Implemented

• The target selecting filter was implemented.
  Now we can set what files are synchronised by regular expression.
• We can configure the size of chunks.
  We can use larger chunks to improve performance.
  (This feature can not be used with IBM Cloudant)
• Read chunks online.
  Now we can synchronise only metadata and retrieve chunks on demand. It reduces local database size and time for replication.

Fixed

• Many typos.

Note

I haven't been well for the last few weeks. Finally, I went doc and realised that I had lost my concentration and prudence by my own thyroid. I now getting be medicated and better, but sorry for the terrible code and messy replies.
I will continue to fix what I have not noticed. I apologise in advance for the big commit.