parser for CSV can't handle common cases found in the wild

[ScottMaxson]

Currently using the csv-parse npm library, which has a few options that make handling real-world CSV files more flexible/adaptable. Cases so far are

• text fields with double-quoted words/phrases
• missing trailing fields in some rows

Could be fixed by passing additional options to the underlying CSV parser such as relax: true etc. at the cost of having to refer to that library, and adding features that may not port to a different parser.