Unescape character entities in raw data

[mdoering]

Raw source data can have character data, especially unicode, escaped in various ways.
During import this needs to be resolved and flagged. The verbatim data should be preserved, but the interpreted data stored in the db anywhere but the verbatim table should have properly unescaped characters:

• xml & html entites
  • named &
  • hex &
  • decimal &
• unicode entities
  • U+0026
• java unicode entites
  • hex \x26
  • octal \046
• CSS & ECMA Javascript
  • Unicode escapes started by "\u": \u00A9
  • Unicode code point escapes indicated by "\u{}": \u{2F804}
  • Hexadecimal escapes started by "\x": \xA9

Apache commons has libraries for this: https://commons.apache.org/proper/commons-text/javadocs/api-release/org/apache/commons/text/StringEscapeUtils.html

[mdoering]

we should probably also strip xml/html tags such as in this title:

A new species of <i>Neamia</i> (Perciformes: Apogonidae) from the West Pacific Ocean.

[mdoering]

Using a UnescapedVerbatimRedord wrapper class that does most of the job and remembers if any values have been modified so we can flag an issue: eee5d67

[mdoering]

Addresses some concerns expressed in CatalogueOfLife/general#37