[ 
https://jira.duraspace.org/browse/DS-1080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=23342#comment-23342
 ] 

Mark Diggory commented on DS-1080:
----------------------------------

Our design choices in the previous version of Discovery to use the dri 
referenceset/references was so that we could reuse the existing processing 
approach that was created in the legacy Manakin ArtifactBrowser. This would 
allow existing instances to turn on Discovery and render facets and still 
maintain their existing search results presentation they may have customized 
int he theme.

But, while there were benefits in this approach for the endusers, there are 
also limitations. Namely, the lack of exposing any of the excellent solr 
capabilities for results rendering into the discovery presentation. Basically 
anything that was available as detail from the original Solor/Lucene search is 
lost because the context of the XSLT rendering because its neither added by the 
transformer, likewise, the xslt processing shifts entirely into the METS 
document and the context of the reference in the referenceset is entirely lost 
in the processing.  Basically, even if we render the solr results into the dri, 
we still need to change the processing approach to allow the xmlui to render 
these details along with the reference set details.

I want to see this corrected by extending the DRI to support xlinking to xml 
resources that should be included into rendering. This approach will make the 
whole concept of mets inclusion of dri:references entirely more generic and 
give us a mechanism for referencing xml content from almost anywhere to be made 
available for processing in the xslt.

If we take for instance the following:


{code}
<referenceSet 
id="aspect.discovery.SimpleSearch.referenceSet.search-results-repository" 
rend="repository-search-results" n="search-results-repository" 
type="summaryList">
<reference repositoryID="10211.6" type="DSpace Item" 
url="/metadata/handle/10211.6/166/mets.xml"/>
<reference repositoryID="10211.6" type="DSpace Item" 
url="/metadata/handle/10211.6/261/mets.xml"/>
<reference repositoryID="10211.6" type="DSpace Item" 
url="/metadata/handle/10211.6/271/mets.xml"/>
{code}

And we renader as the following instead.

{code}

<list id="aspect.discovery.SimpleSearch.referenceSet.search-results-repository" 
rend="repository-search-results" n="search-results-repository" 
type="summaryList">
   <head>Discover</head>
   <item repositoryID="10211.6" type="DSpace Item" 
reference="/metadata/handle/10211.6/166/mets.xml">
        <str name="description">britta is awesome</str>
        <str name="id">poddObject:67</str>
   </item>

{code}

likewise, we will want to process the results of the highlighting from solrj 
back into an xml form that can be used as a source for rendering.

Thus the following output, in fact any output from solr would be rendered into 
DRI lists.  Thus folks who understnd solr results processing will immediately 
comprehend this format.

{code}

<lst name="highlighting">
<lst name="10255/dryad.8252">
<arr name="dc.description">
<str>
cryptic fungal <em>species</em> in the lichen-forming rhizoplaca melanophthalma 
<em>species</em>-complex (lecanoraceae
</str>
</arr>
<arr name="dc.title">
<str>
cryptic fungal <em>species</em> in the lichen-forming rhizoplaca melanophthalma 
<em>species</em>-complex (lecanoraceae
</str>
</arr>
<arr name="dc.subject">
<str><em>species</em> delimitation</str>
</arr>
</lst>


{code}

Would be rendered to:

{code}

<list n="highlighting">
<list n="10255/dryad.8252">
   <list n="dc.description">
        <item type="str">cryptic fungal <em>species</em> in the lichen-forming 
rhizoplaca melanophthalma <em>species</em>-complex (lecanoraceae</item>
   </list>
   <list n="dc.title">
        <item type="str">cryptic fungal <em>species</em> in the lichen-forming 
rhizoplaca melanophthalma <em>species</em>-complex (lecanoraceae</item>
   </list>
   <list n="dc.subject">
        <item type="str"><em>species</em> delimitation</item>
   </list>
</list>
...
</list>

{code}

So, to finalize the resulting output in DRI would be capable of housing the 
solr results as a data structure made available to the theming tier to expose 
rendering additional detail.
                
> Search results preview
> ----------------------
>
>                 Key: DS-1080
>                 URL: https://jira.duraspace.org/browse/DS-1080
>             Project: DSpace
>          Issue Type: New Feature
>          Components: Discovery, XMLUI
>            Reporter: Howard Shand
>            Assignee: Mark Diggory
>            Priority: Major
>         Attachments: DRI-example.xml, example.jpg
>
>
> Show a preview of search results.
> When search results are based on full text search, show the keywords in their 
> context (with some highlighting) similar to the functionality of most search 
> engines.
> Example of the system we were previously using is attached.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://jira.duraspace.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Dspace-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dspace-devel

Reply via email to