Closed jvoigtlaender closed 7 months ago
This has some interplay with other open issues/pull requests. At least #74. Probably also #73. Maybe further ones.
So, some stragegy is needed to make progress without introducing too many source code conflicts.
Being worked on in #91.
The
SynTreeConfig
type now (well, after merging #74) has the fieldsextraText
andextraHintsOnSemanticEquivalence
. That is fine whenSynTreeConfig
is being used as "its own task type" (namely, forTreeToFormula
), but actuallySynTreeConfig
also appears as subpart of other task types' config types. For example, inLegalPropositionConfig
. There, it doesn't make sense to specifyextraHintsOnSemanticEquivalence
. And havingextraText
viaSynTreeConfig
there is also quite strange, since any extra text to give there from the lecturer'S perspective would more reasonably be associated withLegalProposition
(as a task type), rather than withSynTree
(which only happens to be part of the specification ofLegalProposition
).This all comes down to a previously reasonable - but now not so much anymore - double role of
SynTreeConfig
. It would probably be better to separate these roles, by introducing a dedicated config type forTreeToFormula
, and sharingSynTreeConfig
(minus theextra
fields) betweenTreeToFormulaConfig
andLegalPropositionConfig
etc.Originally posted by @jvoigtlaender in https://github.com/fmidue/logic-tasks/issues/74#issuecomment-1879030447