Grouping by property ID should be applied to:
Perhaps grouping should always be applied; in that case, the ungrouped structure should still be accepted as input.
Note: The grouped structure is better for access by property. The an ungrouped list map of statements is better for access by GUID.
https://github.com/wmde/WikibaseDataModelSerialization/issues/50
| Status | Subtype | Assigned | Task | |
|---|---|---|---|---|
| · · · | ||||
| Resolved | Lydia_Pintscher | T73170 [Story] Ensure feature parity of serialization based on WikibaseDataModelSerialization with what we do with WikibaseLib | ||
| Declined | Addshore | T78653 DataModelSerialization: optionally (?!) group statements and snaks. | ||
| · · · |
I am looking at removing the API output in ungrouped lists
https://gerrit.wikimedia.org/r/#/c/225903/
The feature is barely used, in the last 2.5 days it has been used by 1 user for 50 api calls.
I agree that maintaining the ability to deserialize ungrouped lists is important!
Deserialization of ungrouped lists will be possible using the DM deserializers as of 1.7.0
https://github.com/wmde/WikibaseDataModelSerialization/pull/157