Jeff King <peff@xxxxxxxx> writes: >> I am reasonably happy with the "let's keep the vanilla untouched one >> in .git/config-initial, refrain from using [core] and other sections >> that MUST be in the initial configuration for testing, and use a >> wrapper that reads expected addition to the initial one from the >> standard input for validation" approach I came up with, but I am not >> happy with the name 'compare_expect'; 'validate_config_result' might >> be a better name. > > IMHO this is worse than just using "config --file" in most of the tests. > It's more steps to remember to deal with. And most tests do not care at > all what the source file is. "Most tests do not care" only indicates the lack of test coverage. Knowing the implementation, it probably is OK to assume that things would work fine as long as "--file <file>" works correctly, though ;-) Not having to keep the minimum parts of the real configuration file, and being able to use a throw-away file for each test, certainly makes things cleaner. > That said, most of the effort is in the tedium of switching each > individual test. I am happy for whoever volunteers to do that work to > have the final say in the approach. Yup.
