[TIKA-2064] Document type detected incorrectly for Stata datasets (.dta extension)

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.13
• Component/s: detector
• Assignee: Unassigned
• Reporter: Michael Stepner
• Created: 29/Aug/16 03:39
• Updated: 13/Sep/16 20:55

Description

The content type of Stata datasets (created using http://www.stata.com software) is incorrectly detected as `text/html` by Tika. I have tested this using the latest release of Tika, v1.13:

```
$ curl -O http://www.stata-press.com/data/r14/auto.dta
$ java -jar tika-app-1.13.jar --detect auto.dta
text/html
```

I believe that the type should instead be `application/octet-stream` (or the equivalent).

I originally reported this bug downstream (at https://github.com/laurilehmijoki/s3_website/issues/232), and was advised to report upstream to Tika. In addition to the one I downloaded using `curl` in my example, a variety of reference Stata datasets are posted here: http://www.stata-press.com/data/r14/dmain.html