We need a better way to retrieve many documents at same time [LUCENE-7029]

My use case involve that in some situations I need to get a lot of documents from the index. Most of my documents are small (less than 2KB).

I use API: IndexReader.document(id, visitor), but I noticed it keeps decompressing same block over and over again.

If we have an api that let's me get many documents at once, I could do the same job way faster.

After some experiment I realized I can get 2x+ gains on performance with this simple feature.

Disclaimer: I'm a newbie to lucene project. I truly appreciate hard work by all contributors for this awesome library. I would be more than happy to contribute back and get my hands dirty. Any help and pointers is more than welcomed.

---
Migrated from [LUCENE-7029](https://issues.apache.org/jira/browse/LUCENE-7029) by Jacek Migdal
Attachments: [LUCENE-7029.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-7029/LUCENE-7029.patch)

Start with the IndexReader.document(id, visitor) entry point and inspect the attached LUCENE-7029.patch. The work is complete when a bulk document-retrieval API avoids repeatedly decompressing the same block and achieves the reported performance improvement.

Written by the indexing model from the issue text.