Term labels are formatted like byte strings rather than strings, i.e. b'label'

[dhimmel]

Copying part of the JSON result from https://www.ebi.ac.uk/spot/oxo/api/search?ids=EFO:0009623 below:

{
  "_embedded" : {
    "searchResults" : [ {
      "queryId" : "EFO:0009623",
      "querySource" : null,
      "curie" : "EFO:0009623",
      "label" : "nose injury",
      "mappingResponseList" : [ {
        "curie" : "NCIT:C35349",
        "label" : "b'Nose Injury'",
        "sourcePrefixes" : [ "EFO" ],
        "targetPrefix" : "NCIT",
        "distance" : 1
      } ],

Notice the string value for label is b'Nose Injury' rather than the expected Nose Injury.

I haven't looked at the code, but my guess is that labels are actually byte string objects rather than unicode string objects. And when the label object is converted to a string as part of the JSON export, the b'label' notation is applied.

Can the byte strings be decoded to unicode strings prior to the JSON export (or any other option that fixes this)?

[matentzn]

I could swear I fixed this issue (btw, its due to python 3 vs 2 incompatibility and changes in the way string byte arrays are handled), maybe for the browser, and not for the API; cant find my commit :P ? Its important though so I agree this should be chased up.. Thanks for the report!

[rpandey19]

Any update on this. I am seeing the same issue. It shows up on the local installation I have of OXO

[matentzn]

I dont think so! Sorry.