You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@tika.apache.org by Kevin Miller <ke...@oktax.state.ok.us> on 2009/08/19 17:15:16 UTC

Using Tika in Solr to index a Word Document

I am using the Solr nightly build 8/11/09 which uses the tika-core-0.4.jar.  I have set the text field in the solrconfig.xml file to be stored.  I index an MS Word document and when I search for a word in the text of the document and it pulls up the xml format.  The text field is showing the text of the document but there are areas in the document that are FORMDROPDOWNs.  What I want to know is if there is some way that the information that was entered into the FORMDROPDOWNs can be retrieved.  The text field contains the following information (I have entered in parenthesis the actual data from the MS Word document for the FORMDROPDOWNs:

<arr name="text">
−
<str>
       	 OKLAHOMA TAX COMMISSION  
  
  	FISCAL IMPACT STATEMENT AND/OR ADMINISTRATIVE IMPACT STATEMENT
  	FIRST REGULAR SESSION, FIFTY-SECOND OKLAHOMA LEGISLATURE
  
  
  DATE OF IMPACT STATEMENT:	 May 21, 2009
   
  BILL NUMBER:  HB 1097	
  
  STATUS AND DATE OF BILL:	  FORMDROPDOWN(Enrolled Bill)    05/20/2009 
  
  AUTHORS:	House	  FORMTEXT   Dank 			Senate	Brogdon
  
  TAX TYPE (S):    All   SUBJECT:      FORMDROPDOWN(Credit)   
  
  PROPOSAL:	  FORMDROPDOWN(New Law)     
  
  This measure creates a nine (9) member task force to study tax credits.  The measure also includes provisions of procedures and duties for the task force and directs the task force to produce a final written report for the Speaker, the Governor and the Pro Tempore.
  
  
  EFFECTIVE DATE:	August 21, 2009 (Assuming sine die is May 22, 2009)
  
  REVENUE IMPACT: 
  
  Insert dollar amount (plus or minus) of the expected change in state revenues due to this proposed legislation.
  
  FY 09:	None		
  FY 10:	None  FORMTEXT    
  
  ADMINISTRATIVE IMPACT:
  
  Insert the estimated cost or savings to the Tax Commission due to this proposed legislation.
  
  FY 10:	None
  
  
    	  	lrh 
  DATE				DIVISION DIRECTOR
  
                                                                                         
  DATE				XXXXX XXXXXX, ECONOMIST
  
                                     
  DATE				FOR THE COMMISSION
   
</str>
</arr>

Kevin Miller
Web Services

Re: Using Tika in Solr to index a Word Document

Posted by Jukka Zitting <ju...@gmail.com>.
Hi,

On Wed, Aug 19, 2009 at 5:15 PM, Kevin
Miller<ke...@oktax.state.ok.us> wrote:
> What I want to know is if there is some way that the information that was entered into
> the FORMDROPDOWNs can be retrieved.

Hard to say without seeing the actual Word document. Tika uses the
Word parser from the Apache POI project, so if POI can access that
information it should also be possible to expose it in Tika.

If you want to dig deeper into this, I would suggest that you give the
POI project and the OfficeParser code in Tika a look to see whether
there is some change we could make to support your case.

Alternatively, just create a simple test document that illustrates
this issue and attach it to a new a Tika improvement request [1].
There are no guarantees on when issues like that get fixed or even
looked at, but at least it won't get forgotten if it's in the issue
tracker.

[1] https://issues.apache.org/jira/browse/TIKA

BR,

Jukka Zitting