[GeoNode-users] Metadata partly parsed.

[Tom Kralidis]

Hi Margherita:

On Mon, 18 Jan 2016, Margherita Di Leo wrote:

> Date: Mon, 18 Jan 2016 10:12:00 +0100
> From: Margherita Di Leo <diregola at gmail.com>
> To: Tom Kralidis <tomkralidis at gmail.com>
> Cc: "geonode-users at lists.osgeo.org" <geonode-users at lists.osgeo.org>
> Subject: Re: [GeoNode-users] Metadata partly parsed.
> 
> Tom,
>
> On Sun, Jan 17, 2016 at 7:48 PM, Tom Kralidis <tomkralidis at gmail.com> wrote:
>
>> On Fri, Jan 15, 2016 at 5:04 AM, Margherita Di Leo <diregola at gmail.com>
>> wrote:
>>> Dears,
>>>
>>> we would like to deploy our data using GeoNode. However, at present I am
>>> finding multiple issues that i'm not sure I'm able to overcome. I'd
>> probably
>>> better start multiple thread on each issues. I depend upon your kind
>> help.
>>> First issue that I found is that, our metadata, even though 100% INSPIRE
>>> compliant, seem not to be parsed correctly by GeoNode. The only info
>> that I
>>> can visualize after uploading are title, publication date, category and
>>> owner. In the section Abstract appears "No abstract provided", but the
>>> abstract was provided. You can download a sample metadata for testing
>> here .
>>>
>>
>> GeoNode uses the OWSLib package to parse the ISO metadata.  Looking deeper
>> at your metadata sample, your abstract is cast as:
>>
>> <gmd:abstract>
>>   <gmx:Anchor ...>
>> ...
>> </gmd:abstract>
>>
>> OWSLib looks for the following construct:
>>
>> <gmd:abstract>
>>   <gco:CharacterString>
>> ...
>> </gmd:abstract>
>>
>> so in your case gmx:Anchor is not being picked up as an abstract.
>>
>> This is the first time (INSPIRE-based and elsewhere) that I've seen this
>> design pattern.  Is it possible to update your metadata as a workaround?
>> At
>> the same time it would be worth to investigate in OWSLib and possibly apply
>> a issue/patch/workaround.
>>
>
> Thank you for the pointers. Anchor should (and in some catalogues does)
> allow the nice property to have a clickable link in place of the chunk of
> text (the abstract in this case) that lands to a html page. We have used
> this feature because we want to be able to include a long text in the
> abstract and lineage, and new lines are not recognised in xml, resulting in
> a blob of text which is barely readable. I know this is not much used,
> partly because not known and also because usually metadata editors normally
> do not include pro features ;-) but we find it really useful. I think it is
> worth to investigate at the level of OWSLib if a patch could be applied.

Sounds good.  If you/someone can open an OWSLib ticket I can take a look.

..Tom

>
> Thank you again,
> Best,
>
>>
>> ..Tom
>>
>>> Thank you in advance
>>> Cheers,
>>>
>>> --
>>> Margherita Di Leo
>>
>
>
>
> -- 
> Margherita Di Leo
>