[
https://issues.apache.org/jira/browse/TIKA-61?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12732776#action_12732776
]
Jukka Zitting commented on TIKA-61:
-----------------------------------
I would actually rather resolve this as Won't Fix until we have a clear use
case where we actually do need this. So far our parsers extract so little
metadata that namespacing seems overkill. And once we really start parsing and
exposing things like XMP, I'm afraid that we need to come up with a more
complex metadata model than just (namespaced) name-value pairs.
> Add namespaces to our metadata keys
> -----------------------------------
>
> Key: TIKA-61
> URL: https://issues.apache.org/jira/browse/TIKA-61
> Project: Tika
> Issue Type: Improvement
> Components: metadata
> Affects Versions: 0.1-incubating, 0.2, 0.3, 0.4
> Reporter: Bertrand Delacretaz
> Assignee: Chris A. Mattmann
> Priority: Minor
> Fix For: 0.5
>
>
> To avoid collisions, we should namespace our metadata keys, using something
> like
> // a) Set Dublin Core title
> metadata.set("http://purl.org/dc/elements/1.1/";, "title", title);
> // b) Set OpenDocument keywords
> metadata.set(
> "urn:oasis:names:tc:opendocument:xmlns:meta:1.0",
> "keywords", keywords);
> Or also using the namespace prefix in the name:
> // c)
> metadata.set(
> "urn:oasis:names:tc:opendocument:xmlns:meta:1.0",
> "dc:keywords", keywords);
> not sure what the implications of c) are. If the key is composed of the
> namespace URL and item name we should be fine, so for case a) I'd use
> http://purl.org/dc/elements/1.1/#title
> for the key, and escape # if it is used in either the namespace URL or item
> name.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
