rspec test result from csv.read mocking file
I'm using ruby 1.9 and I'm trying to do BDD. My first test 'should read in the csv' works, but the second where I require a file object to be mocked doesn't.
Here is my mode开发者_运维技巧l spec:
require 'spec_helper'
describe Person do
describe "Importing data" do
let(:person) { Person.new }
let(:data) { "title\tsurname\tfirstname\t\rtitle2\tsurname2\tfirstname2\t\r"}
let(:result) {[["title","surname","firstname"],["title2","surname2","firstname2"]] }
it "should read in the csv" do
CSV.should_receive(:read).
with("filename", :row_sep => "\r", :col_sep => "\t")
person.import("filename")
end
it "should have result" do
filename = mock(File, :exists? => true, :read => data)
person.import(filename).should eq(result)
end
end
end
Here is the code so far:
class Person < ActiveRecord::Base
attr_accessor :import_file
def import(filename)
CSV.read(filename, :row_sep => "\r", :col_sep => "\t")
end
end
I basically want to mock a file so that when the CSV method tries to read from the file it returns my data variable. Then I can test if it equals my result variable.
You can stub File.open
:
let(:data) { "title\tsurname\tfirstname\rtitle2\tsurname2\tfirstname2\r" }
let(:result) {[["title","surname","firstname"],["title2","surname2","firstname2"]] }
it "should parse file contents and return a result" do
expect(File).to receive(:open).with("filename","rb") { StringIO.new(data) }
person.import("filename").should eq(result)
end
StringIO
essentially wraps a string and makes it behave like an IO object (a File in this case)
Stubbing File.open
breaks pry
. To avoid this you can stub CSV.read
, which isn't as robust as stubbing file, but will let you use pry
:
let(:data) do
StringIO.new <<-CSV_DATA
0;48;78;108;279;351;405;694;872;1696
1.03;9.28;13.4;18.56;29.9;30.93;42.27;77.32;85.57;100.0
0.0;2.94;8.82;11.76;44.12;97.06;100.0;100.0;100.0;100.0
CSV_DATA
end
let(:csv_config) { { headers: true, return_headers: true, converters: :numeric } }
before { allow(CSV).to receive(:read).with(csv_path, csv_config) { CSV.parse(data, csv_config) } }
精彩评论