foo_parser.rb

ActiveRecord::ConnectionAdapters::Column::TRUE_VALUES << "X"

class FooParser
  def self.from_tabs(stream)
    stream.each do |line|
      while line.split("\t").size < FIELDS.size do
        line += stream.readline
      end
      parts = line.split "\t"
      next if parts.all? {|p| p.to_s.strip.blank?}
      params = {}
      FIELDS.each_with_index do |field, i|
        params[field] = parts[i].strip
      end
      yield BarRegistration.new(params)
    end
  end

  FIELDS = %w(
foo_id
foo_db_key
name
...75 more
)
end

foo_parser_spec.rb

describe FooParser do
  describe "from_tabs" do
    before(:each) do
      @file = File.open support_file("data_files/foodatadec08tab_first_10.txt")
      @registrations = []
      FooParser.from_tabs(@file) {|reg| @registrations << reg}
    end

    it "should yield BarRegistration objects for each row" do
      @registrations.should have(10).items
      lambda { @registrations.each(&:save) }.
        should change(BarRegistration, :count).by(10)
    end

    it "should set the fields in the db" do
      reg = @registrations.first
      reg.foo_id.should include("280039")
      reg.cod_expire_date.should == Date.parse("03/31/2009")
    end

    it "should work even as it gets past a random line break" do
      reg = @registrations[9]
      reg.name.should == "VALKYRIE"
    end

    it "should strip surrounding white space" do
      @registrations.first.foo_id.should == "280039"
    end

    it "should correctly handle X as a boolean true" do
      @registrations[0].should be_recreation
      @registrations[1].should_not be_recreation
    end
  end
end