On Thu, 18 Nov 2021 14:58:52 GMT, Pavel Rappo <pra...@openjdk.org> wrote:
>>> I assume the purpose of this naming scheme is to make visible what >>> combinations of features are covered by each test. I'm not sure I would >>> consider this enough benefit to justify method names which are very verbose >>> and hard to read (especially when you don't have fresh memory of the scheme >>> above). >> >> You are right when saying that this naming scheme helps to survey the test >> landscape. This is especially helpful while mass-developing tests, which is >> what I'm doing at the moment. > > Do you think we could add an underscore here and there to improve > readability? For example, we could separate the first four parts of the name > from the last part, which is a custom string: > > testNegativeExternalTag_FileNotFoundModuleSourcePath > > I'm not against "camelCase", but I have to admit that "snake_case" or a > reasonable mix of both styles might work better for longer names. In general, given the size of the file, I like the verbose names, if only because it gives the reader a strong hint up front about the form (nature) of the test. That being said, now that you have introduced `SnippetTester` (which is great!) you could split this file. I would definitely group positive and negative test cases in the same file, but you could maybe separate `Tag` and `Markup` tests into separate files. (Just a suggestion). FWIW, my experience is that it is easier to test/debug files with fewer methods in, although it is on my to-do list to make it easier to run individual test cases. ------------- PR: https://git.openjdk.java.net/jdk/pull/6359
