Hi,
On Sat, Oct 17, 2009 at 2:43 PM, mastcheshmi
<[email protected]> wrote:
> I use Tika
> for all document this exception occured.
> org/apache/poi/hpsf/MarkUnsupportedException
That seems like a bug. I guess POI expects the given document stream
to support the mark feature, so Tika should explicitly wrap the stream
into a java.io.BufferedInputStream if it does not already support
marks.
Can you please file a bug report [1] about this? Meanwhile, as a
workaround you can wrap the input stream yourself:
new OfficeParser().parse(new BufferedInputStream(in), handler, metadata);
[1] https://issues.apache.org/jira/browse/TIKA
BR,
Jukka Zitting
