Great, thanks for the URL, I'll check that.

I was wondering if maybe Tika would be an overkill solution to my specific case. We don't index PDF, DOC or anything like that, just plain HTML.


I mean, if everything Tika does is to extract text from HTML, maybe I could get the same result using Nokogiri directly in Ruby and send it as plain text to Solr? Am I missing something? What would Tika do besides extracting the text from the HTML?

Thanks in advance,
Rodrigo.

Em 22-09-2016 12:11, Erick Erickson escreveu:
Tika was upgraded from 1.7 to 1.13 in Solr 6.2 so this is likely a
change in Tika.

You could _try_ downgrading Tika, but that's chancy and I have no guarantee
that it'll work.

Or use a SolrJ client to use an older version of Tika and transmit it
to Solr, here's
an example:

https://lucidworks.com/blog/2012/02/14/indexing-with-solrj/

Best,
Erick

On Thu, Sep 22, 2016 at 8:01 AM, Rodrigo Rosenfeld Rosas
<rr_ro...@yahoo.com.br.invalid> wrote:
I forgot to mention that this problem just happened after I upgraded to a
recent version of Solr and tried to reindex all documents. Some documents
that had previously succeeded now failed with this error.

Em 22-09-2016 11:58, Rodrigo Rosenfeld Rosas escreveu:
Hi, thanks. I was talking to @elyograg over freenode#solr and he (or she,
can't know by the nickname) recommended me to create a Java app integrating
SolrJ and Tika to perform the indexing. Is this the only way to achieve that
with Solr? Since I'm not usually a Java developer, I'd prefer another kind
of solution, but if there isn't, I'll have to look at the Java API and
examples for SolrJ and Tika to achieve that...

Just wanted to confirm. I'll try to get a sample HTML yielding to this
problem and attach it to Jira.

Thanks,
Rodrigo.

Em 22-09-2016 11:48, Allison, Timothy B. escreveu:
Y, looks like Nick (gagravarr) has answered on SO -- can't do it in Tika
currently.

-----Original Message-----
From: Allison, Timothy B. [mailto:talli...@mitre.org]
Sent: Thursday, September 22, 2016 10:42 AM
To: solr-user@lucene.apache.org
Cc: 'u...@tika.apache.org' <u...@tika.apache.org>
Subject: RE: Disabling Zip bomb detection in Tika

I don't think that's configurable at the moment.

Tika-colleagues, any recommendations?

If you're able to share the file on Tika's jira, we'd be happy to take a
look.  You shouldn't be getting the zip bomb unless there is a mismatch
between opening and closing tags (which could point to a bug in Tika).

-----Original Message-----
From: Rodrigo Rosenfeld Rosas [mailto:rr_ro...@yahoo.com.br.INVALID]
Sent: Thursday, September 22, 2016 10:06 AM
To: solr-user@lucene.apache.org
Subject: Disabling Zip bomb detection in Tika

Hi, this is my first message in this list.

Is it possible to disable Zip bomb detection in the Tika handler?

I've also described the problem here:


http://stackoverflow.com/questions/39628519/how-to-disable-or-increase-limit-zip-bomb-detection-in-tika-with-solr-config?noredirect=1#comment66575342_39628519

Basically, I get this error when trying to process some big valid HTML
documents:

RSolr::Error::Http - 500 Internal Server Error
Error:

{'responseHeader'=>{'status'=>500,'QTime'=>76},'error'=>{'metadata'=>['error-class','org.apache.solr.common.SolrException','root-error-class','org.apache.tika.sax.SecureContentHandler$SecureSAXException'],'msg'=>'org.apache.tika.exception.TikaException:
Zip bomb detected!','trace'=>'org.apache.solr.common.SolrException:
org.apache.tika.exception.TikaException: Zip bomb detected!
           at

org.apache.solr.handler.extraction.ExtractingDocumentLoader.load(ExtractingDocumentLoader.java:234)
           at

org.apache.solr.handler.ContentStreamHandlerBase.handleRequestBody(ContentStreamHandlerBase.java:68)
           at

org.apache.solr.handler.RequestHandlerBase.handleRequest(RequestHandlerBase.java:154)
           at org.apache.solr.core.SolrCore.execute(SolrCore.java:2089)
           at
org.apache.solr.servlet.HttpSolrCall.execute(HttpSolrCall.java:652)
           at
org.apache.solr.servlet.HttpSolrCall.call(HttpSolrCall.java:459)
           at

org.apache.solr.servlet.SolrDispatchFilter.doFilter(SolrDispatchFilter.java:257)
           at

org.apache.solr.servlet.SolrDispatchFilter.doFilter(SolrDispatchFilter.java:208)

I need to index those documents. Is it possible to disable Zip bomb
detection or to increase the limit using configuration files? I noticed it's
possible to add a tika.config file but I have no idea on how to specify what
I want in such Tika configuration files.

Any help is appreciated!

Thanks in advance,
Rodrigo.




Reply via email to