使用 Behave 运行一系列特征文件

Question

I have a series of embedded applications that I use Behave to test, so I have several folders with feature files and step implementations. So to test each application, I go into each one of the application folders and run Behave.

I was thinking about how can I create a top-level test file that I could run to test all applications.

I know Gherkin doesn't allow importing a feature file into another feature file, so I can't have a top-level file that imports the other feature files.

I think the obvious answer would be to move the individual feature files for each test to as scenarios in a top-level feature file and control what is tested with tags. That would be doable, with is a lot of re-work and I'd rather not deal with all the possible interactions between the different scenarios.

And what if I created this top-level test as a separate Behave test that would go into each individual folder and run Behave? Let's say that I have the following folder structure:

samples/
│
└tests/
 │
 ├── BehaveTest1
 │   ├── BehaveTest1.feature
 │   └── steps
 │       └── test_steps.py
 └── BehaveTest2
     ├── BehaveTest2.feature
     └── steps
         └── test_steps.py

I know I can run Behave in both folders with

behave samples/tests/BehaveTest*

How can I use Behave to automate that?

In the samples folder I'd have the following feature file:

Feature: Test everything

  Scenario Outline: Test all samples
    Given the test folder is"<test_folder>"
     When we run the test
     Then the test is successful

Examples:
  | test_folder     |
  | BehaveTest1     |
  | BehaveTest2     |

The step implementation would change to the correct folder and load the feature file.

How can I do this with Behave?

Many thanks!

Answer 1

why don't you use tags? tags can be used on feature and scenario level

features

• folder1
  • steps
    • your_step_file
  • test1.feature (feature level tag "regression")
  • test2.feature (feature level tag "smoke")
• folder2
  • steps
    • your step file
  • test1.feature (feature level tag "regression")
  • test2.feature (feature level tag "smoke")

i believe you have this kind of nested structure. you can run it any feature file based on ur tag

behave --tags @smoke and if you want to ran them parallel not sequential then try out Behavex, thats another layer on top of behave. behavex -t @tag1,@tag2 --log-capture --parallel-processes 3 --parallel-scheme feature -o Reports