lydeeea ,
it depends mainly on two things:
- How are the input data structured and stored (as richie already pointed out)
- How are the test cases diverse? Shall we transform the test input into ReadyAPI test cases the same way (for all test cases)? Or are there some differences? Higher diversity makes automation more difficult.
- Would data driven tests fit this purpose? You can use couple of test cases which use large data sets as their inputs. This can be done pretty easilly in ReadyAPI.
For simpler automation you can:
- open a new project,
- create a Groovy test step,
- make a conversion script (read the input data and create test cases, test steps and assertions),
- run the step and
- check the results.
In case it fails, you can return to your last version in git. Anyway using git is highly recommended for any ReadyAPI scripting, since you can destroy your project or reverting back might be too difficult.
For larger transformations you can use a programming IDE (e.g. IntelliJ IDEA), use the ReadyAPI SDK and do anything you need with the full support of the Java/Groovy IDE, use any libraries, tools, create the programs, test it etc. With such approach you create the ReadyAPI test projects automatically from scratch. We have used this approach when API definitions and test data changed so often, that we generated ReadyAPI test just before its execution (on the fly).
Best regards,
Karel