Alex Vandiver

NAME

Jifty::DBI::Record - Superclass for records loaded by Jifty::DBI::Collection

SYNOPSIS

  package MyRecord;
  use base qw/Jifty::DBI::Record/;

DESCRIPTION

Jifty::DBI::Record encapsulates records and tables as part of the Jifty::DBI object-relational mapper.

METHODS

new ARGS

Instantiate a new, empty record object.

ARGS is a hash used to pass parameters to the _init() function.

Unless it is overloaded, the _init() function expects one key of 'handle' with a value containing a reference to a Jifty::DBI::Handle object.

id

Returns this row's primary key.

primary_keys

Return a hash of the values of our primary keys for this function.

_accessible COLUMN ATTRIBUTE

Private method.

DEPRECATED

Returns undef unless COLUMN has a true value for ATTRIBUTE.

Otherwise returns COLUMN's value for that attribute.

_primary_keys

Return our primary keys. (Subclasses should override this, but our default is that we have one primary key, named 'id'.)

_init_columns

Sets up the primary key columns.

_init_methods_for_columns

This is an internal method responsible for calling "_init_methods_for_column" for each column that has been configured.

schema_version

If present, this method must return a string in '1.2.3' format to be used to determine which columns are currently active in the schema. That is, this value is used to determine which columns are defined, based upon comparison to values set in till and since.

If no implementation is present, the "latest" schema version is assumed, meaning that any column defining a till is not active and all others are.

_init_methods_for_column COLUMN

This method is used internally to update the symbol table for the record class to include an accessor and mutator for each column based upon the column's name.

In addition, if your record class defines the method "schema_version", it will automatically generate methods according to whether the column currently exists for the current application schema version returned by that method. The schema_version method must return a value in the same form used by since and till.

If the column doesn't currently exist, it will create the methods, but they will die with an error message stating that the column does not exist for the current version of the application. If it does exist, a normal accessor and mutator will be created.

See also "active" in Jifty::DBI::Column, "since" in Jifty::DBI::Schema, "till" in Jifty::DBI::Schema for more information.

null_reference

By default, Jifty::DBI::Record will return undef for non-existent foreign references which don't exist. That is, if each Employee refers_to a Department, but isn't required to, <$model-department>> will return undef for employees not in a department.

Overriding this method to return 0 will cause it to return a record with no id. That is, <$model-department>> will return a Department object, but <$model-department->id>> will be undef.

_to_record COLUMN VALUE

This PRIVATE method takes a column name and a value for that column.

It returns undef unless COLUMN is a valid column for this record that refers to another record class.

If it is valid, this method returns a new record object with an id of VALUE.

prefetched NAME

Returns the prefetched value for column of property NAME, if it exists.

add_column

column

    my $column = $self->column($column_name);

Returns the Jifty::DBI::Column object of the specified column name.

columns

    my @columns = $record->columns;

Returns a sorted list of a $record's @columns.

all_columns

  my @all_columns = $record->all_columns;

Returns all the columns for the table, even those that are inactive.

readable_attributes

Returns the list of this table's readable columns. They are first sorted so that primary keys come first, and then they are sorted in alphabetical order.

serialize_metadata

Returns a hash which describes how this class is stored in the database. Right now, the keys are class, table, and columns. class and table return simple scalars, but columns returns a hash of name =&gt; value pairs for all the columns in this model. See Jifty::DBI::Column/serialize_metadata for the format of that hash.

writable_attributes

Returns a list of this table's writable columns

record values

As you've probably already noticed, Jifty::DBI::Record automatically creates methods for your standard get/set accessors. It also provides you with some hooks to massage the values being loaded or stored.

When you fetch a record value by calling $my_record->some_field, Jifty::DBI::Record provides the following hook

after_column_name

This hook is called with a reference to the value returned by Jifty::DBI. Its return value is discarded.

When you set a value, Jifty::DBI provides the following hooks

before_set_column_name PARAMHASH

Jifty::DBI::Record passes this function a reference to a paramhash composed of:

column

The name of the column we're updating.

value

The new value for column.

is_sql_function

A boolean that, if true, indicates that value is an SQL function, not just a value.

If before_set_column_name returns false, the new value isn't set.

before_set PARAMHASH

This is identical to the before_set_column_name, but is called for every column set.

after_set_column_name PARAMHASH

This hook will be called after a value is successfully set in the database. It will be called with a reference to a paramhash that contains column, value, and old_value keys. If value was a SQL function, it will now contain the actual value that was set. If column has filters on it, value will be the result of going through an encode and decode cycle.

This hook's return value is ignored.

after_set PARAMHASH

This is identical to the after_set_column_name, but is called for every column set.

validate_column_name VALUE

This hook is called just before updating the database. It expects the actual new value you're trying to set column_name to. It returns two values. The first is a boolean with truth indicating success. The second is an optional message. Note that validate_column_name may be called outside the context of a set operation to validate a potential value. (The Jifty application framework uses this as part of its AJAX validation system.)

_value

_value takes a single column name and returns that column's value for this row. Subclasses can override _value to insert custom access control.

__raw_value

Takes a column name and returns that column's raw value. Subclasses should never override __raw_value.

resolve_column

given a column name, resolve it, even if it's actually an alias return the column object.

__value

Takes a column name and returns that column's value. Subclasses should never override __value.

as_hash

Returns a version of this record's readable columns rendered as a hash of key => value pairs

_set

_set takes a single column name and a single unquoted value. It updates both the in-memory value of this column and the in-database copy. Subclasses can override _set to insert custom access control.

load

load can be called as a class or object method.

Takes a single argument, $id. Calls load_by_cols to retrieve the row whose primary key is $id.

load_by_cols

load_by_cols can be called as a class or object method.

Takes a hash of columns and values. Loads the first record that matches all keys.

The hash's keys are the columns to look at.

The hash's values are either: scalar values to look for OR hash references which contain 'operator', 'value', 'case_sensitive' or 'function'

To load something case sensitively on a case insensitive database, you can do:

  $record->load_by_cols( column => { operator => '=',
                                     value => 'Foo',
                                     case_sensitive => 1 } );

load_by_primary_keys

Loads records with a given set of primary keys.

load_from_hash

Takes a hashref, such as created by Jifty::DBI and populates this record's loaded values hash.

_load_from_sql QUERYSTRING @BIND_VALUES

Load a record as the result of an SQL statement

create PARAMHASH

create can be called as either a class or object method

This method creates a new record with the values specified in the PARAMHASH.

This method calls two hooks in your subclass:

before_create

When adding the before_create trigger, you can determine whether the trigger may cause an abort or not by passing the abortable parameter to the add_trigger method. If this is not set, then the return value is ignored regardless.

  sub before_create {
      my $self = shift;
      my $args = shift;

      # Do any checks and changes on $args here.
      $args->{first_name} = ucfirst $args->{first_name};

      return;      # false return vallue will abort the create
      return 1;    # true return value will allow create to continue
  }

This method is called before trying to create our row in the database. It's handed a reference to your paramhash. (That means it can modify your parameters on the fly). before_create returns a true or false value. If it returns undef and the trigger has been added as abortable, the create is aborted.

after_create

When adding the after_create trigger, you can determine whether the trigger may cause an abort or not by passing the abortable parameter to the add_trigger method. If this is not set, then the return value is ignored regardless.

  sub after_create {
      my $self                    = shift;
      my $insert_return_value_ref = shift;

      return unless $$insert_return_value_ref;    # bail if insert failed
      $self->load($$insert_return_value_ref);     # load ourselves from db

      # Do whatever needs to be done here

      return;   # aborts the create, possibly preventing a load
      return 1; # continue normally
  }

This method is called after attempting to insert the record into the database. It gets handed a reference to the return value of the insert. That will either be a true value or a Class::ReturnValue.

Aborting the trigger merely causes create to return a false (undefined) value even thought he create may have succeeded. This prevents the loading of the record that would normally be returned.

delete

Delete this record from the database. On failure return a Class::ReturnValue with the error. On success, return 1;

This method has two hooks:

before_delete

This method is called before the record deletion, if it exists. On failure it returns a Class::ReturnValue with the error. On success it returns 1.

If this method returns an error, it causes the delete to abort and return the return value from this hook.

after_delete

This method is called after deletion, with a reference to the return value from the delete operation.

table

This method returns this class's default table name. It uses Lingua::EN::Inflect to pluralize the class's name as we believe that class names for records should be in the singular and table names should be plural.

If your class name is My::App::Rhino, your table name will default to rhinos. If your class name is My::App::RhinoOctopus, your default table name will be rhino_octopuses. Not perfect, but arguably correct.

collection_class

Returns the collection class which this record belongs to; override this to subclass. If you haven't specified a collection class, this returns a best guess at the name of the collection class for this collection.

It uses a simple heuristic to determine the collection class name -- It appends "Collection" to its own name. If you want to name your records and collections differently, go right ahead, but don't say we didn't warn you.

_guess_table_name

Guesses a table name based on the class's last part.

_handle

Returns or sets the current Jifty::DBI::Handle object

PRIVATE refers_to

used for the declarative syntax

is_distinct COLUMN_NAME, VALUE

Checks to see if there is already a record in the database where COLUMN_NAME equals VALUE. If no such record exists then the COLUMN_NAME and VALUE pair is considered distinct and it returns 1. If a value is already present the test is considered to have failed and it returns a Class::ReturnValue with the error.

run_canonicalization_for_column column => 'COLUMN', value => 'VALUE'

Runs all canonicalizers for the specified column.

has_canonicalizer_for_column COLUMN

Returns true if COLUMN has a canonicalizer, otherwise returns undef.

run_validation_for_column column => 'COLUMN', value => 'VALUE' [extra => \@ARGS]

Runs all validators for the specified column.

has_validator_for_column COLUMN

Returns true if COLUMN has a validator, otherwise returns undef.

unload_value COLUMN

Purges the cached value of COLUMN from the object, forcing it to be fetched from the database next time it is queried.

AUTHOR

Jesse Vincent <jesse@bestpractical.com>, Alex Vandiver <alexmv@bestpractical.com>, David Glasser <glasser@bestpractical.com>, Ruslan Zakirov <ruslan.zakirov@gmail.com>

Based on DBIx::SearchBuilder::Record, whose credits read:

 Jesse Vincent, <jesse@fsck.com> 
 Enhancements by Ivan Kohler, <ivan-rt@420.am>
 Docs by Matt Knopp <mhat@netlag.com>

SEE ALSO

Jifty::DBI, Jifty::DBI::Handle, Jifty::DBI::Collection.




Hosting generously
sponsored by Bytemark