You done found a bug. I reported it as Issue #358.
org.restlet.util.Template.java (rev 2116) 1066 << final String HEXA = DIGIT + "ABCDEF" >> final String HEXA = DIGIT + "ABCDEFabcdef" Good catch! ----- Original Message ----- From: "Paul Kaiser" <[EMAIL PROTECTED]> To: [email protected] Sent: Wednesday, September 5, 2007 9:14:05 PM (GMT-0500) America/New_York Subject: Re: Routing and URL decoding I don't have much in terms of a restlet-based application. I started looking at Restlets about a week ago to build services on top of a legacy database. I took the code from the tutorial page (item 11) and dropped it into an empty main() method. I've attached the source for your reference. The test results are " http://localhost:8182/users/foo%2fbar/orders/aoao " produces "Account of user "foo"" " http://localhost:8182/users/foo%2Fbar/orders/aoao " produces "Order "aoao" for user "foo%2Fbar"". I believe the latter is the correct behavior given that "%2F" should not have the special meaning that "/" does in a path. I haven't had a chance to trace the code to see what is causing the behavior. Just that its really odd that case should make a difference. Thanks, Paul ----- Original Message ---- From: Rob Heittman <[EMAIL PROTECTED]> To: [email protected] Sent: Wednesday, September 5, 2007 9:52:55 AM Subject: Re: Routing and URL decoding Been looking for it, but I'm trying to find where the Restlet framework does any URL decoding at all ... my apps get a lot of special characters in URIs, and we have to decode them explicitly with URLDecoder.decode(...) or Reference.decode(...) which wraps it. Restlet's Reference class is pretty adamant about not doing any transparent decoding of entities. java.net.URLDecoder.decode(...) should support both upper and lowercase hex in all compliant JREs. Can you say any more about the application? I'd suspect something below the Router level. Is it using Restlet-supplied Directory etc. to find some sort of static files, or is there custom code for Finders and Resources (or generic Restlets doing the same functions). ----- Original Message ----- From: "Paul Kaiser" <[EMAIL PROTECTED]> To: "Discuss Restlets" <[email protected]> Sent: Tuesday, September 4, 2007 10:10:03 PM (GMT-0500) America/New_York Subject: Routing and URL decoding I have a "special" situation in an application that allows "/" characters in usernames. If I define a URI that include a username as a path element, the Router gets confused when presented with a URI with an encoded "/" (%2f); something like http://localhost:8182/users/foo%2fbar/orders . Changing it to http://localhost:8182/users/foo%2Fbar/orders works, however. Is there a reason for the case sensitivity? RFC 1738 suggests that lowercase hexadecimal be accepted also. I happened on this behavior when typing URIs into a browser. I'm going to look at a possible work-around if I need it; probably a custom filter placed ahead of the router. If there is a better approach, please let me know. Thanks, Paul ____________________________________________________________________________________ Boardwalk for $500? In 2007? Ha! Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games. http://get.games.yahoo.com/proddesc?gamekey=monopolyherenow Looking for a deal? Find great prices on flights and hotels with Yahoo! FareChase.
