NAME

Yancy::Backend - Interface to a database

VERSION

version 1.088

SYNOPSIS

my $be = Yancy::Backend->new( $url );

$result = $be->list( $schema, $where, $options );
say "Total: " . $result->{total};
say "Name: " . $_->{name} for @{ $result->{items} };

$item = $be->get( $schema, $id );
$be->set( $schema, $id, $item );
$be->delete( $schema, $id );
$id = $be->create( $schema, $item );

DESCRIPTION

A Yancy::Backend handles talking to the database. Different Yancy backends will support different databases. To use a backend, see "SUPPORTED BACKENDS". To make your own backend, see "METHODS" for the list of methods each backend supports, their arguments, and their return values.

Terminology

Yancy backends work with schemas, which are made up of items. A schema is a set of items, like a database table. An item is a single element of a schema, and must be a hashref.

Asynchronous Backends

Asynchronous backends implement both a synchronous and an asynchronous API (using promises).

Synchronous-only Backends

Synchronous-only backends also implement a promises API for compatibility, but will not perform requests concurrently.

SUPPORTED BACKENDS

Other backends are available on CPAN.

EXTENDING

To create your own Yancy::Backend for a new database system, inherit from this class and provide the standard six interface methods: "get", "create", "set", "list", "delete", and "read_schema".

There are roles to aid backend development:

Backends do not have to talk to databases. For an example, see Yancy::Backend::Static for a backend that uses plain files like a static site generator.

METHODS

new

my $url = 'memory://custom_string';
my $be = Yancy::Backend::Memory->new( $url, $schema );

Create a new backend object. $url is a string that begins with the backend name followed by a colon. Everything else in the URL is for the backend to use to describe how to connect to the underlying database and any options for the backend object itself.

The backend name will be run through ucfirst before being looked up in Yancy::Backend::. For example, mysql://... will use the Yancy::Backend::Mysql module.

$schema is a hash reference of schema configuration from the Yancy configuration. See Yancy::Guides::Schema for more information.

list

my $result = $be->list( $schema, $where, %opt );
# { total => ..., items => [ ... ] }

Fetch a list of items from a schema. $schema is the schema name.

$where is a SQL::Abstract where structure.

# Search for all Dougs
$be->list( 'people', { name => { -like => 'Doug%' } } );
# Find adults
$be->list( 'people', { age => { '>=' => 18 } } );
# Find men we can contact
$be->list( 'people', { gender => 'male', contact => 1 } );

Additionally, Yancy backends support the following additional keys in the where structure:

-has (EXPERIMENTAL)

The -has operator searches inside a data structure (an array or a hash). This operator examines the type of the field being searched to perform the appropriate query.

# Create a new page with an array of tags and a hash of author
# information
$be->create( pages => {
    title => 'Release v1.481',
    tags => [ 'release', 'minor' ],
    author => {
        name => 'Doug Bell',
        email => 'doug@example.com',
    },
} );

# All pages that have the tag "release"
$be->list( pages => { tags => { -has => 'release' } } );

# All pages that have both the tags "release" and "major"
$be->list( pages => { tags => { -has => [ 'release', 'major' ] } } );

# All pages that have the author's name starting with Doug
$be->list( pages => { author => { -has => { name => { -like => 'Doug%' } } } } );

This is not yet supported by all backends, and may never be supported by some. Postgres has array columns and JSON fields. MySQL has JSON fields. The "match" in Yancy::Util function matches against Perl data structures. All of these should support -has and -not_has before it can be considered not experimental.

%opt is a list of name/value pairs with the following keys:

  • limit - The number of items to return

  • offset - The number of items to skip

  • order_by - A SQL::Abstract order by clause

  • join - Join one or more tables using a x-foreign-key field. This can be the name of a foreign key field on this schema, or the name of a table with a foreign key field that refers to this schema. Join multiple tables at the same time by passing an arrayref of joins. Fields in joined tables can be queried by prefixing the join name to the field, separated by a dot.

# Get the second page of 20 people
$be->list( 'people', {}, limit => 20, offset => 20 );
# Get the list of people sorted by age, oldest first
$be->list( 'people', {}, order_by => { -desc => 'age' } );
# Get the list of people sorted by age first, then name (ascending)
$be->list( 'people', {}, order_by => [ 'age', 'name' ] );

Returns a hashref with two keys:

items

An array reference of hash references of item data

total

The total count of items that would be returned without limit or offset.

list_p

my $promise = $be->list_p( $schema, $where, %opt );
$promise->then( sub {
    my ( $result ) = @_;
    # { total => ..., items => [ ... ] }
} );

Fetch a list of items asynchronously using promises. Returns a promise that resolves to a hashref with items and total keys. See "list" for arguments and return values.

get

my $item = $be->get( $schema, $id, %opts );

Get a single item. $schema is the schema name. $id is the ID of the item to get: Either a string for a single key field, or a hash reference for a composite key. Returns a hashref of item data.

%opts is a list of name/value pairs of options with the following names:

join

Join one or more tables using a x-foreign-key field. This should be the name of the schema to join, either by a foreign key on this table, or a foreign key on the joined table. Join multiple tables at the same time by passing an arrayref of joins.

get_p

my $promise = $be->get_p( $schema, $id );
$promise->then( sub {
    my ( $item ) = @_;
    # ...
} );

Get a single item asynchronously using promises. Returns a promise that resolves to the item. See "get" for arguments and return values.

set

my $success = $be->set( $schema, $id, $item );

Update an item. $schema is the schema name. $id is the ID of the item to update: Either a string for a single key field, or a hash reference for a composite key. $item is the item's data to set. Returns a boolean that is true if a row with the given ID was found and updated, false otherwise.

Currently the values of the data cannot be references, only simple scalars or JSON booleans.

set_p

my $promise = $be->set_p( $schema, $id );
$promise->then( sub {
    my ( $success ) = @_;
    # ...
} );

Update a single item asynchronously using promises. Returns a promise that resolves to a boolean indicating if the row was updated. See "set" for arguments and return values.

create

my $id = $be->create( $schema, $item );

Create a new item. $schema is the schema name. $item is the item's data. Returns the ID of the row created suitable to be passed in to the get() method|/get.

Currently the values of the data cannot be references, only simple scalars or JSON booleans.

create_p

my $promise = $be->create_p( $schema, $item );
$promise->then( sub {
    my ( $id ) = @_;
    # ...
} );

Create a new item asynchronously using promises. Returns a promise that resolves to the ID of the newly-created item. See "create" for arguments and return values.

delete

$be->delete( $schema, $id );

Delete an item. $schema is the schema name. $id is the ID of the item to delete: Either a string for a single key field, or a hash reference for a composite key. Returns a boolean that is true if a row with the given ID was found and deleted. False otherwise.

delete_p

my $promise = $be->delete_p( $schema, $id );
$promise->then( sub {
    my ( $success ) = @_;
    # ...
} );

Delete an item asynchronously using promises. Returns a promise that resolves to a boolean indicating if the row was deleted. See "delete" for arguments and return values.

query

Run a raw query on the backend. Each backend may have its own arguments. Returns a list of hash references of data.

query_p

Run a raw query on the backend. Each backend may have its own arguments. Returns a promise that resolves into a list of hash references of data.

read_schema

my $schema = $be->read_schema;
my $table = $be->read_schema( $table_name );

Read the schema from the database tables. Returns an OpenAPI schema ready to be merged into the user's configuration. Can be restricted to only a single table.

INTERNAL METHODS

These methods are documented for use in subclasses and should not need to be called externally.

supports

Returns true if the backend supports a given feature. Returns false for now. In the future, features like 'json' will be detectable.

ignore_table

Returns true if the given table should be ignored when doing "read_schema". By default, backends will ignore tables used by:

normalize

This method normalizes data to and from the database.

id_field

Get the ID field for the given schema. Defaults to id.

id_where

Get the query structure for the ID field of the given schema with the given ID value.

AUTHOR

Doug Bell <preaction@cpan.org>

COPYRIGHT AND LICENSE

This software is copyright (c) 2021 by Doug Bell.

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.

1 POD Error

The following errors were encountered while parsing the POD:

Around line 778:

You forgot a '=back' before '=head2'