DIG-2209: Changing /experiments to /biosamples#15
Conversation
mshadbolt
left a comment
There was a problem hiding this comment.
I couldn't understand why the endpoint says it returns arrays but the schema specifies an object. I thought the object types should be consistent not vary amongst different endpoint returns.
|
The result is something like So |
mshadbolt
left a comment
There was a problem hiding this comment.
I think the experiments lists should be unique right? It looks like if there are multiple analyses per experiment it adds the experiment to the list twice e.g.:
{
"analyses": {
"reference_alignment": [
"local-NA02102-cram",
"local-NA02102-bam"
]
},
"biosample_id": "local-SAMPLE_0004",
"experiments": {
"wgs": [
"local-SEQ_0004",
"local-SEQ_0004"
],
"wts": []
},
"program": "local-SYNTH_01",
"runs": []
},
Co-authored-by: Marion <mshadbolt@users.noreply.github.com>
|
I think the problem is that the ingest test file has a duplicate experiment in it; I don't know why. If it's not supposed to have a duplicate (maybe we meant it as a test that shouldn't be double-ingested?), we should fix the problem in the test data. |
|
If we don't want to allow duplicate ingest (which we probably shouldn't), we need a ticket to fix ingest to catch that error. |
|
The double-adding thing was a good catch! It is fixed now. |
mshadbolt
left a comment
There was a problem hiding this comment.
tests passing, working as expected
We are changing the /experiments endpoint because it was confusing: it does not return something that is anything like the ENA model's experiment. We are refactoring that endpoint to something that corresponds more closely to that model.
Testing should be performed in CanDIG/CanDIGv2#1252.