>> Another thought: it'd be nice if the test suite could optionally use >> ISO-8859-1 instead of UTF-8. Even better, the tests that care could >> do both. That would have caught this problem. > >Rather than have the test author decide if it should do both, and be >prone to human error, could the whole suite not be run for multiple >ones, including an unusual little-used one?
At least the tests I've worked on have to be carefully crafted to handle a particular character set conversion (and to deal with the expected output). I don't see how we can make a set of tests that convert between two arbitrary character sets. --Ken _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
