| || |
looks for some strategy
I'm in charge of automatizing functional tests of our web application. This application is data-centric (several CRUD with some business logic interspersed) and based on a 200+ tables schema. The application supports an engineering, manufacturing process which is very sequential, chained. By sequential I mean each step prepares the database for the next one.
We've two options in mind: first, plan the test case as a sequence and each script will depend of the previous script, except the first. Second, each script prepare their data, and run independently of the previous script. But, for the first case each script must be prepared to accept the input data, write for another tester and, for the second case, will be expensive for the last scripts prepare all the information.
Someone has some experience or some reference that can help me?
Re: looks for some strategy
I was understanding your question until the last sentence. Could you rephrase that last part? What I think you're trying to do is fairly simple. If I understand correctly, you'll have a series of scripts. Each script after the first will operate independtly of any manual input. The script will start and either go to the database and fetch the information it needs or navigate through the GUI and store the information in temporary files. The last script should be able to take the original input and derive/verify the expected output. You might want to do an automated database backup between scripts (if feasible) so that you can have a starting point in case of errors and also for debugging purposes.