[
https://issues.apache.org/jira/browse/TIKA-491?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14729432#comment-14729432
]
Ken Krugler commented on TIKA-491:
----------------------------------
Currently the language-detector library I'm integrating (see TIKA-1723) only
has support for the 'no' metalanguage, not the two specific dialects as
described above. I'd recommend opening an issue at that project (see
https://github.com/optimaize/language-detector/). So closing this issue, unless
somebody wants to (a) port the current built-in Tika detector to the new
architecture, and (b) follow up with Jan about getting training text, and (c)
add the new profiles. I'll wait a few days.
> Add language identification support for Norwegian Bokmål and Norwegian Nynorsk
> ------------------------------------------------------------------------------
>
> Key: TIKA-491
> URL: https://issues.apache.org/jira/browse/TIKA-491
> Project: Tika
> Issue Type: New Feature
> Components: languageidentifier
> Affects Versions: 0.7
> Reporter: Jan Høydahl
> Assignee: Ken Krugler
>
> Currently there is one Norwegian language profile in Tika - "no". We need to
> distinguish between the two official Norwegian languages defined by ISO 639-1
> codes "nb" and "nn". Those codes are recommended used instead of the common
> "no" tag.
> Proposed solved by removing the current language profile no.ngp and replacing
> it with two new ones for nb and nn.
> We must also add tests for Norwegian
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
