Some notices have hasAwardRank with a string as its value.

[cristianvasquez]

The following notices were found to contain issues while being loaded into Virtuoso:

• http://data.europa.eu/a4g/resource/2024/00633882_2024
• http://data.europa.eu/a4g/resource/2024/00634694_2024
• http://data.europa.eu/a4g/resource/2024/00634433_2024

The failure occurred due to using strings as values in the hasAwardRank property, despite being declared as integers.

<j.2:hasAwardRank rdf:datatype=http://www.w3.org/2001/XMLSchema#integer>PRIMA</j.2:hasAwardRank>
<j.2:hasAwardRank rdf:datatype=http://www.w3.org/2001/XMLSchema#integer>Prima</j.2:hasAwardRank>
<j.2:hasAwardRank rdf:datatype=http://www.w3.org/2001/XMLSchema#integer>prima</j.2:hasAwardRank>

This appears to be a data quality issue. We need to determine the best approach to handling such mismatches and whether it's done before, during, or after transformation.

[muricna]

Cellar will not load notices that contain concepts with the wrong data type. Therefore if the source xml has a wrong data type it will not be published in Cellar. Since we cannot change the source data one solution could be to simply remove the concept to allow the rest of the notice to be published, otherwise the notice will remain unpublished in TEDSWS.