[OWSLib-devel] Couple questions about owslib.iso:MD_Metadata

[Tom Kralidis]

________________________________
> From: ryan.clark at azgs.az.gov 
> Date: Wed, 13 Mar 2013 11:05:59 -0700 
> To: owslib-devel at lists.osgeo.org 
> CC: steve.richard at azgs.az.gov 
> Subject: [OWSLib-devel] Couple questions about owslib.iso:MD_Metadata 
>  
> I tried to map out for myself the relationship between the  
> owslib.iso:MD_Metadata class and the XPaths in ISO XML. I was trying to  
> get my head around what owslib supports, where it supports  
> multiplicity, and that kind of thing. I wrote it out as a json object  
> and might iterate on it once or twice. If it's every helpful to anyone  
> else, its here: https://gist.github.com/rclark/5154432 
>  

Thanks for this, very valuable!  Wow, given a driver like this, I could see how one could parse on the fly :)

> I came away with two questions: 
> - How come some items are recorded as dict and other's get dedicated  
> classes? For example, srv:containsOperations and  
> gmd:descriptiveKeywords end up in dict objects, but there's a CI_Date  
> class. 

iirc, the code initially came out of tighter requirements/use case, and building it out over time has resulted in the inconsistencies.  The whole owslib.iso.MD_Metadata model would benefit from a once over.  Any proposals/ideas/pull requests are most certainly welcome.

> - It looks like there's discrepancy in how extents are handled between  
> MD_DataIdentification and SRV_ServiceIdentification. If I'm reading it  
> right, I don't think the SRV_ServiceIdentification case will even work  
> properly. Should I adjust to use the same logic in the two situations? 
>  

Not sure I fully understand, don't they both use EX_Extent?  Having said this, pull requests to unify them more would be greatly appreciated.

..Tom