Oftentimes, metalad_core is the only extractor I'll use to generate a catalog entry. And from [the current] information it is not possible to derive a useful dataset name. ..
In the absence of any extra metadata (e.g. from metalad_studyminimeta), IMO the most likely candidate for a dataset name would be the directory name in the path. So if the path is inside the metadata record, the catalog translator could use that to populate a name field.
Once https://github.com/datalad/datalad-metalad/pull/375 is merged, the corresponding translator in core should be updated to get the directory name and populate the name field in the catalog schema with it.
Comment taken partially from https://github.com/datalad/datalad-metalad/issues/373:
Once https://github.com/datalad/datalad-metalad/pull/375 is merged, the corresponding translator in core should be updated to get the directory name and populate the
namefield in the catalog schema with it.