I think the field study.dataAvailability could be more fine grained to 100% correspond with dara, because they have 5 fields and our three fields only correspond to 2 of the five (delivery and not available). Not sure whether that's necessary though.
Also, see #1780
edit: dara:
availabilityType | definition
-- | --
Download | Released for everybody.
Delivery | Can be delivered.
On-site | Can be used on-site only.
Not available | Not available.
Unknown | No information is provided.
we will have to deliver availability to VFDB
(optional k枚nnen weitere Angaben im Freitextfeld availabilityFree gemacht werden, Semikolon als Trennzeichen)
I think we could just map our new system to the VFDB.
Generally we need to discuss the meaning and processing of study.dataAvailability. How does it correspond to release.version? Can we set it automatically? Do we need more values?...
We do not want to use the availabilityTypes from Dara cause they mix up accessWay and availability.
The idea of setting dara's availabilityType automatically for versions >= 1.0.0 came up again. Currently we think that we do not need this field in the study domain object.
Currently @fbaillet has found three studies which are set to unavailable although they are released with a version >= 1.0.0 ...
I think the studies are labeled as "not available" in the MDM? So I guess we just have to relabel?
Yep, but that would not be necessary if we simply removed the field study.dataAvailability. IMHO it is an user error we can prevent...
I think we wanted to use that for releases < 1.0.0, for example, if a project is "in Aufbereitung" (~which has not occurred as of today~).
If we do not use this field anymore, we need to give this information to the users in study.annotations.
I just changed "not available" to "available" for gsl2012 and gsl2015 (I forgot to change it before releasing it)
I think we wanted to use that for releases < 1.0.0, for example, if a project is "in Aufbereitung" (which has not occurred as of today).
We used it for bst02 (If I remember correctly)
The cases mentioned are just proving my statement:
version < 1.0.0 => In Aufbereitung (and not sent to da|ra anyway)
version >= 1.0.0 => data available
shadow copy.hidden => data unavailable
We do not need an extra field...
We need to delete the field from mongo and reindex elasticsearch:
db.getCollection('data_packages').updateMany({}, {$unset: {dataAvailability: ''}})