NAME

DBIx::Class::Migration::Script - Tools to manage database Migrations

SYNOPSIS

dbic-migration status \
  --libs="lib" \
  --schema_class='MyApp::Schema' \
  --dns='DBI:SQLite:myapp.db'

DESCRIPTION

This is a class which provides an interface mapping between the commandline script dbic-migration and the back end code that does the heavy lifting, DBIx::Class::Migration. This class has very little of it's own functionality, since it basically acts as processing glue between that commandline application and the code which does all the work.

You should look at DBIx::Class::Migration and the tutorial over at DBIx::Class::Migration::Tutorial to get started. This is basically API level docs and a command summary which is likely to be useful as a reference when you are familiar with the system.

ATTRIBUTES

This class defines the following attributes.

migration

This contains an instance of DBIx::Class::Migration which is constructed from various attributes described futher in these docs.

This is basically a delegate for all the commands you perform with this interface.

includes

Accepts ArrayRef. Not required.

Similar to the commandline switch I for the perl interpreter. It adds to @INC for when we want to search for modules to load. This is primarily useful for specifying a path to find a "schema_class".

We make no attempt to check the validity of any paths specified. Buyer beware.

schema_class

Accepts Str. Required.

This is the schema we use as the basic for creating, managing and running your deployments. This should be the full package namespace defining your subclass of DBIx::Class::Schema. For example MyApp::Schema.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_SCHEMA_CLASS

If the "schema_class" cannot be loaded, a hard exception will be thrown.

target_dir

Accepts Str. Required.

This is the directory we store our migration and fixture files. Inside this directory we will create a fixtures and migrations sub-directory.

Although you can specify the directory, if you leave it undefined, we will use File::ShareDir::ProjectDistDir to locate the share directory for your project and place the files there. This is the recommended approach, and is considered a community practice in regards to where to store your distribution non code files. Please see File::ShareDir::ProjectDistDir as well as File::ShareDir for more information.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_TARGET_DIR

username

Accepts Str. Not Required

This should be the username for the database we connect to for deploying ddl, ddl changes and fixtures.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_USERNAME

password

Accepts Str. Not Required

This should be the password for the database we connect to for deploying ddl, ddl changes and fixtures.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_PASSWORD

dsn

Accepts Str. Not Required

This is the DSN for the database you are going to be targeting for deploying or altering ddl, and installing or generating fixtures.

This should be a DSN suitable for "connect" in DBIx::Class::Schema, something in the form of DBI:SQLite:myapp-schema.db.

Please take care where you point this (like production :) )

If you don't provide a value, we will automatically create a SQLite based database connection with the following DSN:

DBD:SQLite:[path to target_dir]/[db_file_name].db

Where c<[path to target_dir]> is "target_dir" and [db_file_name] is a converted version of "schema_class". For example if you set schema_class to:

MyApp::Schema

Then [db_file_name] would be myapp-schema.

Please remember that you can generate deployment files for database types other than the one you've defined in "dsn", however since different databases enforce constraints differently it would not be impossible to generate fixtures that can be loaded by one database but not another. Therefore I recommend always generated fixtures from a database that is consistent across enviroments.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_DSN

force_overwrite

Accepts Bool. Not Required. Defaults to False.

Used when building "migration" to en / disable the DBIx::Class::DeploymentHandler option force_overwrite. This is used when generating DDL and related files for a given version of your "_schema" to decide if it is ok to overwrite deployment files. You might need this if you deploy a version of the database during development and then need to make more changes or fixes for that version.

to_version

Accepts Int. Not Required.

Used to establish a target version when running an install. You can use this to force install a version of the database other then your current "_schema" version. You might want this when you need to force install a lower version as part of your development process for changing the database.

If you leave this undefined, no default value is built, however DBIx::Class::DeploymentHandler will assume you want whatever is the value of your $schema->version.

databases

Accepts ArrayRef. Not Required.

Used when building "migration" to define the target databases we are building migration files for. You can name any of the databases currently supported by SQLT. If you leave this undefined we will derive a value based on the value of "dsn". For example, if your "dsn" is "DBI:SQLite:test.db", we will set the valuye of "databases" to ['SQLite'].

fixture_sets

Accepts ArrayRef. Not Required. Defaults to ['all_tables'].

This defines a list of fixture sets that we use for dumping or populating fixtures. Defaults to the ['all_tables'] set, which is the one set we build automatically for each version of the database we prepare deployments for.

sandbox_class

Accepts String. Not required. Defaults to 'SqliteSandbox'

If you don't have a database already running, we will automatically create a database 'sandbox' in your "target_dir" that is suitable for development and rapid prototyping. This is intended for developers and intended to make life more simple, particularly for beginners who might not have all the knowledged needed to setup a database for development purposes.

By default this sandbox is a file based DBD::Sqlite database, which is an easy option since changes are good this is already installed on your development computer (and if not it is trivial to install).

You can change this to either 'PostgresqlSandbox' or 'MySQLSandbox', which will create a sandbox using either DBIx::Class::Migration::MySQLSandbox or DBIx::Class::Migration::PostgresqlSandbox (which in term require the separate installation of either Test::mysqld or Test::postgresql). If you are using one of those open source databases in production, its probably a good idea to use them in development as well, since there are enough small differences between them that could make your code break if you used sqlite for development and postgresql in production. However this requires a bit more setup effort, so when you are starting off just sticking to the default sqlite is probably the easiest thing to do.

YOu should review the documenation at DBIx::Class::Migration::MySQLSandbox or DBIx::Class::Migration::PostgresqlSandbox because those delegates also build some helper scripts, intended to help you use a sandbox.

Uses MooseX::Attribute::ENV to let you populate values from %ENV. Uses key DBIC_MIGRATION_SANDBOX_CLASS

If you need to create your own custom database sandboxes, please see: DBIx::Class::Migration::Sandbox which is the role your sandbox factory needs tp complete. You can signify your custom sandbox by using the full package name with a '+' prepended. For example:

sandbox_class => '+MyApp::Schema::CustomSandbox'

You should probably look at the existing sandbox code for thoughts on what a good sandbox would do.

migration_class

Accepts String. Not Required (Defaults: DBIx::Class::Migration)

Should point to the class that does what DBIx::Class::Migration does. This is exposed here for those who need to subclass DBIx::Class::Migration. We don't expose this attribute to the commandline, so if you are smart enough to do the subclassing (and sure you need to do that), I will assume you will also either subclass DBIx::Class::Migration:Script or override then default value using some standard technique.

dbic_fixture_class

Accepts: Str, Not Required.

You can use this if you need to make a custom subclass of DBIx::Class::Fixtures.

dbic_fixtures_extra_args

Accepts: HashRef, Not Required.

If provided will add some additional arguments when creating an instance of "dbic_fixture_class". You should take a look at the documentation for DBIx::Class::Fixtures to understand what additional arguments may be of use.

COMMANDS

dbic_migration -Ilib install

Since this class consumes the MooseX::GetOpt role, it can be run directly as a commandline application. The following is a list of commands we support as well as the options / flags associated with each command.

help

Summary of commands and aliases.

version

prints the current version of the application to STDOUT

status

Returns the state of the deployed database (if it is deployed) and the state of the current schema

prepare

Creates a fixtures and migrations directory under "target_dir" (if they don't already exist) and makes deployment files for the current schema. If deployment files exist, will fail unless you "overwrite_migrations" and "overwrite_fixtures".

install

Installs either the current schema version (if already prepared) or the target version specified via "to_version" to the database connected via the "dsn", "username" and "password"

upgrade

Run upgrade files to either bring the database into sync with the current schema version, or stop at an intermediate version specified via "to_version"

downgrade

Run down files to bring the database down to the version specified via "to_version"

dump_named_sets

Given listed fixture_sets, dump files for the current database version (not the current schema version)

dump_all_sets

Just dump all the sets for the current database

populate

Given listed fixture_sets, populate a database with fixtures for the matching version (matches database version to fixtures, not the schema version)

drop_tables

Drops all the tables in the connected database with no backup or recovery. For real! (Make sure you are not connected to Prod, for example)

delete_table_rows

does a delete on each table in the database, which clears out all your data but preserves tables. For Real! You might want this if you need to load and unload fixture sets during testing, or perhaps to get rid of data that accumulated in the database while running an app in development, before dumping fixtures.

Skips the table dbix_class_deploymenthandler_versions, so you don't lose deployment info (this is different from "drop_tables" which does delete it.)

make_schema

Creates DBIC schema files from the currently deployed database into your target directory. You can use this to bootstrap your ORM, or if you get confused about what the deployment perl run files get for schema.

install_if_needed

Install the database to the current $schema version if it is not currently installed. Otherwise this is a nop (even if the database is behind the schema).

Command Flags

The following flags are used to modify or inform commands.

includes

Aliases: I, lib Value: String or Array of Strings

dbic_migration --includes lib --includes /opt/perl/lib

Does the same thing as the Perl command line interpreter flag I. Adds all the listed paths to @INC. You will likely use this in order to find your application schema (subclass of DBIx::Class::Schema.

schema_class

Value: Str

dbic_migration prepare --schema_class MyApp::Schema -Ilib

Used to specify the DBIx::Class::Schema subclass which is the core of your DBIx::Class based ORM managed system. This is required and the application cannot function without one.

target_dir

Aliases: D Value: Str

dbic_migration prepare --schema_class MyApp::Schema --target_dir /opt/share

We need a directory path that is used to store your fixtures and migration files. By default this will be the share directory for the application where your "schema_class" resides. I recommend you leave it alone since this is a reasonable Perl convention for non code data, and there's a decent ecosystem of tools around it, however if you need to place the files in an alternative location (for example you have huge fixture sets and you don't want them in you core repository, or you can't store them in a limited space filesystem) this will let you do it. Option and defaults as discussed.

username

Aliases: U Value: Str

password

Aliases: P Value: String

dsn

Value: String

These three commandline flags describe how to connect to a target, physical database, where we will deploy migrations and fixtures. If you don't provide them, we will automatically deploy to a DBD::SQLite managed database located at "target_dir".

dbic_migration install --username myuser --password mypass --dsn DBI:SQLite:mydb.db

force_overwrite

Aliases: O Value: Bool (default: False)

Sometimes you may wish to prepare migrations for the same version more than once (say if you are developing a new version and need to try out a few options first). This lets you deploy over an existing set. This will of course destroy and manual modifications you made, buyer beware.)

dbic_migration prepare --overwrite_migrations

to_version

Aliases: D Value: Str (default: Current VERSION of Schema)

dbic_migration install --to_version 5

Used to specify which version we are going to deploy. Defaults to whatever is the most current version you've prepared.

Use this when you need to force install an older version, such as when you are roundtripping prepares while fiddling with a new database version.

databases

Alias: database Value: Str or Array of Str (default: SQLite)

You can prepare deployment for any database type that SQLT understand. By default we only prepare a deployment version for the database which matches the dsn you specified but you can use this to prepare additional deployments

dbic_migration prepare --database SQLite --database mysql

Please note if you choose to manually set this value, you won't automatically get the default, unless you specify as above

fixture_sets

Alias: fixture_set Value: Str or Array of Str (default: all set)

When dumping or populating fixture sets, you use this to set which sets.

dbic_migration dump --fixture_set roles --fixture_set core

Please note that if you manually describe your sets as in the above example, you don't automatically get the all_tables set, which is a fixture set of all database information and not 'all' the sets.

We automatically create the all_tables fixture set description file for you when you prepare a new migration of the schema. You can use this set for early testing but I recommend you study DBIx::Class::Fixtures and learn the set configuration rules, and create limited fixture sets for given purposes, rather than just dump / populate everything, since that is like to get big pretty fast

My recommendation is to create a core 'seed' set, of default database values, such as role types, default users, lists of countries, etc. and then create a 'demo' or 'dev' set that contains extra information useful to populate a database so that you can run test cases and develop against.

sandbox_class

Alias: T Value: String (default: SqliteSandbox)

If you don't have a target database for your migrations (as you might not for your development setup, or during initial prototyping) we automatically create a local database sandbox in your "target_dir". By default this is a DBD::Sqlite single file database, since this is easy to get installed (you probably already have it) and is easy to work with. However, we can also create database sandboxes for mysql and postgresql (although you will need to get the Test::mysqld and/or Test::postgresql as well as the correcct DBD installed).

This is handy as you move toward a real production target and know the eventual database for production. If you choose to create a postgresql or mysql database sandbox, they will automatically be created in your "target_dir", along with some helper scripts. See DBIx::Class::Migration::PostgresqlSandbox and DBIx::Class::Migration::MySQLSandbox for more documentation.

Assuming you've prepared migrations for an alternative sandbox, such as MySQL:

dbic_migration install --schema_class MyApp::Schema --sandbox_class MySQLSandbox

Would install it. Like some of the other option flags you can specify with an %ENV setting:

export DBIC_MIGRATION_SANDBOX_CLASS=MySQLSandbox

This would be handy if you are always going to target one of the alternative sandbox types.

The default sqlite sandbox is documented at DBIx::Class::Migration::SQLiteSandbox although this single file database is pretty straightforward to use.

If you are declaring the value in a subclass, you can use the pre-defined constants to avoid typos (see "CONSTANTS").

OPTIONAL METHODS FOR SUBCLASSES

If you decide to make a custom subclass of DBIx::Class::Migration::Script, (you might do this for example to better integrate your migrations with an existing framework, like Catalyst) you may defined the following option methods.

default

Returns a Hash of instantiation values.

Merges some predefined values when instantiating. For example:

package MusicBase::Schema::MigrationScript;

use Moose;
use MusicBase::Web;

extends 'DBIx::Class::Migration::Script';

sub defaults {
  schema => MusicBase::Web->model('Schema')->schema,
}

__PACKAGE__->meta->make_immutable;
__PACKAGE__->run_if_script;

This would create a version of your script that already includes the target schema. In this example we will let Catalyst configuration handle which database to run deployments against.

ADDITIONAL INFORMATION FOR SUBCLASSERS

The following methods are documented of interest to subclassers.

run_if_script

Class method that detects if your module is being called as a script. Place it at the end of your subclass:

__PACKAGE__->run_if_script;

This returns true in the case you are using the class as a module, and calls "run_with_options" otherwise. Adding this lets you invoke your class module as a script from the commandline (saving you the trouble of writing a thin script wrapper).

perl -Ilib lib/MyApp/Schema/MigrationScript.pm --status

run_with_options

Given a Hash of initial arguments, merges those with the results of values passed on the commandline (via MooseX::Getopt) and run.

run

Actually runs commands.

CONSTANTS

The following constants are defined but not exported. These are used to give a canonical value for the "sandbox_class" attribute.

SANDBOX_SQLITE

SqliteSandbox

SANDBOX_MYSQL

MySQLSandbox

SANDBOX_POSTGRESQL

PostgresqlSandbox

EXAMPLES

Please see DBIx::Class::Migration::Tutorial for more. Here's some basic use cases.

head2 Prepare deployment files for a schema

dbic_migration prepare --schema_class MyApp::Schema

This will prepare deployment files for just SQLite

dbic_migration prepare --database SQLite --database mysql \
  --schema_class MyApp::Schema

This will prepare deployment files for both SQLite and MySQL

head2 Install database from deployments

dbic_migration install --schema_class MyApp::Schema

Creates the default sqlite database in the share directory.

dbic_migration install --schema_class MyApp::Schema --to_version 2

Same as the previous command, but installs version 2, instead of whatever is the most recent version

dbic_migration populate --schema_class MyApp::Schema --fixture_set seed

Populates the seed fixture set to the current database (matches the database version to the seed version.

SEE ALSO

DBIx::Class::Migration

AUTHOR

See DBIx::Class::Migration for author information

COPYRIGHT & LICENSE

See DBIx::Class::Migration for copyright and license information

cpanm DBIx::Class::Migration

perl -MCPAN -e shell
install DBIx::Class::Migration