Add option to split() to remove matching characters from emitted chunks

[mhart]

This would bring it closer to String.split behaviour (at the moment you're trimming each line in test/split.asynct.js to do the comparison - this wouldn't account for leading spaces for example)

It would also allow for splitting on Regex patterns, because the matched characters wouldn't need to be re-emitted.

Happy to do a pull request for this, but figured I'd wait to see what you thought of #7 first.

[dominictarr]

with your patch this would be really easy to add.
basically, just have to not append the matched character.

there is already a way to do this with string.split:

'a:b:c'.split(/(:)/)

// ['a', ':', 'b', ':', 'c']

this is technically a breaking change, but since 99% of the usecases for this is to make newline seperated json, or otherwise to break on whitespace and then feed it into a parser, I don't think it will affect actual usage.

since it was so simple after your patch, I've implemented this already,

3ac501c
a3f4e66

cheers!

[mhart]

Great! I just added a pull request to keep the behaviour compatible with the old API... up to you:

#9