Re: URL escaping in HTML output

[Joerg Pietschmann]> using Saxon 6.4, i have an URL of
>  /foo?rel=bar/baz
> stuffed literally into the generated HTML code. Yes, it is a href
> attribute of an A element. While reading the URI RFCs i got the
> impression that the slash in the query string should have been
> escaped as %2F.

Looking at rfc 1945, "Hypertext Transfer Protocol -- HTTP/1.0",  there's no
problem with using "/" in the query string.  The relevant part of the BNF is

query          = *( uchar | reserved )
...
reserved       = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+"

So a slash is clearly allowed.  I don't have the HTTP/1.1 rfc handy, but I
doubt that it's changed.  For use with HTML hrefs, the HTTP rfc should be
authoritative, I would think.

>Fortunately, none of the browsers or http servers
> i use has complained. Can somebody tell me whether i'm wrong or
> whether it is a bug in the processor?

Seems it's not a bug at all.

Cheers,

Tom P



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list