
column type. If specified, this string overrides the SQL
column type as computed from the :type
slot value.
:db-reader - If a string, then when
reading values from the database, the string will be used
for a format string, with the only value being the value
from the database. The resulting string will be used as
the slot value. If a function then it will take one
argument, the value from the database, and return the
value that should be put into the slot. If a symbol, then
the symbol-function of the symbol will be used.
:db-writer - If a string, then when
reading values from the slot for the database, the string
will be used for a format string, with the only value
being the value of the slot. The resulting string will be
used as the column value in the database. If a function
then it will take one argument, the value of the slot, and
return the value that should be put into the database. If
a symbol, then the symbol-function of the symbol will be
used.
:base-table - specifies the name
of the SQL database table. The default value is the
class name. Like slot names, hypens in the class name
are converted to underscore characters.
:normalizedp - specifies whether
this class uses normalized inheritance from parent classes.
Defaults to nil, i.e. non-normalized schemas. When true,
SQL database tables that map to this class and parent
classes are joined on their primary keys to get the full
set of database columns for this class. This means that
the primary key of the base class will be copied to all
subclasses as we insert so that all parent classes of an
instance will have the same value in their primary key slots
(see tests/ds-nodes.lisp and oodml.lisp)
Creates a View
Class called name whose
slots slots can map onto the attributes
of a table in a database. If
superclasses is NIL then the
superclass of class will be
standard-db-object, otherwise
superclasses is a list of superclasses
for class which must include
standard-db-object or a descendent of
this class.
Specifying that :normalizedp is T tells CLSQL to normalize the database schema for inheritance. What this means is shown in the examples below.
With :normalizedp equal to NIL (the default) the class inheritance would result in the following:
(def-view-class node ()
((title :accessor title :initarg :title :type (varchar 240))))
SQL table NODE:
+-------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+--------------+------+-----+---------+-------+
| TITLE | varchar(240) | YES | | NULL | |
+-------+--------------+------+-----+---------+-------+
(def-view-class user (node)
((user-id :accessor user-id :initarg :user-id
:type integer :db-kind :key :db-constraints (:not-null))
(nick :accessor nick :initarg :nick :type (varchar 64))))
SQL table USER:
+---------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+---------+--------------+------+-----+---------+-------+
| USER_ID | int(11) | NO | PRI | | |
| NICK | varchar(64) | YES | | NULL | |
| TITLE | varchar(240) | YES | | NULL | |
+---------+--------------+------+-----+---------+-------+
Using :normalizedp T, both view-classes need a primary key to join them on:
(def-view-class node ()
((node-id :accessor node-id :initarg :node-id
:type integer :db-kind :key
:db-constraints (:not-null))
(title :accessor title :initarg :title :type (varchar 240))))
SQL table NODE:
+---------+--------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+---------+--------------+------+-----+---------+-------+
| NODE_ID | int(11) | NO | PRI | | |
| TITLE | varchar(240) | YES | | NULL | |
+---------+--------------+------+-----+---------+-------+
(def-view-class user (node)
((user-id :accessor user-id :initarg :user-id
:type integer :db-kind :key :db-constraints (:not-null))
(nick :accessor nick :initarg :nick :type (varchar 64)))
(:normalizedp t))
SQL table USER:
+---------+-------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+---------+-------------+------+-----+---------+-------+
| USER_ID | int(11) | NO | PRI | | |
| NICK | varchar(64) | YES | | NULL | |
+---------+-------------+------+-----+---------+-------+
In this second case, all slots of the view-class 'node are also available in view-class 'user, and can be used as one would expect. For example, with the above normalized view-classes 'node and 'user, and SQL tracing turned on:
CLSQL> (setq test-user (make-instance 'user :node-id 1 :nick "test-user"
:title "This is a test user"))
#<USER {1003B392E1}>
CLSQL> (update-records-from-instance test-user :database db)
;; .. => INSERT INTO NODE (NODE_ID,TITLE) VALUES (1,'This is a test user')
;; .. <= T
;; .. => INSERT INTO USER (USER_ID,NICK) VALUES (1,'test-user')
;; .. <= T
1
CLSQL> (node-id test-user)
1
CLSQL> (title test-user)
"This is a test user"
CLSQL> (nick test-user)
"test-user"
Notes from a refactor of this code. There are many assumptions that need to be met for normalized classes to work * The each of the classes should have its own single key column (of a different name) that will contain an identical value. EG: node has a node-id, setting which is a node has a node-id and a setting-id which must be equal. You cannot use node-id as the primary key on both tables (as I would have expected). The exception to this seems to be if your class has no slots at all, then you dont need to have a single key column, because your class is fully represented in the db by its parent(s) * more than one parent class per normalized class should be considered experimental and untested (vaya con Dios) * There are a few code paths that just dont pay any attention to normalized classes eg: delete-records-for-instance
The following examples are from the CLSQL test suite.
(def-view-class person (thing)
((height :db-kind :base :accessor height :type float
:initarg :height)
(married :db-kind :base :accessor married :type boolean
:initarg :married)
(birthday :type clsql:wall-time :initarg :birthday)
(bd-utime :type clsql:universal-time :initarg :bd-utime)
(hobby :db-kind :virtual :initarg :hobby :initform nil)))
(def-view-class employee (person)
((emplid
:db-kind :key
:db-constraints :not-null
:type integer
:initarg :emplid)
(groupid
:db-kind :key
:db-constraints :not-null
:type integer
:initarg :groupid)
(first-name
:accessor first-name
:type (varchar 30)
:initarg :first-name)
(last-name
:accessor last-name
:type (varchar 30)
:initarg :last-name)
(email
:accessor employee-email
:type (varchar 100)
:initarg :email)
(ecompanyid
:type integer
:initarg :companyid)
(company
:accessor employee-company
:db-kind :join
:db-info (:join-class company
:home-key ecompanyid
:foreign-key companyid
:set nil))
(managerid
:type integer
:initarg :managerid)
(manager
:accessor employee-manager
:db-kind :join
:db-info (:join-class employee
:home-key managerid
:foreign-key emplid
:set nil))
(addresses
:accessor employee-addresses
:db-kind :join
:db-info (:join-class employee-address
:home-key emplid
:foreign-key aemplid
:target-slot address
:set t)))
(:base-table employee))
(def-view-class company ()
((companyid
:db-kind :key
:db-constraints :not-null
:type integer
:initarg :companyid)
(groupid
:db-kind :key
:db-constraints :not-null
:type integer
:initarg :groupid)
(name
:type (varchar 100)
:initarg :name)
(presidentid
:type integer
:initarg :presidentid)
(president
:reader president
:db-kind :join
:db-info (:join-class employee
:home-key presidentid
:foreign-key emplid
:set nil))
(employees
:reader company-employees
:db-kind :join
:db-info (:join-class employee
:home-key (companyid groupid)
:foreign-key (ecompanyid groupid)
:set t))))
(def-view-class address ()
((addressid
:db-kind :key
:db-constraints :not-null
:type integer
:initarg :addressid)
(street-number
:type integer
:initarg :street-number)
(street-name
:type (varchar 30)
:void-value ""
:initarg :street-name)
(city
:column "city_field"
:void-value "no city"
:type (varchar 30)
:initarg :city)
(postal-code
:column zip
:type integer
:void-value 0
:initarg :postal-code))
(:base-table addr))
;; many employees can reside at many addressess
(def-view-class employee-address ()
((aemplid :type integer :initarg :emplid)
(aaddressid :type integer :initarg :addressid)
(verified :type boolean :initarg :verified)
(address :db-kind :join
:db-info (:join-class address
:home-key aaddressid
:foreign-key addressid
:retrieval :immediate)))
(:base-table "ea_join"))
(def-view-class deferred-employee-address ()
((aemplid :type integer :initarg :emplid)
(aaddressid :type integer :initarg :addressid)
(verified :type boolean :initarg :verified)
(address :db-kind :join
:db-info (:join-class address
:home-key aaddressid
:foreign-key addressid
:retrieval :deferred
:set nil)))
(:base-table "ea_join"))
The actual SQL type for a column depends up the database type
in which the SQL table is stored. As an example, the view
class type (varchar 100) specifies a
SQL column type VARCHAR(100) in MySQL
and a column type VARCHAR2(100) in
Oracle
The actual lisp type for a slot may be different than the
value specified by the :type attribute.
For example, a slot declared with ":type (string
30)" actually sets the slots Lisp type as
(or null string). This is to allow a
NIL value or a string shorter than 30 characters to be
stored in the slot.