Describe the bug
It is not specified how to interpret the combination inlined: False + inlined_as_list: False when the range of the slot has an identifier or a key.
The SchemaLoader is leaving inlined unmodified, giving it implicit precedence over the value of inlined_as_list in this very specific combination. But is somehow inconsistent with the precedence that inlined_as_list: True is given over whatever value inlined might have.
This logic is probably the consequence of assuming that inlined_as_list defaults to False and therefore inlined_as_list: False would be interpreted as "unclear what the user wanted". But in fact inlined_as_list will be None if the attribute has not been specified at all, only SchemaLoader or SchemaView will infer a boolean, when required. And in that case some logic can be applied to decide what to resolve a None to.
Version of LinkML you are using 1.11.1
Please provide a schema (and if applicable, a data file) that replicates the issue
Describe the bug
It is not specified how to interpret the combination
inlined: False+inlined_as_list: Falsewhen the range of the slot has an identifier or a key.The SchemaLoader is leaving
inlinedunmodified, giving it implicit precedence over the value ofinlined_as_listin this very specific combination. But is somehow inconsistent with the precedence thatinlined_as_list: Trueis given over whatever valueinlinedmight have.This logic is probably the consequence of assuming that
inlined_as_listdefaults toFalseand thereforeinlined_as_list: Falsewould be interpreted as "unclear what the user wanted". But in factinlined_as_listwill beNoneif the attribute has not been specified at all, only SchemaLoader or SchemaView will infer a boolean, when required. And in that case some logic can be applied to decide what to resolve aNoneto.Version of LinkML you are using 1.11.1
Please provide a schema (and if applicable, a data file) that replicates the issue