@rreitmann @AndyDaniel1 I think this would be a good candidate for the next sprint. question-image jsons are still not valid and contain irrelevant stuff. Not sure about question-jsons.
You're completly right in stressing the importance of this topic. If you hadn't prioritised this issue with 1 I would've done it. Still I need an update on the state of zofar import topic in general.
@rreitmann fyi our feedback for the zofar question export of gra2005.3 https://github.com/dzhw/zofar/issues/143#issuecomment-482171084
The following list describes the reasons why we currently use https://dzhw.github.io/questionMetadataPreparation and do not directly import the zofar exprt:
conceptIds hence they are missing in the question jsons and we need to add them"resolution": { "widthX": 1200, "heightY": 800}imageType, file, questionId, dataAcquisitionProjectId should not be present cause they are filled during import by the MDMid, dataAcquisitionProjectId, studyId should not be present cause they are filled during import by the MDMquestionNumber (name of the json) is not consistent with the number in the instrument pdf (see https://github.com/dzhw/zofar/issues/143#issuecomment-482171084), we correct the questionNumber in the handcrafted excel~@UteH Have you informed Zofar about the requirements in my previous comment?
@rreitmann
No. As I recall, we had agreed to just list the problems but not investigate further (yet).
I can provide the most current export for nacaps2018.2. You are welcome to review it an give me the most currenct version of _your_ requirements. I can then ask zofar about it as we will be in contact soon.
(However, I will get in touch regarding https://github.com/dzhw/metadatamanagement/issues/2606 as we discussed)
@UteH I had a quick look on the Nacaps export. I have updated my https://github.com/dzhw/metadatamanagement/issues/1412#issuecomment-511739404. The current issues are all mitigated by converting the zofar export to the handcrafted one. IMHO there will always be the need for manually correcting the zofar export (at least due to conceptIds) and therefore there is no need to fix this on the Zofar side.
@rreitmann thanks for the quick review. So I could ask them to omit/change the remaining attributes mentioned above (but we do not depend on it)
conceptIds will not be altered as it has to be filled by us
@UteH if they change the resolution we would need to change questionMetadataPreparation. It is a minor change but we would need to know when they change it...
Most helpful comment
The following list describes the reasons why we currently use https://dzhw.github.io/questionMetadataPreparation and do not directly import the zofar exprt:
conceptIdshence they are missing in the question jsons and we need to add them"resolution": { "widthX": 1200, "heightY": 800}imageType,file,questionId,dataAcquisitionProjectIdshould not be present cause they are filled during import by the MDMid,dataAcquisitionProjectId,studyIdshould not be present cause they are filled during import by the MDMquestionNumber(name of the json) is not consistent with the number in the instrument pdf (see https://github.com/dzhw/zofar/issues/143#issuecomment-482171084), we correct the questionNumber in the handcrafted excel~