Re: [PATCH] Add userdiff patterns for Ada

Adrian Johnson [off-list ref] writes:

quoted

I do not seem to find anything interesting after @@, which means
that xfuncname is not tested at all even though the log message
claims the patch adds some.

I probably misunderstood how the tests work. I thought t4034 tested
wordRegex while t4018 is for xfuncname. I based the Ada tests on the
Pascal tests but changed the operators to Ada operators. I'm not really
sure what else the test needs.

The per-language for-loop you added ada to in t4018 is only to make
sure there is no regexp syntax error in the built-in xfuncname, and
does not check if the patterns make sense for the language at all.

You could add test vectors to check if the built-in xfuncname
catches beginning of functions in Ada correctly if you wanted to,
but I think observing what appears on @@ lines in t4034 test vector
would be a sufficient test.