Wait for data to be drained before writing new central file header

[falk0pr0ss]

When adding a huge amount of files to an archive there will likely be a memory out of heap exception when writing the central file header in the zip module.

Here is a small test for reproduction where testfile can be an empty file:
<script src="https://gist.github.com/falk0pr0ss/d92f77c342f2505b40faed42946f73b4.js"></script>

Here is the behaviour before the proposed fix - the huge increase of memory usage is when the writeCentralFileHeader method is called repeatedly and a memory out of heap exception is caused:

Here is the behaviour with the proposed fix, where the risk of a memory out of heap exception is reduced by waiting for the drain event to be fired after the write method returns false:

[stixx200]

@ctalkington
Any updates here? I also need this fix :)