The \A and \z anchors are allowed in regexen

Henrik explained that the wrong note in the spec came from a weird and probably
unwanted bit of code in the lexer that would sometimes strip those anchors
out... but only if they were occurring in positions where they'd never match any
string. He thinks either we were trying to "help," or we might have been working
around a bug in some ancient Ruby.

future_regex.pp

@@ -0,0 +1,14 @@
+notice("my string" =~ /^my string$/) # true
+notice("my string" =~ /\Amy string\z/) # true
+notice("my string" =~ /\Amy stringgg\z/) # false
+# .... so the bit about \A and \z being disallowed at https://github.com/puppetlabs/puppet-specifications/blob/master/language/lexical_structure.md#regular-expressions seems to not be true? What would definitively prove this?
+notice("my string\nhey second line" =~ /\Amy string\Z/) # false
+notice("my string\nhey second line" =~ /^my string$/) # true, So, ok....
+notice("my string\n" =~ /\Amy string\z/) # false
+notice("my string\n" =~ /\Amy string\Z/) # true
+notice("hey\nmy string\n" =~ /\Amy string\Z/) # false
+
+# this is looking like that's not the case at all.
+
+
+