# Birdspotting Some add-ons on `ActiveRecord::Migration` to make migration safer in the context of zero downtime deployment. [![Gem Version](https://badge.fury.io/rb/birdspotting.svg)](https://badge.fury.io/rb/birdspotting) [![Maintainability](https://api.codeclimate.com/v1/badges/4a272a3f849869a200df/maintainability)](https://codeclimate.com/github/drivy/birdspotting/maintainability) [![Test Coverage](https://api.codeclimate.com/v1/badges/4a272a3f849869a200df/test_coverage)](https://codeclimate.com/github/drivy/birdspotting/test_coverage) ## Installation Add this line to your application's Gemfile: ```ruby gem 'birdspotting' ``` And then execute: $ bundle Or install it yourself as: $ gem install birdspotting ## Usage ### Configuration You can configure the gem (for instance by creating a `config/initializers/birdspotting.rb`) with the following options (here with the default values): ```ruby Birdspotting.configure do |config| config.start_check_at_version = nil config.check_bypass_env_var = "BYPASS_SCHEMA_STATEMENTS_CHECK" config.add_column_position_check = true config.encoding_check = true config.encoding_check_message = "\n/!\\ You are dealing with a %s field" \ "(%s): did you think about emojis and used the appropriate encoding? /!\\ \n\n" config.rename_column_check = true config.rename_column_message = "Don't use rename_column! https://stackoverflow.com/a/18542147" config.remove_column_check = true end ``` #### Common configuration `start_check_at_version` allows to start the checks after some migration version only. Set it to a migration timestamp like 20151209000000 for instance. When nil, all migrations will be checked. `check_bypass_env_var` specify the ENV var allowing to bypass the checks. Use it to bypass temporarily all the checks so you do it intentionally. You can set it to any value, it's just testing it's set. For instance if check_bypass_env_var is set to BYPASS_SCHEMA_STATEMENTS_CHECK (the default) you can do: ``` BYPASS_SCHEMA_STATEMENTS_CHECK=true rails db:migrate:up VERSION=20180806142044 ``` ### add_column request position We like to keep or columns organised for the case where we don't use the ORM but some other client. This will raise a `Birdspotting::ColumnPositionMissingError` error if neither `:first` or `:after` is in the add_columns option. You can skip this validation by setting `add_column_position_check` to `false`. ### add_column encoding warning This will add a warning when adding a string (or text) column to warn us to think about encoding issues. Like do we want to support emojis, or unusual characters? You can skip this validation by setting `encoding_check` to `false`. You can customise the warning message by using the `encoding_check_message` setting. ### rename_column By default, we don't want to use the rename column possibility as it will break any live application. And we want to be able to release and run migration without downtime. Though when a rename_column is used, it will raise a `Birdspotting::RenameColumnForbiddenError`. You can skip this validation by setting `rename_column_check` to `false`. You can customise the warning message by using the `rename_column_message` setting. You might like to customize the warning message to be a link to an internal set of instructions for the correct way to do this. ### remove_column By default, we don't want to be able to remove a columns which is still in use by a the application. Thus we check if the column is still present in the columns list. - If we are not able to find the model, we issue a `Birdspotting::ModelNotFoundError`. - If the column is still present in the model, we issue a `Birdspotting::RemoveColumnForbiddenError`. We advise to set the column in the `ignored_columns` of the model. (See [this blog article](https://blog.bigbinary.com/2016/05/24/rails-5-adds-active-record-ignored-columns.html)) You can skip this validation by setting `remove_column_check` to `false`. ### reorder_columns [mySql only] As said above, we like to keep or columns organised for the case where we don't use the ORM but some other client. This helper allow to reorder all the columns of a table. Usage: ````ruby class ReorderPostsColumns < ActiveRecord::Migration[5.2] include Birdspotting::ReorderColumns def change reorder_columns_for :posts, %i{ id author body subject posted_at created_at updated_at } end end ```` **CAVEAT:** * All columns must be passed in parameters (or it will raise a `Birdspotting::MismatchedColumnsError`). * For now, it only works on mysql (or it will raise a `Birdspotting::UnsupportedAdapterError`). ## Development After checking out the repo, run `bin/setup` to install dependencies. Then, run `rake spec` to run the tests. You can also run `bin/console` for an interactive prompt that will allow you to experiment. To install this gem onto your local machine, run `bundle exec rake install`. To release a new version, update the version number in `version.rb`, and then run `bundle exec rake release`, which will create a git tag for the version, push git commits and tags, and push the `.gem` file to [rubygems.org](https://rubygems.org). ## Contributing Bug reports and pull requests are welcome on GitHub at https://github.com/drivy/birdspotting. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the [Contributor Covenant](http://contributor-covenant.org) code of conduct. ## License The gem is available as open source under the terms of the [MIT License](https://opensource.org/licenses/MIT). ## Code of Conduct Everyone interacting in the Birdspotting project’s codebases and issue trackers is expected to follow the [code of conduct](https://github.com/drivy/birdspotting/blob/master/CODE_OF_CONDUCT.md).