Pretty-printing in "datatype answers"

[jvoigtlaender]

For example, in MatchAd tasks, the "input format" looks like this:

but then in feedback the correct solution is printed like this:

[nimec01]

The problem here is that AutoTool uses the ToDoc instance of MatchAdSolution to pretty print the entered solution. We however use the standard Show instance in the task description and evaluation.

Should we just replace the show function call with render . toDoc (from Autolib.ToDoc) or is there a better way?

[jvoigtlaender]

That would probably work for that task type at least.

In general, one has to be careful, first checking: Is the student submission parsed according to ToDoc? (The point being that while read and show are inverses of each other, it it is not the case that render . toDoc and show always agree, not even modulo layout/whitespace.)

(Meta: I was going to assign you another issue. You are not at the moment the only contributor for whom the possibly-hilfskraft labels are relevant.)