Mix scenario outline and parser

[haomingyin]

I'm trying to mix scenarios and parser to prevent repeating step definitions.

For example,

  Scenario Outline: Create users
    Given I want to create a user
    And my 'given_name' is <given_name>
    And my 'family_name' is <family_name>
    And my 'username' is <username>
    When I create the user
    Then the user will be created

    Examples:
      | given_name | family_name | username |
      | Tom        | Bob         | tom.bob  |

and the step definition is

@given("my '{attr_name}' is '<attr_value>'")
def build_request(attr_name, attr_value):
    # I would expect when attr_name is given_name and attr_value will be Tom
    # when attr_name is family_name and attr_value is Bob ...
    pass

obviously, it does not work. But I'm wondering if pytest-bdd supports use cases like that. As the user has 10+ attributes, it would be great to not repeat the step definitions.

[dcendents]

This already works, I have steps that mix both, but you need to use a parser to parse arguments (one of parse, cfparse or re).

e.g.:

@given(parsers.parse("my '{attr_name}' is '<attr_value>'"))
def build_request(attr_name, attr_value):
    pass

reference: https://github.com/pytest-dev/pytest-bdd#step-arguments

[youtux]

This should be fixed by #445