east - node.js database migration tool for different databases (extensible via adapters).
east connects to the db using particular adapter (mongodb, sqlite, postgres,
mysql, couchbase), keeps track of executed migrations by storing their names
inside db and makes connect to the db available inside migrate
and rollback
functions. east encourages you to use for migrations driver/syntax with which
you are already familiar with (apparently you use it for work with db at your
application) and doesn't provide universal api for working with any kind of
database.
east itself requires node.js >= 0.10 to work.
Please note that particular adapter may have another requirements (see documentation for specific adapter).
npm install east -g
alternatively you could install it locally
go to project dir and run
east init
after that you can create
, migrate
, rollback
your migrations.
Run east -h
to see all commands:
Usage: east [options] [command]
Commands:
init initialize migration system
create <basename> create new migration based on template
migrate [options] run all or selected migrations
rollback [options] rollback all or selected migrations
list [status] list migration with selected status (`new`, `executed` or `all`), `new` by default
*
Options:
-h, --help output usage information
-V, --version output the version number
--adapter <name> which db adapter to use
--config <path> config file to use
--timeout <timeout> timeout for migrate/rollback
--template <path> path to template for new migrations
--dir <dir> dir where migrations stored
--url <url> db connect url
--trace verbose mode (including error stack trace)
run east <command> -h
to see detail command help.
All options described above can be set via command line or at .eastrc
file
located at current directory, e.g.:
{
"dir": "./dbmigration",
"template": "./lib/node/utils/customMigrationTemplate.js"
}
.eastrc
also can be a regular nodejs script (instead of json file):
var path = require('path');
module.exports = {
dir: path.join(__dirname, 'dbmigration'),
template: './lib/node/utils/customMigrationTemplate.js'
};
east create doSomething
produces something like this
New migration `1_doSomething` created at migrations/1_doSomething.js
created file will contain
exports.migrate = function(client, done) {
done();
};
exports.rollback = function(client, done) {
done();
};
client
is connect to current db and he determined by adapter (see adapters section)
done
is function which should be called at the end of migration (if any
error occured you can pass it as first argument)
migration also can be synchronous - declare only client
at migrate
or rollback
rollback
function is optional and may be omitted
Migration file is normal node.js module and you can migrate any database e.g.
// include your database wrapper which you already use in app
var db = require('./db');
exports.migrate = function(client, done) {
db.connect(function(err) {
if (err) done(err);
db.things.insert({_id: 1, name: 'apple', color: 'red'}, done);
});
};
exports.rollback = function(client, done) {
db.connect(function(err) {
if (err) done(err);
db.things.remove({_id: 1}, done);
});
};
or you can use special adapter for database (see adapters section)
let's create one more migration
east create doSomethingElse
then executes both of them
east migrate
it sequentially executes all new migrations and produces
target migrations:
1_doSomething
2_doSomethingElse
migrate `1_doSomething`
migration done
migrate `2_doSomethingElse`
migration done
selected migrations can be executed by passing their names (or numbers or basenames or paths) as argument
east migrate 1_doSomething 2
in our case this command will skip all of them
skip `1_doSomething` because it`s already executed
skip `2_doSomethingElse` because it`s already executed
nothing to migrate
you can pass --force
option to execute already executed migrations.
This is useful while you develop and test your migration.
You also can export tags
array from migration and then migrate only
migrations that satisfied expression specified by option --tag
. Expression
consists of tag names and boolean operators &
, |
and !
. For example,
following command will migrate all migrations that have tag tag1
and not have
tag tag2
:
east migrate --tag 'tag1 & !tag2'
rollback
has similar to migrate
command syntax but executes rollback
function from migration file
east rollback
will produce
target migrations:
2_doSomethingElse
1_doSomething
rollback `2_doSomethingElse`
migration successfully rolled back
rollback `1_doSomething`
migration successfully rolled back
east list
shows new migrations e.g.
new migrations:
1_doSomething
2_doSomethingElse
target status could be specified as an argument e.g.
east list executed
adapter determines where executed migration names will be stored and what will be
passed to migrate
and rollback
function as client
.
Default adapter store executed migration names at file .migrations
which is
located at migrations directory and pass null
as client
.
Other adapters:
into cloned repository run
npm test
For writing your own adapter you should implement methods for connection, mark transaction as executed, etc see details inside built-in adapter and other adapters.
You also can run tests from current repository against your adapter:
- Clone current repository
- Create file
.eastrc
insidetest
directory with path and parameters for your adapter e.g.
{
"adapter": "../../east-mysql/lib/adapter",
"url": "mysql://user:password@localhost/east_test_db",
"createDbOnConnect": true
}
- Run
npm test
at root of the cloned repository.
MIT