Comment 7 for bug 579669

That's correct. The patch uses percent encoding (http://en.wikipedia.org/wiki/Percent-encoding), except '%' is also a reserved character here, so we escape '-' and then use '-' and the escape character by replacing '%' with '-'.

The patch can be tidied slightly by using urllib2.quote rather than urllib.quote, thus removing the need to import urllib (I didn't realise urllib2 exposed urlib's quote at the time I wrote the patch)