------- Comment #3 from paolo dot carlini at oracle dot com 2010-01-02 03:06 ------- I'm working on this, and I want to make sure we agree about the third case. As I understand the Standard, the expected is 1, not -1, that is, I don't see anything wrong with parsing Pndedelnik as Pnd followed by an 'e'. Of course the parsing (at variance with the current behavior) must stop on that 'e', with error status good instead of eof. In fact, in this case the abbreviation is *not* followed by characters that could match a full name (as is the case of Monday in English, for example), and the facet simply doesn't continue reading beyond that 'e'. In other terms, I don't see why Pndedelnik should be different from 'Pnd ', that is Pnd followed by a space (which I'm sure we agree it's fine).
-- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38081