Closed alexhuang091 closed 3 years ago
Arguably this isnt really an issue. If the user requests columns for fields we don't have indexed, then blank column values seems one acceptable outcome. The other outcome would be to throw an error. A case can be made for both, but I'm happy with the current functionality.
run this query
https://nci3-biocache-service.ala.org.au/ws/occurrences/offline/download?disableAllQualityFilters=true&fields=decimalLatitude,decimalLongitude,eventDate,scientificName,taxonConceptID,recordID,dataResourceName&qa=alex,nick&emailNotify=false&sourceId=2004&reasonTypeId=10&email=hua091@csiro.au&dwcHeaders=true&fq=(lsid:urn:lsid:biodiversity.org.au:afd.taxon:51dc6bc2-9165-4fc3-a374-1010ac3037d0)&fq=(year:"1930")
with
qa=alex,nick
in SearchDAOImpl::addPostProcessingFields,
alex
is mapped toalex
,nick
is mapped tonick
so there will be 1 column foralex
and 1 column fornick
so
fieldMappingUtil.translateFieldValue
needs a change to filter out all invalid qa values?