URLs Containing "#"

The CERN httpd running as a proxy fails on URLs like:
gopher://146.138.63.106:70/00%5Cdir%5Cnfs%5Coeweekly%5Coeweek95%5Coe%2395-01.txt
When you translate the escaped characters, this is:
gopher://146.138.63.106:70/00\dir\nfs\oeweekly\oeweek95\oe#95-01.txt

Using Netscape directly (not through a proxy) there is no problem getting
this URL.

The function "scan" in HTParse.c reads the "#" as a delimiter between the
path and the anchor, even though in this case, it's just another character
in the path.

The #ifdef'ed out OLD_CODE in scan suggests JFG and TBL have struggled with
this problem before.  Does anyone have a fix for the proxy so it can get
gopher files with "#" in the pathname?

Or, would it be better to tell the gophermaster at 146.138.63.106 that "#"
is a reserved character (in URLs) and shouldn't be used in filenames?

		Randy Simons
		Sandia National Laboratories

Received on Wednesday, 7 June 1995 12:40:57 UTC