[Spike] Investigate usage of ReGex \w character class in Puppet

The \w character class in Ruby does not support Unicode and will only match on [a-zA-z0-9_] which is limiting.

Instead, we should generally be using [[:word:]] where Unicode is important within a regex. This ticket involves auditing the existing code to identify critically important areas where this may be a problem in types / providers, gems, etc. Note that the Unicode compliant word character class will be slower than the ASCII version, so it would be good to keep an eye on performance as changes are proposed / made

This could also extend to supported modules, but should probably be a new ticket.

In scope for Agent + Platform Team

• Identify areas of concern, file tickets, assess impact and priority.
• Analysis of gems should limit scope (using --without development flag to bundler).