Closed nflorin closed 2 years ago
This is going to be very difficult, because for a lot of errors we'll have to find bad WorldCat records and then periodically check them to make sure that they haven't been fixed.
I've found it's easy enough to temporarily change the test inputs to add whatever issues you want to test for. So I'm going to close this.
The test inputs are based on real data files and so have very good success rates. We should manipulate them so that there is at least one example of every possible error in every file. Also the files have a lot of records in them (2000 in the mrk and txt files; 1120 in the csv). They probably don't need to be this big, given that running all three for the first time will require something like 4500 API hits that will take quite a while to run.
So in brief: