Re: JSR311: Bug in resource method matching algorithm [old contains little mistake]

From: Stephan Koops <>
Date: Mon, 22 Sep 2008 20:08:33 +0200

Hi Marc,

yes, you are right.

best regards

Marc Hadley schrieb:
> I don't think there's a problem. When you enter step 3 all of the
> methods in the set M already have the same regex (so there's no need
> to sort on it):
> - if you enter from 2.(a) then the methods don't have a regex
> - if you enter from 2.(h) then you've already chosen the best matching
> regex in 2.(g) and all the methods in M have that exact regex.
> Marc.
> On Sep 20, 2008, at 12:36 PM, Stephan Koops wrote:
>> Hi,
>> (change against previous email: from "thrid bullet" to "first bullet")
>> sorry that it is so late, but I think I found a bug in the JAX-RS
>> specification: We forget to add "the number of capturing groups with
>> non-default regular expressions (i.e. not '([^/]+?)') as the tertiary
>> key (descending order)" to chapter 3.7.2 (Request Matching), point 3
>> (identify method) as in the points one and two. It must be placed as
>> first bullet to 3 (b) "Sort M in descending order as follows".
>> I hope it's not too late.
>> best regards
>> Stephan