Is there a way to configure Robot Framework so the execution is not stopped by a failure

Question

I'm aware of the Run Keyword and Continue on Failure / Run Keyword And Ignore Error / Run keyword and return status Builtin keywords but I have a very wide set of test cases that should not be stopped for any reason on a specific scenario and I was wondering if there is an option not to make the execution stop on a failure by default, without having to manage it through these keywords and adding a non-business related syntax in my upper layer keywords.

Answer 1

Although it feels a bit counter intuitive that you should want to continue when you've encountered an erronous situation, given that you may no longer be in control of the application. This in itself should be prevented. However, that said.

Given that you are already familiar with the family of Run and continue keywords, there is not much else to suggest and to answer the question with an affirmative: No.

The only approach is to wrap the keywords in a Run and Continue keyword.

Answer 2

Generally speaking, robot simply isn't designed to work the way you want. It's designed to exit a test when a keyword fails unless you explicitly run that keyword with one of the special keywords (eg: run keyword and continue on failure).

In some very limited cases, you can get this behavior by using a template that calls run keyword and continue on failure for every test step. This technique will only work if your test case is made up strictly of keywords, and doesn't try to save keyword results to variables.

For example, consider this test:

*** Test cases ***
Example
    log   step one
    log   step two
    fail  something went wrong
    fail  something else went wrong
    log   last step

If you run the above test, it will stop on the first failure. However, by adding a test template that uses run keyword and continue on failure, all the steps will run before continuing to the next test:

*** Test cases ***
Example
    [Template]  Run keyword and continue on failure
    log   step one
    log   step two
    fail  something went wrong
    fail  something else went wrong
    log   last step

This is what the report looks like with the above test: