DataTable#symbolic_hashes followed by DataTable#hashes produces undefined method: 'tr'

[Ben-Behar]

Summary

When using a gherkin step with a DataTable, if in my step definition I use table.symbolic_hashes and then later use table.hashes, then I am presented with the error undefined method 'tr' for :my_key:Symbol.

Expected Behavior

I should be able to access my table as a list of symbolic hashes as needed without changing any of the state of the DataTable. After this, I should be able to access my table as a list of string hashes without issue.

Current Behavior

With a feature file:

Feature: Testificate
  Scenario: Testify
    When test
      | my_key   |
      | my_value |

And a step definition file:

When('test') do |table|
  table.symbolic_hashes
  table.hashes
end

I will receive the following error during execution:

Scenario: Testify
    When test      
      | my_key   |
      | my_value |
      undefined method `tr' for :my_key:Symbol
      Did you mean?  try (NoMethodError)
      ./features/step_definitions/testify_steps.rb:139:in `/^test$/'
      ./features/testben.feature:6:in `When test'

Possible Solution

In the following method:

def symbolic_hashes
  @header_conversion_proc = lambda { |h| symbolize_key(h) }
  @symbolic_hashes ||= build_hashes
end

Do not set the @header_conversion_proc, but instead pass an optional block through the method build_hashes that will define the appropriate strategy for header conversion.

Context & Motivation

I ran into the issue when trying to minimize use of string hashes from the testing framework at my company.

Your Environment

• Version used: 2.4.0
• Operating System and version: macOS Sierra 10.12.6
• Link to your project: N/A

[danascheider]

Thanks for reporting, @Ben-Behar! Is this something you'd be able to put together a pull request for? If so, we'd be happy to help you figure out the best implementation, etc.

[Ben-Behar]

Sure, I'll give it a shot. I will definitely need some hand holding though.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.