From the user standpoint, I would prefer not for the database to be partially completed (as then you would have to delete the entries prior to the error to avoid double entries). However, I am not sure what is preferred: do we simply leave it as is, half-entered, or should all entries prior to the error in the CSV be rolled back and deleted?

kknaur wrote:

I am not sure I understand this question.

What I mean is , let's say I have 5 entries in the csv I selected. The 3rd one has an error that is handled by the program giving the user the a message and returning to the main menu. Should the first 2 entries in the CSV be committed to the database? Or should all entries leading up to the error in the same CSV file be removed from the database? So that way the user can fix the error in their data and recommit it all with ease.