The parsing of the header of a gradebook NG spreadsheet is broken. This is because the regular expressions used to validate and parse a the column titles are wrong.
The columns can be of one of four types:
#SOMETHING - a comment, meaning to ignore the column
* SOMETHING2 - a "comment column". SOMETHING2 must match the name of an assignment
SOMETHING3 [dd.dd] - Here SOMETHING3 should be the name of an assignment and dd.dd is a point value to assign
SOMETHING4 - Here SOMETHING4 should be the name of an assignment
The problem is that the regexps that are present make the assumption that the SOMETHING strings in these consist solely of a-zA-Z0-9_ and space, due to their use of a \w special character. This is wrong for several reasons. First, the dash - is allowed in assignment names. Second, this fails if someone uses non-ascii characters in an assignment name, this will also fail to match.
I should note that Java regexps are a mess with respect to non-ascii strings (see e.g. here: https://stackoverflow.com/questions/4304928/unicode-equivalents-for-w-and-b-in-java-regular-expressions ) so you should be quite careful writing them.
A second, unrelated bug is that the error messages given don't give any hint as to what the problem might be. I have at least added the exception messages to the screen, which, while possibly not terribly user friendly are better than nothing.