Closed jdries closed 1 year ago
Current parsing of these search results happens here: https://github.com/Open-EO/openeo-opensearch-client/blob/master/src/main/scala/org/openeo/opensearch/OpenSearchResponses.scala#L280
Solving this issue will also help fix this one: https://github.com/Open-EO/openeo-geopyspark-driver/issues/229
need to wait before tested in deployment before closing
This creodias query returns two products that are actually the same:
https://finder.creodias.eu/resto/api/collections/Sentinel1/search.json?maxRecords=10&startDate=2021-04-11T00%3A00%3A00Z&completionDate=2021-04-11T23%3A59%3A59Z&productType=GRD&sensorMode=IW&geometry=POLYGON((5.785404630537803+51.033953432779526%2C5.787426293119076+51.021746940265956%2C5.803195261253003+51.018694814851074%2C5.803195261253003+51.02912208053834%2C5.785404630537803+51.033953432779526))&sortParam=startDate&sortOrder=descending&status=all&dataset=ESA-DATASET
S1B_IW_GRDH_1SDV_20210411T054146_20210411T054211_026415_032740_6184 S1B_IW_GRDH_1SDV_20210411T054146_20210411T054211_026415_032740_EBF8
If we forward this into openEO, we will process the same product twice. I propose to only return the most recent product, by publication date. To find duplicates, we cannot look at the product ID, because this is in fact different, so we need to look at all other properties.
Design wise, we could add this as a kind of postprocessing function that can optionally be applied to the results of a catalog search.