capybara - Find with xPath is leaving the within scope

I am trying to build a date selector with Capybara using the default Rails date, time, and datetime fields. I am using the within method to find the select boxes for the field but when I use xPath to find the correct box it leaves the within scope and find the first occurrence on the page of the element.

Here is the code I am using. The page I am testing on has 2 datetime fields but I can only get it to change the first because of this error. At the moment I have an div container with id that wraps up the datetime field but I do plan on switching the code to find by the label.

module  Marketron
  module DateTime

    def select_date(field, options = {})
      date_parse = Date.parse(options[:with])

      year = date_parse.year.to_s
      month = date_parse.strftime('%B')
      day = date_parse.day.to_s

      within("div##{field}") do
        find(:xpath, "//select[contains(@id, \"_#{FIELDS[:year]}\")]").select(year)
        find(:xpath, "//select[contains(@id, \"_#{FIELDS[:month]}\")]").select(month)
        find(:xpath, "//select[contains(@id, \"_#{FIELDS[:day]}\")]").select(day)
      end

    end

    def select_time(field, options = {})
      require "time"
      time_parse = Time.parse(options[:with])

      hour = time_parse.hour.to_s.rjust(2, '0')
      minute = time_parse.min.to_s.rjust(2, '0')

      within("div##{field}") do
        find(:xpath, "//select[contains(@id, \"_#{FIELDS[:hour]}\")]").find(:xpath, "option[contains(@value, '#{hour}')]").select_option
        find(:xpath, "//select[contains(@id, \"_#{FIELDS[:minute]}\")]").find(:xpath, "option[contains(@value, '#{minute}')]").select_option

      end

    end

    def select_datetime(field, options = {})
      select_date(field, options)
      select_time(field, options)
    end

    private

      FIELDS = {year: "1i", month: "2i", day: "3i", hour: "4i", minute: "5i"}

  end
end

World(Marketron::DateTime)

You should specify in the xpath that you want to start with the current node by adding a . to the start:

find(:xpath, ".//select[contains(@id, \"_#{FIELDS[:year]}\")]")

Example:

I tested an HTML page of this (hopefully not over simplifying your page):

<html>
    <div id='div1'>
        <span class='container'>    
            <span id='field_01'>field 1</span>
        </span>
    </div>
    <div id='div2'>
        <span class='container'>
            <span id='field_02'>field 2</span>
        </span>
    </div>  
</html>

Using the within methods, you can see your problem when you do this:

within("div#div1"){ puts find(:xpath, "//span[contains(@id, \"field\")]").text }
#=> field 1
within("div#div2"){ puts find(:xpath, "//span[contains(@id, \"field\")]").text }
#=> field 1

But you can see that but specifying the xpath to look within the current node (ie using .), you get the results you want:

within("div#div1"){ puts find(:xpath, ".//span[contains(@id, \"field\")]").text }
#=> field 1
within("div#div2"){ puts find(:xpath, ".//span[contains(@id, \"field\")]").text }
#=> field 2

来源：https://stackoverflow.com/questions/10250181/capybara-find-with-xpath-is-leaving-the-within-scope

标签

ruby

ruby-on-rails-3

cucumber

capybara

ruby-1.9.2