[
https://issues.apache.org/jira/browse/STANBOL-56?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Fabian Christ updated STANBOL-56:
---------------------------------
Priority: Minor (was: Major)
> [RICK] Implement Caches for ReferencedSites
> -------------------------------------------
>
> Key: STANBOL-56
> URL: https://issues.apache.org/jira/browse/STANBOL-56
> Project: Stanbol
> Issue Type: Improvement
> Components: RICK
> Reporter: Fabian Christ
> Assignee: Rupert Westenthaler
> Priority: Minor
>
> Reported by project member rupert.westenthaler, Nov 16, 2010
> ReferencedSites can define a cache strategy.
> NONE ... means that no cache should be used
> USED ... indicated, that Entities loaded from the site should be cached
> locally
> ALL ... indicated, that the Cache holds all entities of the referenced site.
> The caching functionality will be implemented based on the Yard interface.
> This means that all the cached data will be stored within a Yard (The Yard is
> the storage component of the RICK)
> To give the user control about the information to be caches the already
> existing mapping functionality needs to be extended.
> Users need to be able to define
> - fields to be cached
> - languages to be cached for human language texts like labels and
> descriptions.
> - data type restrictions (e.g. one might not want to cache string values for
> latitude and longitude)
> - data type conversions (e.g convert the string 2,576 to a double)
> This functionality should be implemented based on the mapping functionality
> already provided by the "eu.iksproject.rick.servicesapi.mapping" package.
> Cache need to keep there own configuration. To allow easy copy&paste
> operation with pre-computed caches such configurations MUST BE stored within
> the Cache.
> A change to the configuration of the cache normally needs to recompute every
> item in the cache. This might not be feasible especially for FULL caches.
> However it might still be useable to allow users to define additional fields
> that can be cached for entities downloaded form the site.
> To allow that Caches need two configurations:
> - base: Information present for each entity stored by the cache
> - additional: Information that are cached if entities are updated in the
> cache.
> Caches need to analyse queries if they can be executed and/or answered by
> using the cache. For this the base configuration MUST BE used.
> - queries can be executed if all fields with constraints are present in the
> cache
> - queries can be answered if all selected fields are stored in the cache
> Comment 1 by project member rupert.westenthaler, Nov 20, 2010
> Revision 1074 added
> - support for Caches
> - base/additional Mappings
> - language filters
> - data type filters
> - data type conversions
> Not Implemented:
> - analyzing queries if the can be executed based on the cache if cache
> strategy is FULL
> - currently all queries are executed on the cache if cache strategy is FULL
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
