TestGen - The Test Data Generation Toolset

It's difficult to create fresh, distinctive data:

• Unless you can completely reset things, once a given test case has been used, you need a new one
• You can't use "real" data: that's forbidden by GDPR and corporate data policies
• You may not have access to test instances of all the related systems
• Even if you have a source of fictionalised data it's hard to manually map it into your test cases
• It's not unusual to see all the test cases being "Testy McTester of Test Street" or variations on your tester's name and address – that means you're not getting good test coverage

It's hard to ensure you exercise all required scenarios, and understand the expected outcome of each test:

• This takes a lot of work if done manually
• For each new batch of data you need to re-apply the same variations
• This is repetitive and requires a lot of attention to detail, making it error prone

It's hard to track relationships between cases, especially across different systems and interfaces:

• You can manually track data relationships in a single interactive test script, but this becomes almost impossible for integration testing, where cases in different batches for different flows need to have defined relationships (e.g. one flow sets up the donors, and another their donations)
• You may need to update some cases once earlier ones have been run
• You then need to make sure that all the related cases are executed in the right order

Wouldn’t it be great if you could…

• Define a recipe to output one record for each stream of data
• List your scenarios, prescribing each one by selecting the output recipe and base data
• Define scenario-specific changes using a rich but simple change vocabulary
• Define the expected outcome of each test
• Add built-in variations to test things like address validation
• Cross-refer between scenarios so that Test B can share specific data items from Test A

…And then push a button to generate a fresh set of test data to your chosen recipe

• With brand-new, distinctive but realistic names, addresses and similar
• With each batch, test and variation given a clear structured identifier, so you can easily relate a given test to data in the target systems
• With tests to be run together organised into shared, structured files and batches
• With an overview document listing all the data and its expected destination

TestGen does exactly this!

TestGen is a Test Data Generator. It creates sets of distinct, fictionalised test data to test integration flows

With TestGen you can specify and generate data which has specific, realistic relationships:

• between items in a test (e.g. a sequence of dates)
• between tests (e.g. two updates to the same target)
• between batches of tests (e.g. the recruitment file creates a donor’s commitment, and then the transactions file creates donations from that donor)
• This allows testing of defined complex scenarios through chains of systems and interfaces

The focus is on integration functionality. However there’s nothing to stop the data being used for interactive testing through a user interface. It can be used for bulk/load testing as well.

TestGen doesn’t run tests like Selenium or LoadRunner, but it generates the data they need

TestGen ensures that data elements which should be the same between tests are, while those which need to be different are fresh

Once set up, generating a fresh data set is a simple job taking a few minutes