Matching Hubot's name in Coffeescript

Question

I want to create a Hubot script which listens for strings that look like:

thanks Hubot

where Hubot is the bots name. If we assume that bot is always named hubot, then it's trivial:

robot.hear /thanks hubot/i, (msg) -> 
    ...

But the name of course can vary, though is contained in the robot.name variable.

Question: how do you interpolate the bot's name into the regular expression? This question seems to indicate you can just use block regular expressions, but I tried:

robot.hear ///thanks #{robot.name}///i, (msg) ->
    ....

And it didn't seem to work.

Edit: I have a workaround, I can extract the second word and compare to the bot's name, like so:

  robot.hear /thanks (.*)/i, (msg) ->
    name = msg.match[1]
    if name.toLowerCase() is robot.name.toLowerCase()
      ....

But this feels rather kludgy, I'd like to be able to just interpolate a variable into the regex itself.

Answer 1

The ///s in Coffeescript don't just allow interpolation, they engage the full extended mode which means that whitespace is no longer significant inside those regexes, ie. if you want to match a space then you have to use something like \s

So try this:

robot.hear /// thanks \s+ #{robot.name} ///i