Hi,

Is it possible to return just the raw text of the document extracted by Tika? In other words, we don't want it in XML or JSON, just the text in it.

So if the document contained "Hello World", we want to get back "Hello World" without any wrapping XML. Is this possible? Thanks.

Sincerely,
Alex

On Thu, 16 Aug 2012, Alexander Cougarman wrote:
> Is it possible to return just the raw text of the document extracted by
> Tika? In other words, we don't want it in XML or JSON, just the text in
> it.

Yes. Are you using the TikaApp jar, calling the Tika facade class, or
calling a parser directly?

Nick

I'm using this C# code to call the parser directly via its URL; it returns JSON:

var url = @"http://localhost:8983/solr/update/extract";

var client = new WebClient();
client.QueryString.Add("extractOnly","true");
client.QueryString.Add("wt","json");
var data = client.UploadFile(url, "input.txt");
var json = ASCIIEncoding.ASCII.GetString(data);
Sincerely,
Alex
-----Original Message-----
From: Nick Burch [mailto:[EMAIL PROTECTED]]
Sent: 16 August 2012 6:36 PM
To: [EMAIL PROTECTED]
Subject: Re: Return raw text from document

On Thu, 16 Aug 2012, Alexander Cougarman wrote:
> Is it possible to return just the raw text of the document extracted
> by Tika? In other words, we don't want it in XML or JSON, just the
> text in it.

Yes. Are you using the TikaApp jar, calling the Tika facade class, or calling a parser directly?

Nick

On Fri, 17 Aug 2012, Alexander Cougarman wrote:
> I'm using this C# code to call the parser directly via its URL; it
> returns JSON:
>
> var url = @"http://localhost:8983/solr/update/extract";

You might have more luck asking on the SOLR lists, as it looks like your
question is with how SOLR integrates Tika. Tika will very happily give you
just the plain text, your only query is how SOLR exposes that...

Nick

Hi Alex,

On 17 Aug 2012, at 08:37, Alexander Cougarman <[EMAIL PROTECTED]> wrote:

> I'm using this C# code to call the parser directly via its URL; it returns JSON:
>
> var url = @"http://localhost:8983/solr/update/extract";
>
> var client = new WebClient();
> client.QueryString.Add("extractOnly","true");
> client.QueryString.Add("wt","json");
> var data = client.UploadFile(url, "input.txt");
> var json = ASCIIEncoding.ASCII.GetString(data);
>
> Sincerely,
> Alex

There is parameter called extractFormat that you can use in extractOnly mode.  This will give you the serialised content back as plain text within a <str> element within the full XML response.

Not sure your full use-case, and if you are using the SOLR instance for other features later,  but you could also use the JSR-311 Tika Server to do this extraction for you http://wiki.apache.org/tika/TikaJAXRS

Cheers,
Dave