[
https://issues.apache.org/jira/browse/TIKA-1434?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Marco Quaranta updated TIKA-1434:
---------------------------------
Attachment: ascii_file.txt
> Plain text file reported as binary
> ----------------------------------
>
> Key: TIKA-1434
> URL: https://issues.apache.org/jira/browse/TIKA-1434
> Project: Tika
> Issue Type: Bug
> Components: detector, mime
> Affects Versions: 1.6
> Reporter: Marco Quaranta
> Priority: Minor
> Attachments: ascii_file.txt
>
>
> Mime type detection of the attached file, an ASCII file, reports
> application/octect-stream. The negative result is caused by ASCII-extended
> characters in the file; TextStatistics.isMostlyAscii() calls return false. We
> could try to call TextDetector.detect() two times: if the first call fails,
> then we can try to convert bytes in utf-8 and call it again. What do you
> think?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
