Import/ validation script

[nbumbarger]

Develop node script to transform CSV into JSON data, then check JSON data for errors.
The validation will primarily check for missing columns and types errors against a template schema. It should be able to report a minimum of missing columns and type errors (line number of first error, maybe a count of total corrupted records). Should it break on the first error, like a typical validator, which would potentially turn correction into an iterative process but may be simpler from a UI perspective, or should it deliver a comprehensive report of all errors at one time?