[M] (DI-2-8) Moles-to-MDIP transform needs extending

[selatham]

The Moles-to-MDIP and MDIP-to-Moles transforms are not picking up very much. Need more content.

Rough Priority:-

• DatasetIdentifier?
• Target keyword
• Title
• Abstract
• Coordinates
• OnlineResource?
• Originator
• Temporal info(do they have any? is this dataset start and end dates?)
• Other names
• Other keywords
• Constraints
• Other spatial stuff (vertical extent etc.)

[ko23]

OK, extended to include:

• DatasetIdentifier?
• Target keyword
• Title
• Abstract
• bounding box aka CoOrdinates
• Dataset temporal extremes

[mpritcha]

Problem coverting longitude values? Is is trying to cast them as integers?

org.exist.xquery.XPathException: failed to convert '-12.0' to an integer: For input string: "12.0" [at line 1, column 4840]

This was attempting to covert the DIF for
neodc.nerc.ac.uk__DIF__NEODC_ARSF_ATM_DAED.xml into MDIP. "-12.0" seems to be the <Westernmost_Longitude>

[ko23]

Quite right, should be cast as decimal. Fixed.

[selatham]

Warning that MDIP will be getting Edina to 'finish off' their work on the front-end. They have not used MDIP format records yet - none were available. Now that they are available they will be utilising them. This will include BODC records which will be being transformed from DIF. So the transform needs to be sufficient for this requirement. They will probably flag up issues with Matt in the first instance.

[selatham]

Remaining extension is lower priority than DIF or ISO work.

[sdonegan]

Have added much content to MDIP Xqueries, just need to check that we are getting out everying we should be..

[sdonegan]

Now returning original format documents to providers - where MDIP was original input format the original doc is returned. This has solved a content problem with the MDIP portal.