It is a bit late, and the question has certainly been answered. However if you are still open to suggestions, and have not yet spent your time converting the indicated parser from C# to JavaScript, I have written a parser (originally in PHP) which I converted myself[2] to JavaScript. It is available at my website under the 3-clause BSD license. The parser seems to be reasonably fast, but I haven't performed any analysis on its speed.
It may not be as flexible in some ways as other possible projects out there, but it does allow defining your own codes ("bb-code" or not, with quite a few properties), and is also all contained within the one file. This is not a simple find-and-replace parser, and is not based on regex. Due to there currently being no option to disable escaping of "content", the output of one parser cannot be used as input to another (to perform more than one pass on different types of codes, such as :smile:-type codes and regular bb-code) as the output of the first parser would be escaped by the second.
If this is of any interest to you, it might save you from having to convert that other library. Technically, I'm a relative "unknown", but that's the great things about JS/OSS: you can check out the source to see what I've done.
[2] As a result, there are a few remaining "compatibility functions", but I rewrote things which had native equivalents available.