Glint checks grade files for errors in format and for some errors in content.
Usage
glint [-?] [selection options] [files]
Details
Glint reads its arguments, emitting the same error messages as do the other grading programs when they encounter errors, except that all problems are reported as "errors" rather than "warnings". However, its error messages go to the standard output, not the error output, because the messages are the entire purpose of glint. The return status is 0 (indicating success) regardless of what errors are found, unless there are IO errors.
The choices of output stream and return status may arguably be poor, but they have the weight of history on their side. To check whether glint is happy with a file, or not, you need to see whether its output is empty or not.
Here is an outline of the checks that glint performs beyond what the other grading programs do:
- The selected student records are checked for oddities such as excessive blanks in the name, or a student number that is the same as the number for another selected student record. These oddities are reported in student-number order.
- Glint recomputes the calculated grades — as opposed to the raw grades, in the sense of the description of the file format. The new results are not entered in the file, but they are compared with the existing values and any discrepancies are reported. If you haven't yet calculated some of your grades, glint simply skips the check for those grades.
Glint readily produces multiple error messages for a single error, because after reporting an error it tries to keep going, so as to discover as many other problems as it can. If it finds errors while reading a file from input, glint does not try to recompute student grades, but that still leaves some scope for redundant error messages.
Glint's checks on names may confuse you, because the name displayed in a student record is "normalized", with blank sequences compressed and capitalization standardized. Look for the raw name shown with the error message.
If glint reports a digit at the end of a name, that very likely indicates a missing separator (tab character) before the first mark. A text editor can help to verify this and to fix the problem.
It is probably a poor idea to use the selection options unless you are very sure that the omitted student records are of no further use, and even then the only good reason for omitting them is to avoid complaints about repeated student numbers.
If no files are specified, input comes from the standard input and output goes to the standard output.