Scrubbing made easy
Acts as scrubbable give you model level access to scrub your data per object.
It runs using the parallel gem for faster processing which is dependent on the amount of cores available on the box. More cores == faster scrubbing
gem 'acts_as_scrubbable'
Simple add the configuration for your fields that map directly to your columns
class User < ActiveRecord::Base
...
acts_as_scrubbable :first_name, :last_name
# optionally you can add a scope to limit the rows to update
scope :scrubbable_scope, -> { where(some_value: true) }
...
end
Incase the mapping is not straight forward
class Address
acts_as_scrubbable :lng => :longitude, :lat => :latitude
end
The confirmation message will be the db host
rake scrub
....
2015-11-19 10:52:51 -0800: [WARN] - Please verify the information below to continue
2015-11-19 10:52:51 -0800: [WARN] - Host: 127.0.0.1
2015-11-19 10:52:51 -0800: [WARN] - Database: blog_development
Type '127.0.0.1' to continue.
-> 127.0.0.1
2015-11-19 10:52:51 -0800: [WARN] -- : Scrubbing classes
2015-11-19 10:52:51 -0800: [WARN] -- : Scrubbing ClassToScrub
...
2015-11-19 10:52:51 -0800: [WARN] -- : Scrub Complete!
In the case you are automating the rake task and want to skip the confirmation
rake scrub SKIP_CONFIRM=true
If you want to limit the classes you to be scrubbed you can set the SCRUB_CLASSES
variable
rake scrub SCRUB_CLASSES=Blog,Post
If you want to skip the afterhook
rake scrub SKIP_AFTERHOOK=true
You may find the need to extend or add additional generators or an after_hook
ActsAsScrubbable.configure do |c|
c.add :email_with_prefix, -> { "prefix-#{Faker::Internet.email}" }
c.after_hook do
puts "Running after commit"
ActiveRecord::Base.connection.execute("TRUNCATE some_table")
end
end
By default, the scrubbing proces will run independent UPDATE statements for each scrubbed model. This can be time
consuming if you are scrubbing a large number of records. As an alternative, some
databases support doing bulk database updates using an upsert using the INSERT command. activerecord-import
gives us easy support for this and as such it is a requirement to using
this feature. Some details about the implementation can be found here. https://github.com/zdennis/activerecord-import#duplicate-key-update
Note that we only support the MySQL implementation at this time.
To use UPSERT over UPDATE, it can be enabled by specifying the environment variable USE_UPSERT='true'
or through configuration.
ActsAsScrubbable.configure do |c|
c.use_upsert = true
end