Improve CSV processing error message

[stsccfr]

I'm filing this on behalf of the WRY coordinator, who requests that we use more conventional notation when reporting errors to transcribers after processing an uploaded CSV file. Consider the following FlexCSV header line:

chapman_code,place_name,church_name,register_type,register_entry_number,death date

The last field contains a space instead of an underscore (it should be death_date), so it is invalid. The email from the server to the transcriber will say:

The field order definition at position 5 contains an invalid field: death date (is it blank?)}.

Because people working with FlexCSV will be working with spreadsheets, and because spreadsheets refer to columns using letters rather than numbers, we should use letters, too. Also we are using 0-based enumeration scheme in our emails, which is completely baffling to users. Finally, the wording of the error message could be less technical sounding, so I suggest the following wording:

The field name in Row 6 Column F is invalid: death date

The changes are:

1. Include the row number, using 1-based enumeration
2. Use letters to refer to columns so that it matches spreadsheet notation
3. Omit (is it blank?) since we already know that the field is not blank (it is death date in this case). The suggestion (is it blank?) should only be offered when the field name is truly missing.

[jayto581]

Code updated to improve CSV processing error messages

• Column are referenced with position number and letter
• Rows are referenced with data line number and physical file row number
• code also addressed issue 1574 (Verify who gets copies of email when transcriber uploads a file )

The CSV batch processing email changes may impact other projects such as FreeCEN. It is recommended that the other projects also review and test robustness of code.