- Automatically discovers schema and replication topology
- Zero boilerplate
- Exposes CRUD methods
- Saves nested objects
- Object caching & query memoization
- PostgreSQL 9+
npm install oreo
npm install pg
† see the example database schema below
var oreo = require('oreo')
// discover schema and replication topology
var db = oreo({
driver: 'pg',
hosts: ['localhost:5432'],
name: 'my_db',
user: 'postgres',
pass: 'password'
}, runExampleQueries)
function runExampleQueries(err) {
// Insert a new book and its author
db.books.insert({
title: 'Fear and Loathing in Las Vegas',
author: {
name: 'Hunter S.Thompson'
}
}, function(err, book) {
console.log(book) // { id: 1, title: Fear and Loathing in Las Vegas, author_id: 1 }
// Get a linked object
book.hydrate('author', function(err, author) {
console.log(book.author) // { id: 1, name: Hunter S. Thompson, books: [] }
// Get multiple books using array of primary keys
db.books.mget(author.books, function(err, books) {
console.log(books)
})
})
// Get an author by primary key
db.authors.get(1, function(err, author) {
console.log(author)
})
// Find authors by criteria
db.authors.find({
where: {
author_id: 1
}
}, function(err, authors) {
console.log(authors) // [{ id: 1, name: Hunter S. Thompson, books: [] }]
})
// Update the book
book.update({
title: 'The Rum Diary'
}, function(err, book) {
console.log(book) // { id: 1, title: The Rum Diary, author_id: 1 }
})
})
}
Example database schema:
CREATE TABLE authors (
id SERIAL,
name VARCHAR,
books INTEGER[],
CONSTRAINT author_pkey PRIMARY KEY(id)
);
CREATE TABLE books (
id SERIAL,
title VARCHAR,
author_id INTEGER,
CONSTRAINT book_pkey PRIMARY KEY(id),
CONSTRAINT author FOREIGN KEY (author_id) REFERENCES authors(id)
);
Pro Tip: Create a trigger to auto-populate author.books[]
.
Schema configurationNaming conventionsMigrationsJoin-based hydration
## oreo( opts, [cb] )
Instantiates the db
object and configures the database connection string(s).
- opts {Object} options
- cb {Function} (optional) callback(err)
var oreo = require('oreo')
var db = oreo({
driver: 'pg',
hosts: ['localhost:5432'],
name: 'database',
user: 'username',
pass: 'password',
debug: false,
memoize: 150, // ms to cache data objects in app ram
cache: null // object with get/set methods to cache data objects, i.e. redis client
}, function(err) {
db.execute('select now() as now', function(err, rs) {
console.log('now:', rs[0].now)
})
})
Hacker Tip: Replicate to Redis so your cache is never stale.
## db.discover( [cb] )Re-discover the schema in the database.
- cb {Function} (optional) callback(err)
For each table in the database, defines a property db.<table_name>
whose value is a Table
object.
Automatically runs when oreo is instantiated. Also, you can specify methods that will be bound to each Row
object that is returned by Table.get()
.
db.discover(function(err) {
// the Table API (see docs below) is now available:
// db.authors
// db.books
// bind a method to all "book" objects
db.books._methods.getTitle = function() {
return this.title
}
})
Executes an arbitrary SQL query.
- query {String|Array} the SQL statement
- data {Object} (optional, unless
options
is specified) parameterized query data - options {Object} (optional) query options
write
(optional) if truthy, forces query to run on master db, otherwise attempts to run on a read-only hostconString
(optional) the connection string of the db
- cb {Function} (optional) callback(err, results)
db.execute([
'select now()', // arrays can be used for multi-line convenience
'as now'
], function(err, rs) {
console.log(rs[0]) // 2014-06-24 21:03:08.652861-04
})
Parameterized query (SQL injection safe):
db.execute([
'select id',
'from authors',
'where name = :name'
], {
name: 'Jack Kerouac',
}, function(err, rs) {
console.log(rs[0].id) // 1
})
If no callback is provided a stream is returned:
db.execute('select now()')
.on('data', function(row) {
})
.on('error', function(error) {
})
.on('end', function(result) {
})
Finds one or more rows:
db.authors.find({
where: ["name ilike 'Jack%'"],
order: 'name asc',
offset: 5,
limit: 5
}, function(err, authors) {
console.log(authors[0].id) // 1
})
The where
option has several valid formats:
- {String}
where: "field = 'abc' and field2 > 1"
- {Array}
where: ["field = 'abc'", "field2 > 1"]
- {Object} recommended, blocks SQL injection
where: { field: 'abc', field2: ['>', 1] }
If no callback is provided a stream is returned.
## db.table.findOne( opts, [cb] )Finds exactly one row:
db.authors.findOne({
where: ["name ilike 'Jack%'"],
order: 'name asc',
offset: 5
}, function(err, author) {
console.log(author.id) // 1
})
If no callback is provided a stream is returned.
## db.table.get( primaryKey, [cb] )Finds a row by primary key:
db.authors.get(1, function(err, author) {
console.log(author) // { id: 1, name: Jack Kerouak, books: [1] }
})
Multi-column (composite) primary key:
db.parts.get({
company: 'Cogswell Cogs',
part_no: 'A-12345'
}, function(err, part) {
})
Inserts a new row.
db.books.insert({
title: 'On the Road',
author_id: 1
}, function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1 }
})
Insert multiple rows into related tables in a single transaction:
db.books.insert({
title: 'On the Road',
author: {
name: 'Jack Kerouac'
}
}, function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1 }
book.hydrate(function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1, author: { id: 1, name: Jack Kerouac, books: [1] } }
})
})
Gets many rows from the database by primary key:
var bookIds = [1]
db.books.mget(bookIds, function(err, books) {
console.log(books)
// [ { id: 1, title: On the Road, author_id: 1 } ]
})
If no callback is provided a stream is returned.
## db.table.save( data, [cb] )Inserts or updates depending on whether the primary key exists in the db.
var formPOST = {
id: 1,
title: 'New Title'
}
db.books.save(formPOST, function(err, book) {
console.log(book)
// { id: 1, title: New Title, author_id: 1 }
})
Gets the linked record (foreign key)
db.books.get(1, function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1 }
book.hydrate('author', function(err, author) {
console.log(book)
// {
// id: 1,
// title: On the Road,
// author_id: 1,
// author: { id: 1, name: Jack Kerouac, books: [1] }
// }
})
})
Saves the modified property values to the database (recursively):
db.books.get(1, function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1 }
book.author_id = 2
book.save(function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 2 }
})
})
Sets multiple property values but does not save yet:
db.books.get(1, function(err, book) {
console.log(book)
// { id: 1, title: On the Road, author_id: 1 }
book.set({
title: 'New Title',
author_id: 2
})
book.save()
})
Update an existing row:
book.update({
title: 'New Title'
}, function(err, book) {
console.log(book)
// { id: 1, title: New Title, author_id: 1 }
})