A simple and language-independent SQL schema migration tool
You can download the latest release on the release page of the project.
Go users can also simply compile it from source and install it as a go executable using the following command :
go install github.com/elwinar/rambler
Releases are compiled using the wonderful XGo. Don't hesitate to check it out, it really kicks some serious ass.
In rambler, migrations are kept in the simplest form possible: a migration is a list of sections (up
and down
), each section being an SQL statement. Example:
-- rambler up
CREATE TABLE foo (
id INTEGER UNSIGNED AUTO_INCREMENT,
bar VARCHAR(60),
PRIMARY KEY (id)
);
-- rambler down
DROP TABLE foo;
Sections are delimited by SQL comments sufixed by the rambler marker (white-spaces sensitive). While applying a migration, rambler will execute each up
section in order, and while reversing it it will execute each down
section in reverse order.
Migrations are executed in alphabetical order, thus a versionning scheme of the form version_description.sql
is highly recommended, version being an integer value, and description an underscored string. Examples:
201409272258_Added_table_foo.sql
01_First_migration.sql
Rambler configuration is lightweight: just dump the credentials of your database and the path to your migrations' directory into a JSON file, and you're done. Here is an example or JSON configuration file with the default values of rambler:
{
"driver": "mysql",
"protocol": "tcp",
"host": "localhost",
"port": 3306,
"user": "root",
"password": "",
"database": "",
"directory": "."
}
When running, rambler will try to find a configuration file in the working directory and use its values to connect to the managed database.
Rambler now supports HJSON configuration files, which is by the way retrocompatible with JSON.
Rambler supports actually 3 drivers:
mysql
postgresql
sqlite
Don't hesitate to get in touch if you want to see another one supported, provided a golang database/sql
driver exist for your database vendor.
To apply a migration, use the apply
command.
rambler apply
Rambler will compare the migrations already applied and the available migrations in increasing order to find the next migration to apply, then execute all its up
sections' statements in order.
To reverse a migration, use the reverse
command.
rambler reverse
Rambler will compare the migrations already applied and the available migrations in decreasing order to find the last applied migrations, then execute all its down
sections' statements in reverse order.
You can tell rambler to repeat the process while there is a migration to apply (or reverse) with the all
flag (or its shorthand, a
).
To ensure database schema consistency, rambler will complain and stop when encountering a new migration in the middle of the already existing ones or if it can't find a migration already applied.
An environment is an additionnal configuration that is given a name, and can be used to create multiple configurations for a single application (for example, to differenciate production, testing, etc).
Environments are defined in the configuration file, under the environments
item.
Each environment is defined as an attribute of this item, the key being the name and the value being the configuration options.
Environments configuration are derived from the default configuration of rambler (at the configuration file's root), so you only need to override the needed options:
{
"driver": "mysql",
"protocol": "tcp",
"port": 3306,
"user": "root",
"password": "",
"database": "rambler_default",
"directory": "migrations",
"environments": {
"development": {
"database": "rambler_development"
},
"testing": {
"database": "rambler_testing"
}
}
}
Here we have three environments defined:
default
, will use therambler_default
database,development
, will use therambler_development
database,testing
, will use therambler_testing
database;
Feel free to give feedback, make pull requests or simply open issues if you find a bug or have an idea.