Multimedia mapping for Audubon Core is confusing in the IPT

[ManonGros]

There are two fields for "rights", "source", "creator", "format" fields. See example below:

And it seems that when those "duplicated" fields are mapped to data, the extension isn't being ingested and the images don't show on GBIF.org. If the data can't be ingested, could the IPT give a warning? Alternatively, we can make sure that the data is ingested? Or maybe those fields should be twice in the IPT mapping UI?

@MattBlissett what do you think?

[MattBlissett]

The AC standard includes these pairs of terms. One is for string data like Image (for dc:type), and the other for a URI like http://purl.org/dc/dcmitype/StillImage (dcterms:type).

We should probably add the dataType to our XML definition of the extension and then expose this in the IPT, if it isn't already.

Do you have an example of the failing dataset?

[ManonGros]

This dataset's version wasn't displaying the new images: http://patrimonio.ambiente.gob.ec/iptmae/resource?r=varamientos-ecuador&v=2.7

[vechocho]

I was converting some agouti.eu export to DwC and noticed that the converted file uses "dcterm:rights" and "dc:type" as field name, after changing the names on my dataset the images are now indexed.

On the Audubon Media Descripotion the fields are named "source", "rights" and so on so i was using that as names https://rs.gbif.org/extension/ac/audubon_2020_10_06.xml

[peterdesmet]

@vechocho can you indicate what the original term names were in your agouti.eu export and how you fixed them? We want to resolve this in the camtraptor write_dwc() function, so it is resolved for anyone using this function.

[vechocho]

@peterdesmet The Agouti export was fine, i used the workflow that you send me (https://github.com/inbo/mica-occurrences/blob/master/datasets/mica-agouti-occurrences/src/dwc_mapping.Rmd) to convert to darwin core (i attach the result).

I use de Agouti converted dataset as template for my stranding dataset.

dwc_audubon - copia.csv