[
https://issues.apache.org/jira/browse/TIKA-257?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jukka Zitting resolved TIKA-257.
--------------------------------
Resolution: Fixed
Fix Version/s: 0.4
Assignee: Jukka Zitting
I found a pretty accurate magic byte pattern (the file name string
[Content_Types].xml at offset 30) for OOXML files. This still doesn't tell
whether the document is a spreadsheet, a presentation or something different,
but at least it's enough to allow Tika to correctly send the document to
OOXMLParser for more detailed processing with POI.
I added the byte pattern and made some related adjustments in revision 793696.
The above test case now passes.
Resolving as Fixed.
> Uncorrect mime-type detection for ooxml
> ---------------------------------------
>
> Key: TIKA-257
> URL: https://issues.apache.org/jira/browse/TIKA-257
> Project: Tika
> Issue Type: Bug
> Components: general
> Affects Versions: 0.4
> Reporter: Maxim Valyanskiy
> Assignee: Jukka Zitting
> Fix For: 0.4
>
>
> MimeTypes detects docx (and other office XML documents) as 'application/zip'
> when file does not have proper extension:
> $ java -jar tika-app/target/tika-app-0.4-SNAPSHOT.jar -m
> /home/maxcom/download-tmp/proto.docx
> Content-Type:
> application/vnd.openxmlformats-officedocument.wordprocessingml.document
> resourceName: proto.docx
> $ cat /home/maxcom/download-tmp/proto.docx | java -jar
> tika-app/target/tika-app-0.4-SNAPSHOT.jar -m
> Content-Type: application/zip
> This breaks text extraction when filename is not known
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
