Token Seperation Issue

I have discovered an issue when tokenize and matchAllTokens is enabled. When you search for white (with a trailing whitespace), the tokens are ['white', '']. This causes fuse to return no results due to one empty token. Would removing empty tokens as below be a fix for this?

var tokens = pattern.split(options.tokenSeparator).filter(function(token){
    return token !== '';
})

[dlosie]

I solved this by using String.prototype.trim() on my search query to remove unwanted leading /trailing whitespace.

fuse.search(query.trim());

[jwmann]

@jeffal I pulled in and re-built with the most recent master changes. Thank god Travis-CI didn’t fall off a cliff this time and built successfully.

Just needs a merge.