avoid downloading again already existing chunks

[drzraf]

• Since we know where chunks are expected to be written before starting the download
  and
• Since chunks are written after they have be fully downloaded (and stored into memory)

Then it should be possible to avoid downloading chunks already present on-disk

The only possible issue with this are chunk partially written on-disk (eg: insufficient disk-space at the time of writing)
To avoid corrupted chunk being stored on-disk, a callback() could remove the partially written file (== if fs.writeFile() returns an error)

[nurrony]

Thanks for opening this issue. I will start working on it as soon as I can allocate some bandwidth for this.