httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "James Park (pencil_ethics)" <pencilethics.l...@gmail.com>
Subject Re: Best way for a module to retrieve the full original URL?
Date Mon, 03 Dec 2007 13:49:22 GMT
Jeff Tharp wrote:
> The problem with this approach is that I seem to loose URL fragments,
> such that when I request http://server.domain.com/index.html#anchor,
> orig_uri = "http://server.domain.com/index.html", not
> "http://server.domain.com/index.html#anchor".  Is this expected, or am I
> going about this in the wrong fashion?

Jeff,

A superficial analysis of RFC 2396 Section 4.1 and RFC 3986 Section 3.5
suggests the fragment identifiers are exclusively handled by the
user-agent. This makes sense from the common sense perspective:
fragments are internal to the document and the server does not parse the
document, it merely serves it to the user-agent. So the server does not
need to understand fragments.

Simplified: you will never see fragment identifiers on server-side. The
Referer header does not include them either (by virtue of RFC 2616
Section 14.36); so if you really _do_ need them (you shouldn't, to be
honest), you are out of luck. The lack of fragment identifiers is not
only expected behavior; it is mandatory.

I hope this settles your question.

- James Park (pencil_ethics)

Mime
View raw message