[ 
https://issues.apache.org/jira/browse/TIKA-1561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14337878#comment-14337878
 ] 

Nick Burch commented on TIKA-1561:
----------------------------------

Are any of those DIF files you mention under a suitable license where we can 
include them as part of Apache Tika? (eg Apache Licensed, BSD Licensed, Public 
Domain, something like that)

If we can get a very small DIF file with a suitable license, it would be good 
to pop that under {{src/test/resources/test-documents}} then add a unit test 
for DIF detection in 
{{tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java}} , to 
verify that the new mime magic is working

> GCMD Directory Interchange Format (.dif) identification
> -------------------------------------------------------
>
>                 Key: TIKA-1561
>                 URL: https://issues.apache.org/jira/browse/TIKA-1561
>             Project: Tika
>          Issue Type: Improvement
>          Components: mime
>    Affects Versions: 1.7
>            Reporter: Luke sh
>            Assignee: Chris A. Mattmann
>            Priority: Trivial
>         Attachments: 
> carbon_isotopic_values_of_alkanes_extracted_from_paleosols.dif
>
>
> cited from the http://gcmd.nasa.gov/add/difguide/WRITEADIF.pdf 
> "The Directory Interchange Format (DIF) is metadata format used to create 
> directory entries that describe scientific data
> sets. A DIF holds a collection of fields, which detail specific information 
> about the data."
>  The .dif file respect proper xml format that describe the scientific data 
> set, the schema xsd files can be found inside the .dif xml file.
> i,e, http://gcmd.gsfc.nasa.gov/Aboutus/xml/dif/dif_v9.8.4.xsd
> The reason opening this ticket is tika parser for this dif file is being 
> under consideration with development, the support to identify the type of xml 
> file is needed.
> Although dif file in this case seems to be an proper xml file which can be 
> parsed by xmlparser, still it might need a specific process on some of the 
> fields to be extracted and injected into the Solr System for analysis.
> Then it is proposed that the following type 'text/dif+xml' is appended and 
> used in the tika-mimetypes.xml to be able to support the specific xml type 
> detection which extends the application/xml, so that some special process can 
> be applied to this particular xml file.
> <mime-type type="text/dif+xml">
>    <root-XML localName="DIF"/>
>    <root-XML localName="DIF" 
> namespaceURI="http://gcmd.gsfc.nasa.gov/Aboutus/xml/dif/"/>
>    <glob pattern="*.dif"/>
>    <sub-class-of type="application/xml"/>
> </mime-type>
> Expected MIME type: text/dif+xml
> The following is the link to the dif format guide
> http://gcmd.nasa.gov/add/difguide/
> example dif files:
> 1) 
> https://www.aoncadis.org/dataset/id/005f3222-7548-11e2-851e-00c0f03d5b7c.dif
> 2) 
> https://www.aoncadis.org/dataset/id/0091cf0c-7ad3-11e2-851e-00c0f03d5b7c.dif
> 3) 
> https://www.aoncadis.org/dataset/id/02a6301c-3ab3-11e4-8ee7-00c0f03d5b7c.dif
> an example dif file has also been attached.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to