[
https://issues.apache.org/jira/browse/JENA-1249?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15609769#comment-15609769
]
ASF subversion and git services commented on JENA-1249:
-------------------------------------------------------
Commit d865a839c9268b5f2e23c223c193f409b887a28c in jena's branch
refs/heads/master from [~andy.seaborne]
[ https://git-wip-us.apache.org/repos/asf?p=jena.git;h=d865a83 ]
JENA-1249 - Don't touch prefixes unless non-empty to be merged.
Cope with initialization ordering effects.
> ModelCom constuction triggers prefix reading in TDB
> ---------------------------------------------------
>
> Key: JENA-1249
> URL: https://issues.apache.org/jira/browse/JENA-1249
> Project: Apache Jena
> Issue Type: Bug
> Reporter: Andy Seaborne
> Attachments: ReportPerf.java
>
>
> For TDB, the Dataset-level wrapper is {{DatasetImpl}}. The model-level
> operations {{Dataset.getDefaultModel}} and {{Dataset.getNamedModel}} use a
> {{ModelCom}} to wrap the graph.
> The {{ModelCom}} construct also calls {{ModelCom.getPrefixMapping}} which in
> turn triggers the TDB graph level prefix mapping to be created.
> Due to JENA-81, the prefix map is read to the cache when the graph starts.
> Overall result - creating the model causes the prefix mappings to be read.
> We should avoid calling {{getPrefixMapping}} inside the {{ModelCom}}
> constructor.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
