searching mechanism and vector in index

[Amna Waqar <am...@gmail.com>]

Hi all,
I ve read the code of indexer in nutch-1.2 which states that store ,index
and vector are used for each field in the index.
what is the reason for using 'vector'.
Also
i ve understood the indexing steps like formation of documents containg
fields objects like
url, content ,title as field name and their values
Then feeding of document objects to analyzer which do tokernizing and form
the index
Now i want to know in which form(like data structure) this index is stored
and how does a query is searched in the index.
Does it is being searched sequentially until the match is found.
THanks
Amna Waqar

Re: searching mechanism and vector in index

[Markus Jelsma <ma...@openindex.io>]

Hi,

I haven't read the code on this one but it might be code that couples with 
Lucene, which is a dependency of Nutch < 1.3. In Lucene, there is more going 
on that just creating stored and analyzed (indexed) field values. There are 
norms, term vectors and more. The Lucene docs will tell you  about the file 
format, query language etc.

http://lucene.apache.org/java/2_9_1/

Cheers,

> Hi all,
> I ve read the code of indexer in nutch-1.2 which states that store ,index
> and vector are used for each field in the index.
> what is the reason for using 'vector'.
> Also
> i ve understood the indexing steps like formation of documents containg
> fields objects like
> url, content ,title as field name and their values
> Then feeding of document objects to analyzer which do tokernizing and form
> the index
> Now i want to know in which form(like data structure) this index is stored
> and how does a query is searched in the index.
> Does it is being searched sequentially until the match is found.
> THanks
> Amna Waqar