2 Automated Data Class Generation
A powerful feature of Racquel is the ability to generate data-class mappings automatically
using database schema metadata. This allows for data classes to be defined for all the tables in a
database without the tedious effort of manually coding the mappings.
| (gen-data-class db-connection | | table-name | | #:db-system-type db-system-type-kw | | #:generate-joins? generate-joins-kw | | #:generate-reverse-joins? generate-reverse-joins-kw | | #:schema-name schema-name | | #:inherits base-class | | #:table-name-normalizer proc | | #:column-name-normalizer proc | | #:join-name-normalizer proc | | #:table-name-externalizer proc | | #:print? print-kw | | #:prepare? prepare-kw | | data-class-clause ...) |
|
| |
| db-system-type-kw | | = | | | | | | | | | #:db-system-type db-system-type | | | | | | | | generate-joins-kw | | = | | | | | | | | | #:generate-joins? (or/c #t #f) | | | | | | | | generate-reverse-joins-kw | | = | | | | | | | | | #:generate-reverse-joins? (or/c #t #f) | | | | | | | | schema-name-kw | | = | | | | | | | | | #:schema-name (string?) | | | | | | | | inherits-kw | | = | | | | | | | | | #:inherits (string?) | | | | | | | | table-name-normalizer-kw | | = | | | | | | | | | #:table-name-normalizer | | | | | | | | column-name-normalizer-kw | | = | | | | | | | | | #:column-name-normalizer | | | | | | | | join-name-normalizer-kw | | = | | | | | | | | | #:join-name-normalizer | | | | | | | | table-name-externalizer-kw | | = | | | | | | | | | #:table-name-externalizer | | | | | | | | print-kw | | = | | | | | | | | | #:print? (or/c #t #f) |
|
Generates a data class from the specified table-name using the db-connection. If
the database system type is an ODBC connection, then the particular system type can be specified
using the #:db-system-type keyword. (This is not necessary if the database system type has
already been defined by calling set-odbc-dbsystem-type!. If the #:db-system-type is
not specified, then the database system type is determined from the dbsystem-name of the
db-connection, where the 'odbc database system is assumed to be SQL Server.
The generate-joins? and generate-reverse-joins keywords control the
automatic generation of joins and reverse joins. Joins are determined based on foreign key
constraints in the database, and reverse joins are based on any foreign key contraints referencing
it. By default, joins are generated, but reverse joins are not.
In some cases, the table name may not be unique in the database server’s metadata. In those cases,
the schema-name can be used to specify the schema that the table resides in.
The base class of the generated data class can be specified using the #:inherits keyword.
The default base-class is object%.
Behavior of the map generation can be controlled by customizing the normalizer procedures which
normalize database names into Racket symbols. Generation of external names from the symbol can also
be controlled using the externalizer procedures. A normalizer or externalizer can be customized by
specifying the procedure to use with the appropriate keyword. An example of a situation where one
would want to override the default normalizers is if table names in a database started with a prefix.
Below are the default normalizers and externalizers.
This normalizer converts database table names into Racket class names, using a set of rules. First,
the normalizer will convert mixed-case names, e.g. "MixedCase", and make the all lower-case with
hyphens between the names, e.g. "mixed-case". It will then convert any underscores to hyphens.
Finally, it will append a percent sign to the end of the name, since that is the Racket standard for
naming classes.
| > (table-name-normalizer "ExampleTable_Name") |
| example-table-name% |
This is default normalizer for table names if the #:table-name-externalizer keyword is not
specified.
This converts column names of a table into Racket symbols, following a set of rules. The
rules are similar to those for the table-name-normalizer. First mixed-case names are
converted to lower-case with hyphens, then underscores are converted to hyphens.
| > (column-name-normalizer "ExampleColumn_Name") |
| example-column-name |
This is default normalizer for table names if the #:column-name-externalizer keyword is not
specified.
This converts joined table names into Racket symbols, following a set of rules. The
rules are similar to those for the column-name-normalizer. First mixed-case names are
converted to lower-case with hyphens, then underscores are converted to hyphens. Also, if the
cardinality of the join is 'one-to-many, an "s" is appended to the end of the name (or "es"
if the name ends with an "s".)
| > (join-name-normalizer "JoinExample_Address") |
| join-example-addresses |
This is default normalizer for table names if the #:join-name-externalizer keyword is not
specified.
Converts a database table name to an external name for JSON or XML serialization. The default is no
change at all to the table name.
(set-odbc-dbsystem-type! odbc-sys-type) → (void?)
|
| odbc-sys-type : (or/c 'sqlserver 'oracle 'db2) |
If the database system being used is either Oracle or DB/2, then the database system type needs to be
set to distinguish the ODBC connection from a SQL Server connection, which is assumed for an ODBC
connection is it is not specified using this procedure.