Improve user experience for large files

[bruth]

• Log records evaluated so far to stdout and the rate of evaluation, and time remaining
  • If this is a file on disk, the size can be inspected and the evaluation rate can be used to estimate how long it will take to parse the file.
• On exit, print the line number of the record that was last parsed
• Add command line option for starting at a specific line in case parsing needs to be resumed

[gracecodes144]

These sound like good additions. Can't think of anything to add at the moment.