Scheman

SQL schema parser.

Usage

Create diff from 2 schema files or input.

# before.sql
CREATE TABLE `table1` (
  `column1` INTEGER(11) PRIMARY KEY NOT NULL AUTO_INCREMENT,
  `column2` DATETIME DEFAULT NOW()
);

CREATE TABLE `table2` (
  `column1` INTEGER(11) NOT NULL AUTO_INCREMENT,
  PRIMARY KEY (`column1`)
);

# after.sql
CREATE TABLE `table1` (
  `column1` CHAR(11) NOT NULL AUTO_INCREMENT,
  `column2` DATETIME DEFAULT CURRENT_TIMESTAMP(),
  `column3` VARCHAR(255) NOT NULL DEFAULT "a",
  PRIMARY KEY (`column2`)
);

CREATE TABLE `table3` (
  `column1` INTEGER(11) NOT NULL AUTO_INCREMENT,
  PRIMARY KEY (`column1`)
);

$ scheman diff --before before.sql --after after.sql
BEGIN;

SET foreign_key_checks=0;

CREATE TABLE `table3` (
  `column1` INTEGER(11) NOT NULL AUTO_INCREMENT,
  PRIMARY KEY (`column1`)
);

ALTER TABLE `table1` ADD COLUMN `column3` VARCHAR(255) NOT NULL DEFAULT "a",
  CHANGE COLUMN `column1` CHAR(11) NOT NULL AUTO_INCREMENT,
  DROP PRIMARY KEY,
  ADD PRIMARY KEY `column2`;

DROP TABLE `table2`;

SET foreign_key_checks=1;

COMMIT;

STDIN

You can input schema data into scheman diff command via STDIN, instead of --before. For instance, this interface is useful when you want to use mysqldump command to get your current schema.

$ mysqldump --no-data --compact db_name | scheman diff --after after.sql

./schema.sql

Scheman use ./schema.sql as a default value of --after option.

$ mysqldump --no-data --compact db_name | scheman diff

Pipes

Here is an example workflow of schema modification, using UNIX pipes.

$ vi schema.sql
$ mysqldump --no-data --compact db_name | scheman diff | mysql db_name

Rails

scheman-rails provides some rake tasks to use scheman with Rails.