Closed sentry-io[bot] closed 11 months ago
I'm pretty sure we've reported this issue already in the past, but I am tagging @allakulov as we are discussing improvements to their OCDS implementation again
So yes, it only happens a few times and the format is always as you described. For the record, this is the full list of release ids without an ocid:
ocds-hu01ve-7608611-01 ocds-hu01ve-7446430-01 ocds-hu01ve-7442999-01 ocds-hu01ve-7450526-01 ocds-hu01ve-7403060-01 ocds-hu01ve-7431124-01 ocds-hu01ve-7419421-01 ocds-hu01ve-7437826-01 ocds-hu01ve-7460648-01 ocds-hu01ve-7461737-01 ocds-hu01ve-7452379-01 ocds-hu01ve-7499991-01 ocds-hu01ve-7375344-01 ocds-hu01ve-7408813-01 ocds-hu01ve-7473526-01 ocds-hu01ve-7486516-01 ocds-hu01ve-7411328-01 ocds-hu01ve-7473453-01 ocds-hu01ve-7456352-01 ocds-hu01ve-7460480-01 ocds-hu01ve-7465257-01 ocds-hu01ve-7472880-01 ocds-hu01ve-7467979-01 ocds-hu01ve-7395176-01
Okay, so can we make the spider fill in the OCID if missing, then close this issue?
Yeap, I'm already on it :)
(@jpmckinney writing)
ANAC omits OCIDs occasionally, but the IDs might be based on the OCID e.g. this ID is
ocds-hu01ve-7467979-01
, so maybe the OCID can be assumed to beocds-hu01ve-7467979
? I haven't checked against their other data, but if that's the case, we should fill in the OCID in the spider (and report the issue to ANAC, if we haven't already).This causes errors to be logged by Kingfisher Process when attempting to merge releases without OCIDs.
Sentry Issue: REGISTRY-KINGFISHER-PROCESS-6P