On Sun, 11 Sep 2016, David Bremner <david@tethera.net> wrote: > Jani Nikula <jani@nikula.org> writes: > >> We largely use the corpus under test/corpus for >> testing. Unfortunately, many of our tests have grown to depend on >> having exactly this set of messages, making it hard to add new message >> files for testing specific cases. >> >> Move the basic corpus under tests/corpora/default, and make it >> possible to add new, independent corpora along its side. This means >> tons of renames with a few tweaks to add_email_corpus function in >> test-lib.sh to let tests specify which corpus to use. >> >> --- >> >> This is to address id:87pooc1t4s.fsf@zancas.localnet > > It seems unobjectionable, but tbh I don't understand the problem it is > solving. Do you plan to ship static test corpuses for individual test > files, rather than using add_message? Even for the *one* message in the referenced patch, I don't know where I would put it as a message file. I worked around the issue by using a here document (which you objected to), but it's a general problem. If there's an issue someone sees, we can't just shove the message (or a set of messages) that reproduces the issue under corpus, because it would screw up all the other tests. There needs to be some other place. This seemed like a generic approach to the problem. I could also imagine a shell helper function to add a single message from somewhere under corpora to the db for testing. For some cases, it might be more readable than using add_message. (And for my case, it's impossible to use add_message because I need a broken message.) BR, Jani.