Broken default selector for pronunciation

[stephenmac7]

I wasn't getting any results from running wikipron (e.g. wikipron fra) and realized that the class names on wikipedia may have changed, breaking the XPath selectors used in config.py and extract/default.py. In particular, the span containing the pronunciation is now marked IPA nowrap instead of IPA.

Things seem to start working again if I replace span[@class = "IPA"] with span[contains(@class, "IPA")].

[kylebgorman]

I believe you; we have to change that selector frequently. I can look at it in a week or so; would you like to take a stab at it yourself? I will probably be able to review a PR quicker than I can implement one myself.