[jira] [Commented] (TIKA-3023) Text files starting with MOVI are detected as X-SGI-Movie

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Commented] (TIKA-3023) Text files starting with MOVI are detected as X-SGI-Movie

Sebastian Nagel (Jira)

    [ https://issues.apache.org/jira/browse/TIKA-3023?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17011634#comment-17011634 ]

Nick Burch commented on TIKA-3023:
----------------------------------

Assuming that the byte after MOVI is part of a version or length field, perhaps safest to check for 0x00 or 0x01 or 0x02? That shouldn't trigger on text, but probably would help with the detection of the real thing. (Some sort of "not a printable ascii field" is probably a step too far and probably wouldn't help much either!)

> Text files starting with MOVI are detected as X-SGI-Movie
> ---------------------------------------------------------
>
>                 Key: TIKA-3023
>                 URL: https://issues.apache.org/jira/browse/TIKA-3023
>             Project: Tika
>          Issue Type: Bug
>    Affects Versions: 1.23
>         Environment: Issue recreated on
> Windows 10 Professional 64bit running the runnable Jar
> Ubuntu 16.04.6 LTS running Tika-Python
>            Reporter: Steve
>            Priority: Minor
>         Attachments: capitalmovie.txt
>
>
> If a plaintext file starts with "MOVI" Tika labels it as an SGI Movie.
> The hex conversion for MOVI is 4D 4F 56 49 which is the same as the header for the SGI Movie file format
> [https://reposcope.com/mimetype/video/x-sgi-movie]
>  
> This SGI format isn't supported so any information from a text file starting like this would be lost. I've attached a simple file that should recreate the problem.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)