ids can also change as the result of an add(), not just optimize(). An add can trigger a segment merge which can squeeze out deleted docs and thus change the ids. I think everything else you said is pretty much correct.
On 10/6/05, Jack McBane <[EMAIL PROTECTED]> wrote: > > I know that in general if you optimize an index the document id that a > HitCollector would see is not guarenteed to stay the same since any > deleted > documents would get removed from the index completely and all later > documents pushed up to fill in the gaps. I'm wondering though, are the > document ids are preserved accross an optimize if there were no deletions > (which could be checked with a call to IndexReader.hasDeletions())? > > Similarly, if you populate an index using IndexWriter.addIndexes > (Directory[]) > are the document ids preserved except with offsets added? That is, if you > have two indices with 10 documents each (so that each has document ids > 0-9) > and combine them in this way, does document 0 in the new index correspond > to > document 0 in the Directory[0] index, ..., document[9] correspond to > document 9 in Directory[0], document[10] correspond to document 0 in > Directory[1],...,document[17] correspond to document 9 in Directory[1]. > > I've been creating a prototype of something that for a first pass assumed > that documents are preserved as described above, but some of my results > are > starting to look like that is not the case. > > Thanks. > > -- -Yonik Now hiring -- http://tinyurl.com/7m67g
