Database wrapper for Crystal. Inspired by Ecto for Elixir language.
See documentation on http://docs.crecto.com
Add this to your application's shard.yml
:
dependencies:
crecto:
github: fridgerator/crecto
Include a database adapter:
Include crystal-pg in your project before crecto
in your application:
require "pg"
require "crecto"
Include crystal-mysql in your project before crecto
in your application:
require "mysql"
require "crecto"
Include crystal-sqlite3 in your project before crecto
in your appplication:
require "sqlite3"
require "crecto"
Micrate is recommended. It is used and supported by core crystal members.
First create a Repo. The Repo maps to the datastore and the database adapter and is used to run queries. You can even create multiple repos if you need to access multiple databases.
Note: For those coming from Active Record: Repo provides a level of abstraction between database logic (Repo) and business logic (Model).
Let's create a repo for Postgres:
module Repo
extend Crecto::Repo
config do |conf|
conf.adapter = Crecto::Adapters::Postgres # or Crecto::Adapters::Mysql or Crecto::Adapters::SQLite3
conf.database = "database_name"
conf.hostname = "localhost"
conf.username = "user"
conf.password = "password"
conf.port = 5432
# you can also set initial_pool_size, max_pool_size, max_idle_pool_size,
# checkout_timeout, retry_attempts, and retry_delay
end
end
And another for SQLite:
module SqliteRepo
extend Crecto::Repo
config do |conf|
conf.adapter = Crecto::Adapters::SQLite3
conf.database = "./path/to/database.db"
end
end
Optionally you can use constants shorcuts using:
Query = Crecto::Repo::Query
Multi = Crecto::Multi
Define table name, fields, validations, and constraints in your model
Defining a new class using Crecto::Model
:
class User < Crecto::Model
schema "users" do
field :age, Int32 # or use `PkeyValue` alias: `field :age, PkeyValue`
field :name, String
field :is_admin, Bool, default: false
field :temporary_info, Float64, virtual: true
field :email, String
has_many :posts, Post, dependent: :destroy
end
validate_required [:name, :age]
validate_format :name, /^[a-zA-Z]*$/
unique_constraint :email
end
Defining another one:
class Post < Crecto::Model
schema "posts" do
belongs_to :user, User
end
end
Creating a new User
:
user = User.new
user.name = "123"
user.age = 123
changeset = User.changeset(user)
puts changeset.valid? # false
puts changeset.errors # {:field => "name", :message => "is invalid"}
puts changeset.changes # {:name => "123", :age => 123}
user.name = "test"
changeset = User.changeset(user)
changeset.valid? # true
Repo returns a new changeset.
changeset = Repo.insert(user)
puts changeset.errors # []
user.name = "new name"
changeset = Repo.update(user)
puts changeset.instance.name # "new name"
post = Post.new
post.user = user
Repo.insert(post)
query = Query
.where(name: "new name")
.where("users.age < ?", [124])
.order_by("users.name ASC")
.order_by("users.age DESC")
.limit(1)
users = Repo.all(User, query)
users.as(Array) unless users.nil?
user = Repo.get(User, 1)
user.as(User) unless user.nil?
Repo.get_by(User, name: "new name", id: 1121)
user.as(User) unless user.nil?
changeset = Repo.delete(user)
user = Repo.get(User, id).as(User)
posts = Repo.get_association(user, :posts)
post = Repo.get(Post, id).as(Post)
user = Repo.get_association(post, :user)
users = Repo.all(User, preload: [:posts])
users[0].posts # has_many relation is preloaded
posts = Repo.all(Post, preload: [:user])
posts[0].user # belongs_to relation preloaded
If an association is not loaded, the normal accessor will raise an error.
users = Repo.all(User)
users[0].posts? # => nil
users[0].posts # raises Crecto::AssociationNotLoaded
For has_many
preloads, the result will always be an array.
users = Repo.all(User, preload: [:posts])
users[0].posts? # => Array(Post)
users[0].posts # => Array(Post)
For belongs_to and has_one preloads, the result may still be nil if no
record exists. If the association is nullable, always use association?
.
post = Repo.insert(Post.new).instance
post = Repo.get(Post, post.id, preload: [:user])
post.user? # nil
post.user # raises Crecto::AssociationNotLoaded
Can use the following aggregate functions: :avg
, :count
, :max
, :min:
, :sum
Repo.aggregate(User, :count, :id)
Repo.aggregate(User, :avg, :age, Query.where(name: 'Bill'))
Create the multi instance
multi = Multi.new
multi.insert(insert_user)
multi.delete(post)
multi.delete_all(Comment)
multi.update(update_user)
multi.update_all(User, Query.where(name: "stan"), {name: "stan the man"})
multi.insert(new_user)
Repo.transaction(multi)
If there are any errors in any of the transactions, the database will rollback as if none of the transactions happened
multi.errors.any?
If you wish to access attributes of a model without having to check for nil, in the case that you are using a NOT NULL
database constraint you can use the non-nillable attribute accessors. CAUTION: Mis-use of this could lead to Nil reference runtime errors.
user.name!
user.age!
(Postgres only)
class UserJson < Crecto::Model
schema "users_json" do
field :settings, Json
end
end
user = UserJson.new
user.settings = {"one" => "test", "two" => 123, "three" => 12321319323298}
Repo.insert(user)
query = Query.where("settings @> '{\"one\":\"test\"}'")
users = Repo.all(UserJson, query)
(Postgres only)
class UserArray < Crecto::Model
schema "users_array" do
field :string_array, Array(String)
field :int_array, Array(Int32)
field :float_array, Array(Float64)
field :bool_array, Array(Bool)
end
end
user = UserArray.new
user.string_array = ["one", "two", "three"]
Repo.insert(user)
query = Query.where("? = ANY(string_array)", "one")
users = Repo.all(UserArray, query)
By default nothing is logged. To enable pass any type of IO to the logger. For STDOUT
use:
Crecto::DbLogger.set_handler(STDOUT)
f = File.open("database.log", "w")
f.sync = true
Crecto::DbLogger.set_handler(f)
- Fork it ( https://github.com/fridgerator/crecto/fork )
- Create your feature branch (git checkout -b my-new-feature)
- Commit your changes (git commit -am 'Add some feature')
- Push to the branch (git push origin my-new-feature)
- Create a new Pull Request
When developing against crecto, the database must exist prior to
testing. There are migrations for each database type in spec/migrations
,
and references on how to migrate then in the .travis.yml
file.
Create a new file spec/repo.cr
and create a module name Repo
to use for testing.
There are example repos for each database type in the spec folder: travis_pg_repo.cr
,
travis_mysql_repo.cr
, and travis_sqlite_repo.cr
When submitting a pull request, please test against all 3 databases.