[ 
https://issues.apache.org/jira/browse/TIKA-4536?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18034241#comment-18034241
 ] 

Tilman Hausherr commented on TIKA-4536:
---------------------------------------

+1. I've seen code that checks whether an InputStream is a TikaInputStream, 
this will get rid of it.

> Discuss: Consider requiring TikaInputStream for parsers and detectors
> ---------------------------------------------------------------------
>
>                 Key: TIKA-4536
>                 URL: https://issues.apache.org/jira/browse/TIKA-4536
>             Project: Tika
>          Issue Type: Improvement
>            Reporter: Tim Allison
>            Priority: Major
>
> On 4.x, we changed all embedded file processing to use TikaInputStream. This 
> made a bunch of code a lot simpler.
> We could simplify code dramatically if we required TikaInputStream in parsers 
> and detectors.
> Caching to disk is no longer a bottleneck for many users, and users can 
> choose to limit InputStreams by wrapping their InputStreams in a 
> BoundedInputStream or simiar:
> This would be a major breaking change, but would clean up a lot of code.
> WDYT?
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to