This dataset extension provides the #in_batches method. The method splits dataset in parts and yields it.

Note: currently only PostgreSQL database is supported.

Add this line to your application's Gemfile:

gem 'sequel-batches'

In order to use the feature you should enable the extension:

DB.extension(:batches)

After that the #in_batches method becomes available on dataset:

User.where(role: "admin").in_batches(of: 4) do |ds|
  ds.delete
end

Finally, here's an example including all the available options:

options = {
  of: 4,
  pk: [:project_id, :external_user_id],
  start: { project_id: 2, external_user_id: 3 },
  finish: { project_id: 5, external_user_id: 70 },
  order: :desc,
}

Event.where(type: "login").in_batches(**options) do |dataset|
  dataset.delete
end

You can set the following options:

Overrides primary key of your dataset. This option is required in case your table doesn't have a real PK, otherwise you will get Sequel::Extensions::Batches::MissingPKError.

Note that you have to provide columns that don't contain NULL values, otherwise this may not work as intended. You will receive Sequel::Extensions::Batches::NullPKError in case batch processing detects a NULL value on it's way, but it's not guaranteed since it doesn't check all the rows for performance reasons.

Sets chunk size (1000 by default).

A hash { [column]: <start_value> } that represents frame start for batch processing. Note that you will get Sequel::Extensions::Batches::InvalidPKError in case you provide a hash with wrong keys (ordering matters as well).

Same as start but represents the frame end.

Specifies the primary key order (can be :asc or :desc). Defaults to :asc.

Bug reports and pull requests are welcome on GitHub at https://github.com/umbrellio/sequel-batches.

The gem is available as open source under the terms of the MIT License.