Feature: Processing Decision Logic
  As a data engineer
  I want the system to decide whether to process datasets
  So that I avoid reprocessing completed work

  Background:
    Given an authenticated sheets tracker
    And a sheet with the following datasets
      | Dataset Name | Status      | Error |
      | wordnet      | NOT STARTED |       |
      | yago         | IN PROGRESS |       |
      | dbpedia      | DONE        |       |
      | geonames     | SKIP        |       |

  Scenario: Should process NOT STARTED dataset
    When I check if "wordnet" should be processed
    Then the decision should be to process
    And the reason should contain "NOT STARTED"

  Scenario: Should not process DONE dataset
    When I check if "dbpedia" should be processed
    Then the decision should be to skip
    And the reason should contain "DONE"

  Scenario: Should not process IN PROGRESS dataset
    When I check if "yago" should be processed
    Then the decision should be to skip
    And the reason should contain "IN PROGRESS"

  Scenario: Should not process unknown status
    When I check if "geonames" should be processed
    Then the decision should be to skip
    And the reason should contain "Unknown status"

  Scenario: Force processing overrides status
    When I check if "dbpedia" should be processed with force enabled
    Then the decision should be to process
    And the reason should contain "Force processing"

  Scenario: Process dataset not in sheet
    When I check if "nonexistent" should be processed
    Then the decision should be to process
    And the reason should contain "No status found"

  Scenario: Process when tracker not authenticated (tracking is optional)
    Given an unauthenticated sheets tracker
    When I check if "wordnet" should be processed
    Then the decision should be to process
    And the reason should contain "not available"