[
https://issues.apache.org/jira/browse/TIKA-2084?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15504321#comment-15504321
]
Luis Filipe Nassif commented on TIKA-2084:
------------------------------------------
I think the reset could be optional, because some cases the first parser, even
throwing an exception, can extract valuable content, for example, when the
exception is thrown while parsing the last page of a docx or pdf (when the flag
to catch expections per page is not set)
> Create resettable OutputStream to support "backoff on exception" strategy
> -------------------------------------------------------------------------
>
> Key: TIKA-2084
> URL: https://issues.apache.org/jira/browse/TIKA-2084
> Project: Tika
> Issue Type: New Feature
> Components: core
> Reporter: Tim Allison
>
> If we want a backoff on exception strategy, "try xmlparser, if that fails,
> try the TXTParser", we'll need to have a resettable
> outputstream/contenthandler to clear what had been written by the first
> parser.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
