[
https://issues.apache.org/jira/browse/TUSCANY-2552?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Phillipe Ramalho updated TUSCANY-2552:
--------------------------------------
Attachment: tuscany_2552_phillipe_ramalho_07_04_2009.patch.txt
Second patch, search functionality implemented. How the results are displayed
on the domain manager webpage is still a working in progress.
> Implement a search function for the domain manager home page
> ------------------------------------------------------------
>
> Key: TUSCANY-2552
> URL: https://issues.apache.org/jira/browse/TUSCANY-2552
> Project: Tuscany
> Issue Type: New Feature
> Components: Java SCA Domain Management
> Reporter: Jean-Sebastien Delfino
> Assignee: Adriano Crestani
> Fix For: Java-SCA-Next
>
> Attachments: tuscany_2552_phillipe_ramalho_06_04_2009.patch,
> tuscany_2552_phillipe_ramalho_07_04_2009.patch.txt
>
>
> I'd like to have the ability to search artifacts in an SCA domain. The
> current domain manager app home page shows a search field + button but that's
> currently not implemented yet.
> I suggest the following, which should be easy to implement:
> - use Apache Solr to index all text artifacts in contributions
> - implement the search function in a new optional module (to not overload the
> domain manager, which should remain easily embeddable)
> - start with a simple text indexing scheme (don't need to develop a
> structured Solr schema in a first stage, until we better understand the
> actual usage patterns)
> - when a contribution (JAR or folder) is added to a domain, index the
> artifacts in it
> - when a contribution is requested from the domain manager and we get the
> opportunity to detect that it contribution has changed (that happens when a
> node starts and requests artifacts from the domain manager for example), add
> the new content to the index
> - every 5/10 mns, list the contributions in the domain, check if they have
> changed, and add the new content to the index
> - initially no need to handle deletes, as it may actually be a good thing to
> keep all the history indexed, just keep a copy of the indexed content in the
> index itself or some form or archive.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
