I'm using nutch .8 and using multiple indexes with it. I just point the crawler to different folders, depending on which index I am updating. When searching, I point it to the index directory, not indexes. I merge my indexes (the one in index and indexes) and I store the result in the index folder. I think the indexes folder stores an index for each segment (maybe?) but if you merge them together with the main index I think the index folder will have everything. At least it seems to be working in my implementation.
On 10/31/07, Pratyush Banerjee <[EMAIL PROTECTED]> wrote: > > Hi Sir, > > I am currently working on a cross lingual search engine that requires > building separate index for every language we are going to support. We are > using Nutch as the basis of the engine. However i was trying to find any > help regrading handling multiple indexes in nutch. > > The nutch indexer creates an index and an indexes directory. I am not sure > about which one is the actual index directory. I tried tweaking around > with > the code but with little effect. Can any one tell me, why there are two > separate directories and which one is used for what ? > > secondly since my index will be on the basis of language, i would try to > identify the language using the language identifier (which is working > currently). So which files need to be modified and also if someone can > provide me with some idea about how to go about it. > > Am current;ly using nutch 0.9 on a fc6 machine with jdk1.6. > > Anybody please help... > > -- > Pratyush Banerjee > SPO, CLIA > IIT Kharagpur >
