xhqiao89
commented
7 years ago kimschreuders[CodePlex]
I made this comment on issue #13, but it applies here as well:
One other thought, you need to be careful when updating the metadata catalog that the associated tagging gets handled properly. The tagging needs to be carefully managed to make sure that it is updated appropriately so that the tagging is handled by one of these method, i.e. if series identifiers are changed: 1) tagging that can be saved (i.e. it is clear that tags for a particular old series should be transfered to a particular new series) should be pointed at the new identifiers, 2) tagging that cannot be saved should be either deleted or indicated as orphaned and messages should go out to the administrators of the data set letting them know that their series are not tagged and/or that they have orphaned tags that need to be cleaned up.
kimschreuders[CodePlex]
I performed this search in HydroDesktop:
Area: Cache County Date: 5/27/2911 - 5/27/2011 Services: Little Bear River Experimental Watershed Keywords: Hydrosphere
and looked at the results with the Attribute table tool, it showed 574 series returned. However, 287 of them had a Vocabulary Code of quotLBRquot and 287 had a Vocabulary Code of quotLittleBearRiverquot, and each of these 287 records seemed to be duplicates of the other 287 records. The quotLBRquot all had an EndDate of 9/24/2010 or earlier. Many of the quotLittleBearRiverquot records have an EndDate of 5/27/2011 (today).
It looks to me that all this means that Jeff changed the Vocabulary Code on 9/24/2010 from quotLBRquot to quotLittleBearRiverquot and the metadata harvester at SDSC has a bug that causes it to not delete existing metadata catalog entries for matadata that no longer exists. So there are now metadata entries are in the central catalog for both the old series with quotLBRquot and the current series with quotLittleBearRiverquot.
If this is the case, then 2 things need to be fixed. 1) the duplicate entries need to be removed from the catalog. 2) the harvester needs to be fixed so that it deletes metadata from the catalog for series that no longer exists so that we do not have this problem again.