Update default extractor for purgeCSS to account for escaped class names

[ronaldcurtis]

Describe the problem:

Not sure if this is a bug or a feature request, but I just noticed that the default extractor for purgeCSS does not extract class names that have been escaped. For example, in translation JSON files, we have a need to include html elements as such:

{
  "helloThere": "Hello there, <span class=\"whitespace-no-wrap\">Mr. Jones</span>"
}

But it seems like the patterns used in the default extractor don't collect these class names properly. Using a Regex tester, it looks like the final \ is also included:

I guess it's not too much of a problem, as one can just add another extractor for JSON files. But it took as a while to figure out the work around, so it might help others in the future if the patterns could also account for escaped strings.

[adamwathan]

Fixed in 1.8.7 👍