Re: [cpp-netlib-devel] URI unescaping

[Jeroen H. <vex...@gm...>]

Hi,

On Fri, Feb 5, 2010 at 17:17, Nelson, Erik - 2
<eri...@ba...> wrote:
> In the http server example, the request.uri is still escaped, e.g., has
> characters like '%2f' in it... Is there a 'best' way to unescape it
> within cpp-netlib?  Or should I use another source of un-escaping?
>
> Thanks
>
> Erik
>

The short answer, no cpp-netlib currently doesn't have any features to
unescape a URI except for an implementation of unescape itself.

Now for the longer answer, as unescaping a URI isn't as easy as it
might look at first glance. It is part of the normalisation process
normally applied to URI's, and is described in section 6.2.2.2. in the
RFC  which can be found at http://www.ietf.org/rfc/rfc3986.txt. Each
section of the URI has a set of reserved characters, and those should
be left escaped whilst any other character can be unescaped safely.
Unescaping the reserved characters would change the meaning of the
URI, and might even make it invalid. An example would be
http://user%40n...@do... versus
http://user@nam...@do..., where the escaped character is in the
user_info section, and unescaping it invalidates the URI.

The whole normalisation process is something I hope to tackle after
I've completed the parser, but I'm unsure when this might be. For the
time being you could have a look at google-url, a URI parser written
for the chrome browser which can be found at
http://code.google.com/p/google-url/.

Jeroen