https://bz.apache.org/bugzilla/show_bug.cgi?id=63824
--- Comment #1 from Mark Thomas <ma...@apache.org> --- HTTP header values are case sensitive (HTTP header names are insensitive). Both RFC 2616 and RFC 7230 define the relevant token as "close" so the current case sensitive comparison is correct. The current code also assumes "close" is the entire header value. My reading of the spec does not support that although I can't think of anything else that would be present. The worst that could happen is that the "close" token would be added twice which, apart from the extra bandwidth, should not be an issue. I want to look into the performance implications of parsing this "properly" but I am leaning towards WONTFIX at this point. -- You are receiving this mail because: You are the assignee for the bug. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
