Re: Is it wrong to create index writer on each query request.

Hi, What are the other disadvantages (other than the time factor) of creating index for every request? Manjula. On Thu, Jun 5, 2014 at 2:34 PM, Aditya wrote: > Hi Rajendra > > You should NOT create index writer for every request. > > >>Whether it is time consuming to update index writer when

Re: ASCIIFoldingFilterFactory

Hi Michael, Questions about Solr should go to the Solr user mailing list, rather than this list, which is for Lucene users - see for how to subscribe. I’ve never heard of ASCIIFoldingExpansionFilterFactory, but ASCIIFoldingFilterFactory has a new

ASCIIFoldingFilterFactory

Hi there I am a relative newbie Solr user so please be gentle with me. I am experimenting with various phonetic filters and the tokens created can vary depending on whether the words contain diacritical characters. My problem is that the documents being indexed are not always consistent in terms

SpanQuery not working as expected

I read through the http://searchhub.org/2009/07/18/the-spanquery/ which provided a good overview for how one can construct fairly complex span queries.  I was particularly interested in the ability to construct nested span queries.  I'm trying to apply this concept to search a field that contain

Re: Indexing integer ranges for point search

It all depends on the statistics: how the ranges are correlated. If the integer range is small: from 1-2, for example, you might consider indexing every integer in each range as a separate value, especially if most documents will only have a small number of small ranges. If there are too

Re: How to approach indexing source code?

If you already have a parser for the language, you could use it to create a TokenStream that you can feed to Lucene. That way you won't be trying to reinvent a parser using tools designed for natural language. -Mike On 6/5/2014 6:42 AM, Johan Tibell wrote: I will definitely try a prototype.

Re: How to approach indexing source code?

I will definitely try a prototype. My main question is whether I'm better off creating documents directly or if I should try to parse the compiler output using an analyzer/tokenizer. On Thu, Jun 5, 2014 at 12:24 PM, Aditya wrote: > It is up to your requirement. You could either index source fil

Re: How to approach indexing source code?

It is up to your requirement. You could either index source file or compiler output. Try doing some proof of concept. You will get some idea of how to move forward. Regards Aditya www.findbestopensource.com On Thu, Jun 5, 2014 at 2:48 PM, Johan Tibell wrote: > By "index the entire source fil

Re: How to approach indexing source code?

By "index the entire source file" do you mean "don't index the compiler output"? If so, that doesn't sound very appealing as it loses most of the benefit of having a search engine built for searching source code. On Thu, Jun 5, 2014 at 11:11 AM, Aditya wrote: > Just keep it simple. Index the en

Re: Reader reopen

You can reopen non-NRT readers, too. It just means you have to call IW.commit in order for the non-NRT reader to see the changes. DR.openIfChanged accepts either kind of reader. Mike McCandless http://blog.mikemccandless.com On Thu, Jun 5, 2014 at 4:24 AM, Gergő Törcsvári wrote: > Hy, > > Th

Re: How to approach indexing source code?

Just keep it simple. Index the entire source file. One source file is one document. While indexing preserve dot (.), Hypen(-) and other special characters. You could use whitespace analyzer. I hope it helps Regards Aditya www.findbestopensource.com On Wed, Jun 4, 2014 at 3:29 PM, Johan Tibell

Re: Is it wrong to create index writer on each query request.

Hi Rajendra You should NOT create index writer for every request. >>Whether it is time consuming to update index writer when new document will come. No. Regards Aditya www.findbestopensource.com On Thu, Jun 5, 2014 at 12:24 PM, Rajendra Rao wrote: > I have system in which documents and Quer

Re: Indexing integer ranges for point search

Hi, Continuing your example, you could do the following: Document: range1_from:1 range1_to:3 range2_from:12 range2_to:20 range3_from:13290 range3_to:16509 ... other fields... Query (for "2"): (+range1_from:[* TO 2] +range1_to:[2 TO *]) OR (+range2_from:[* TO 2] +range2_to:[2 TO *]) OR (+ran

Re: Reader reopen

Hy, Thanks the fast answear. So the only way to reopen a reader if I use a near realtime reader or if I know where is the indexes and what kind of reader I had before? (And manually close and open an another reader.) Gergő 2014-06-03 12:37 GMT+02:00 Michael McCandless : > Sure, just use Direc