Closed sitag closed 4 years ago
@juettemann you might need to keep this in mind.
Hello @sitag,
I'm not sure I agree with your reading of the spec. It says: CHIP_PROTOCOL - The ChIP protocol used, or 'Input'.
.
As I understand it, the XML example you provided should not validate.
Cheers,
Daniel
@dzerbino Sorry, I meant nullable attributes like "CHIP_ANTIBODY"
OK, here's one possible solution: each time a property is of type "string"
without any constraints (e.g. enum, regex etc), then we replace "string"
with ["string","null"]
@dzerbino There was some discussion (with @dbujold ) about supporting lists for attributes and not strings, in which case a NA value is an empty list (?) ..are we moving forward with lists or staying with strings?
Hello @sitag,
That's a good point:
Cheers,
Daniel
submitted ["string", "null"] fix. https://github.com/IHEC/ihec-ecosystems/pull/60/commits/15bf5ec2e39d41423894ea19e913ae6c77afb44f
Can we close then?
sure ...assuming it looks good to you
As per https://github.com/IHEC/ihec-ecosystems/blob/master/docs/metadata/1.0/Ihec_metadata_specification.md , things like
should validate. But when we parse this xml, the json gets None for "chip_protocol" and this fails against https://github.com/IHEC/ihec-ecosystems/blob/master/schemas/json/experiment.json#L135 as it is not a string.