Patio

Patio is a Sequel inspired query engine.

Installation

To install patio run

npm install comb patio

If you want to use the patio executable for migrations

npm install -g patio

Running Tests

To run the tests

grunt test

To run just the postgres tests

grunt test-pg

To run just the mysql tests

grunt test-mysql

Running Tests with Docker

In order to provide a consistent test environment and make it easier to test, we have included a Dockerfile and a docker-compose.yml to make it easy to test in an isolated environment. You can do so with:

docker-compose build
docker-compose up -d mysql postgres
sleep 10 # Wait for databases to come up
docker-compose up patio

Why Use Patio?

Patio is different because it allows the developers to choose the level of abtraction they are comfortable with.

If you want to use the ORM functionality you can. If you don't you can just use the Database and Datasets as a querying API, and if you need to you can write plain SQL

Concepts

1. Model definitions are defined by the tables in the database.

   As you add models the definition is automatically defined from the table definition. This is particularly useful when you want to define your model from a schema designed using another tool (i.e. ActiveRecord, Sequel, etc...)

2. Patio tries to stay out of your way when querying.

   When you define a model you still have the freedom to do any type of query you want.

   Only want certain columns?

   MyModel.select("id", "name", "created").forEach(function(record){
       //record only has the id, name, and created columns
   });

   You want to join with another table?

   MyModel.join("otherTable", {id: patio.sql.identifier("myModelId"}).forEach(function(record){
       //Record has columns from your join table now!
   });

   You want to run raw SQL?

   MyModel.db.run("select * from my_model where name = 'Bob'").all().chain(function(records){
       //all records with a name that equals bob.
   });

   You want to just query the database and not use a model?

   var DB = patio.connect("pg://test:[email protected]:5432/test_db");
   DB.from("myTable").filter({id: [1,2,3]}).all().function(records){
      //records with id IN (1,2,3)
   });

Getting Started

All the code for this example can be found here

1. Create a new database

   PostgreSQL

   psql -c "CREATE DATABASE reademe_example"
   

   MySQL

   mysql -e "CREATE DATABASE readme_example"
   

2. Create a migration

   mkdir migration
   patio migration-file -n createInitialTables ./migration
   

   This will add a migration name createdInitialTables in your migration directory.

3. Add the following code to your migration

   module.exports = {
       //up is called when you migrate your database up
       up: function (db) {
           //create a table called state;
           return db
               .createTable("state", function () {
                   this.primaryKey("id");
                   this.name(String);
                   this.population("integer");
                   this.founded(Date);
                   this.climate(String);
                   this.description("text");
               })
               .chain(function () {
                   //create another table called capital
                   return db.createTable("capital", function () {
                       this.primaryKey("id");
                       this.population("integer");
                       this.name(String);
                       this.founded(Date);
                       this.foreignKey("stateId", "state", {key: "id", onDelete: "CASCADE"});
                   });
               });
       },
   
       //down is called when you migrate your database down
       down: function (db) {
           //drop the state and capital tables
           return db.dropTable("capital", "state");
       }
   };

4. Run your migration

   patio migrate -v --camelize -u "<DB_CONNECTION_STRING>" -d ./migration
   

5. Connect and query!

   var patio = require("patio");
   
   //set camelize = true if you want snakecase database columns as camelcase
   patio.camelize = true;
   patio.connect("pg://[email protected]:5432/readme_example");
   
   //define a State model with a relationship to capital
   var State = patio.addModel("state").oneToOne("capital");
   
   //define a Capital model with a relationship to State
   var Capital = patio.addModel("capital").manyToOne("state");
   
   //save a state
   State
       .save({
           name: "Nebraska",
           population: 1796619,
           founded: new Date(1867, 2, 4),
           climate: "continental",
           //notice the capital relationship is inline
           capital: {
               name: "Lincoln",
               founded: new Date(1856, 0, 1),
               population: 258379
           }
       })
       .chain(function () {
           //save a Capital
           return Capital.save({
               name: "Austin",
               founded: new Date(1835, 0, 1),
               population: 790390,
               //define the state inline
               state: {
                   name: "Texas",
                   population: 25674681,
                   founded: new Date(1845, 11, 29)
               }
           });
       })
       .chain(function () {
           //Query all the states by name
           return State.order("name").forEach(function (state) {
               //Get the associated capital
               return state.capital.chain(function (capital) {
                   console.log("%s's capital is %s.", state.name, capital.name);
               });
           });
       })
       .chain(process.exit, function (err) {
           console.log(err)
           process.exit(1);
       });

Guides

• Migrations
• Models
  • Associations
  • Inheritance
  • Validation
  • Plugins
• Querying
• DDL
• Logging

Features

• Comprehensive documentation with examples.
• > 80% test coverage
• Support for connection URIs
• Supported Databases
  • MySQL
  • Postgres
  • Redshift
• Models
  • Associations
  • Inheritance
  • Validation
  • Plugins
• Simple adapter extensions
• Migrations
  • Integer and Timestamp based.
• Powerful Querying API
• Transactions with
  • Savepoints
  • Isolation Levels
  • Two phase commits
• SQL Datatype casting
• Full database CRUD operations
  • createTable
  • alterTable
  • dropTable
  • insert
  • multiInsert
  • update
  • remove
  • query