I am guessing you have used 0.8-dev, I am not sure how things are in 0.7

bin/nutch index <index> <crawldb> <linkdb> <segment>

where <index> = name of your index directory! NOTE you cannot call it
"index" (Please correct me if I am wrong). This is why it is good to
call it "indexes" or "id"
<crawldb> = name of your crawldb directory
<linkdb> = name of your linkdb directory
<segment> = name of your segment directory! not "segments" directory
meaning  "segments/200612120937"  or something similler.

Now you need to index all your segments using the command above. And
the <index> folder should be the same but the segment folder will
change offcourse.
Example

>bin/nutch index indexes crawldb linkdb segment/200612029232
>bin/nutch index indexes crawldb linkdb segment/200612022212

Once the indexing is done for all the fetched segments then try the following

>bin/nutch merge index indexes

Note "index" above is where the merged index is placed for searching
(i.e. output index) and "indexes" folder is where your segments index
are.

Hope this helps. I don't know how things are in windows with cygwin.


On 3/27/06, Vertical Search <[EMAIL PROTECTED]> wrote:
> So far, I have been able to work through minor obstacles in setting up Nutch
> for a vertical search.
> Now am kind of stuck from past 24 hours to merge the indexes.
> I have crawled multiple sites. but want to merge the indexes.
> I am using merge command as follows
>
>
> bin/nutch merge C:/vSearch/DB/index C:/global_search/index
>
> All it says is "Adding C:/vSearch/DB/index" but the indexed data itself is
> not consolidated and merged.
>
> Can some one point me to the correct mail in archieve or help me get over
> this problem...
> Another question is in IndexMerger, I see -workingdir as a flag. tried with
> that too with no avail...
>
> Please.......... help..
>
> Thanks
>

Reply via email to