Implements migration scripts #268
Conversation
| @@ -8,6 +8,8 @@ gem 'eventmachine' if ENV['EM'] | |||
| gem "activesupport", "< 5.0.0" | |||
| gem "activemodel", "< 5.0.0" | |||
|
|
|||
| gem 'generator_spec' | |||
There was a problem hiding this comment.
Should be in the 'development' group? or we want it as a real dependency?
There was a problem hiding this comment.
As far as I understood the way it work, Gemfile in a gem is just like add_development_dependency and therefore shouldn't be a real dependency.
Only the one declared in the gemspec file with add_dependency will be real dependency.
If that statement is wrong, then I should update it, of course!
| raise unless respond_to?(:down) | ||
|
|
||
| begin | ||
| down |
There was a problem hiding this comment.
Print an error message right before calling down with the exception, and the fact that you are running down. People need to know what's going on.
| begin | ||
| down | ||
| raise | ||
| rescue StandardError |
There was a problem hiding this comment.
This doesn't seem useful?
|
|
||
| def rollback! | ||
| down if respond_to?(:down) | ||
| rescue StandardError |
There was a problem hiding this comment.
I'm not sure what's that rescue block gives you, can you explain?
| # | ||
| class Migrator | ||
| def self.migrations_table_name | ||
| 'nobrainer_migration_versions' |
There was a problem hiding this comment.
Create a document class for the migration table, similar to lib/no_brainer/lock.rb
You can call it MigrationState, and the table can be called 'nobrainer_migrations'
This should simplify your table manipulation logic.
| @@ -16,7 +16,8 @@ module NoBrainer | |||
| # Code that is loaded through the DSL of NoBrainer should not be eager loaded. | |||
| autoload :Document, :IndexManager, :Loader, :Fork, :Geo, :SymbolDecoration | |||
| eager_autoload :Config, :Connection, :ConnectionManager, :Error, | |||
| :QueryRunner, :Criteria, :RQL, :Lock, :ReentrantLock, :Profiler, :System | |||
| :QueryRunner, :Criteria, :RQL, :Lock, :ReentrantLock, :Profiler, | |||
| :System, :Migrator, :Migration | |||
There was a problem hiding this comment.
Why should they be eager loaded? and not just loaded on demand ?
| @@ -37,5 +43,22 @@ namespace :nobrainer do | |||
| task :reset => [:drop, :sync_schema_quiet, :seed] | |||
|
|
|||
| task :create => [:sync_schema] | |||
| task :migrate => [:sync_schema] | |||
There was a problem hiding this comment.
Shouldn't we sync_schema still?
| @path = path | ||
| @version, snake_class_name = path.scan(%r{^.*/(\d+)_(.*)\.rb}).flatten | ||
|
|
||
| @name = snake_class_name ? snake_class_name.camelize : nil |
| def initialize(path) | ||
| @path = path | ||
| @version, snake_class_name = path.scan(%r{^.*/(\d+)_(.*)\.rb}).flatten | ||
|
|
This PR adds support for migration scripts with nobrainer, heavily inspired from the ActiveRecord Migration Script concept (meaning familiar to RoR people that we are).
Migration script is really helpful in the data migration concern. Fix wrongly created data, deleting old data and more, can be easily done with migration scripts.
It is especially true as I've implemented the Capistrano integration too, exactly as AR is doing.
I've added a very basic generator which generates a migration script file named the same way as AR does. The migration scripts are inheriting from the new
NoBrainer::Migrationclass which holds the logic of runningupthendownifupfailed.Finally the new
NoBrainer::Migratorclass contains the heart of this PR and take care of:Benchmark.measureblock the migration script and shows the execution time (I've stoled theannouncemethod from ActiveRecord)The migrator supports both migrating all non executed scripts, and rollbacking the last script only.
So it's a very minimal implementation with tests everywhere AFAIK.
I'm using this branch for a production project at work and it works pretty well. I really hope you'll accept it 🙏. I'm of course ready to fix all the things that would need it.