Re: Patch for mailto: link urlencoding

* Michael A. Puls II <[email protected]> [01-06-2010 23:03]:
>
> On Wed, 06 Jan 2010 13:30:54 -0500, Jon Nathan  
> <[email protected]> wrote:
> 
> > We ran into an issue with squirrelmail 1.4.19.  When you click on
> > a mailto: link that has special characters (plus signs, etc), the
> > resulting compose.php pane replaces them with spaces.
> 
> This is very common with webmails. They often use http-based encoding,  
> decoding and parsing functions when dealing with mailto URIs.  
> mail.live.com, mail.yahoo.com and mail.google.com are all examples that  
> have issues with this in one part of their code or another.
> 
> Unfortunately, '+' is allowed to be unencoded in an hfvalue in a mailto  
> URI. You have to take that into account when you place that mailto data in  
> an http URI. (Of course, an author of a mailto URI should just use %2B  
> anyway when they want a + for good measure)
> 
> I address this as follows:
> 
> 1. The function that percent-decodes a mailto hfvalue must not decode '+'  
> to ' '.
> 
> 2. The function that percent-encodes the resulting data gotten from #1  
> must percent-encode all '+' to '%2B' if the data will end up in an HTTP  
> URI. That way, stuff like PHP's #_GET[] etc. will produce a  
> percent-decoded value that matches the original mailto data properly.
> 
> 3. You can do without #1 though if you take the source mailto URI and  
> replace all +'s in it with %2B before percent-decoding.
> 
> 4. Another alternative is to manually parse the query string and treat the  
> data as mailto data where a '+' is a '+' instead of http data where a '+'  
> is a ' '.
> 
> I can confirm the problem in SM. I didn't test your patch though. Does it  
> do the fix for all values (subject, send_to, send_to_cc, send_to_bcc and  
> body)?
> 
> For example, if you compose a message and put:
> 
> --------
> <mailto:1+2?subject=1+2&body=1+2&cc=1+2&bcc=1+2>

SM doesn't recognize this as a valid mailto: link, either with or
without my patch.  

> <mailto:?to=1+2&subject=1+2&body=1+2&cc=1+2&bcc=1+2>


It sees this as a mailto: link and sets:

 To: 1+2&amp;subject 

with my patch.  Looks like I am urlencoding the ampersand unnecessarily.


Without my patch, it sets:

 To: 1 2 Cc: 1 2 Bcc: 1 2 Subject: 1 2
 Body: 1 2 

That is, it sets them all properly, albeit without the plus sign..



> --------
> 
> in the message body and send it, in the received message, will SM show  
> them via the following markup?
> 
> <pre>
> &lt;<a  
> href="http://smserver.com/src/compose.php?send_to=1%2B2&subject=1%2B2&body=1%2B2&cc=1%2B2&bcc=1%2B2">mailto:1+2?subject=1+2&body=1+2&cc=1+2&bcc=1+2</a>&gt;
> 
> &lt;<a  
> href="http://smserver.com/src/compose.php?send_to=1%2B2&subject=1%2B2&body=1%2B2&cc=1%2B2&bcc=1%2B2">mailto:?to=1+2&subject=1+2&body=1+2&cc=1+2&bcc=1+2</a>&gt;
> </pre>
> 
> That might be something to test.
> 
> -- 
> Michael

Thanks for your feedback.  What do you think the appropriate way
to adress this is?  I'd hate to only urlencode specific characters.

Thanks,
-Jon


------------------------------------------------------------------------------
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
-----
squirrelmail-devel mailing list
Posting guidelines: http://squirrelmail.org/postingguidelines
List address: [email protected]
List archives: http://news.gmane.org/gmane.mail.squirrelmail.devel
List info (subscribe/unsubscribe/change options): https://lists.sourceforge.net/lists/listinfo/squirrelmail-devel