Question > Shouldn't limit(x) be used instead of looping over cursor?

Hi guys,

I reported a question some time ago, that end up with this enhancement. That was pretty good, but I still think it can be fine tuned even further.

Let's assume the following scenario:

Configuration: cursor-batch-size: 100, default-pagesize: 25 and max-pagesize: 100
I want to get the first page

So, according to the code we have here, the cursor will pull the first 100 documents, and code will return the first 25.

I think this can be improved if instead of counting the number items (_pagesize) in a loop, we would use .skip(toSkip).limit(pagesize), in other words:

            cursor = getFindIterable(coll, sortBy, filters, hint, keys);
            cursor.skip(toskip)
                  .limit(pagesize);

            MongoCursor<BsonDocument> mc = cursor.iterator();

            while (mc.hasNext()) {
                ret.add(mc.next());
            }

IMO, the advantage of this approach is that the cursor will just pull 25 items, instead of the full batch (100). Note that when documents have certain size, or for big collections, the improvement is noticeable.

Thoughts?

SoftInstigate / restheart

Question > Shouldn't limit(x) be used instead of looping over cursor? #343