Re: CXF JAXRS Client : HttpAsyncClient and KeepAlive
I've updated the JAXRS test, JAXRSAsyncClientTest:
@Test
public void testPatchBook() throws Exception {
String address = "http://localhost:; + PORT + "/bookstore/patch";
WebClient wc = WebClient.create(address);
wc.type("application/xml");
WebClient.getConfig(wc).getRequestContext().put("use.async.http.conduit", true);
WebClient.getConfig(wc).getHttpConduit().getClient().setReceiveTimeout(1000L);
Book book = wc.invoke("PATCH", new Book("Patch", 123L),
Book.class);
assertEquals("Patch", book.getName());
wc.close();
}
and after putting a breakpoint in BookStore patchBook method I see it
staying on the breakpoint without timing out...
Cheers, Sergey
On 12/04/16 14:01, Sergey Beryozkin wrote:
Hi
I've only opened an issue to make sure this issue is being tracked
https://issues.apache.org/jira/browse/CXF-6846
Cheers, Sergey
On 12/04/16 13:50, chandra shekhar wrote:
Hi Sergey,
Sorry to bother you.
Did you find something ?
Regards,
Chandra Shekhar
--
View this message in context:
http://cxf.547215.n5.nabble.com/CXF-JAXRS-Client-HttpAsyncClient-and-KeepAlive-tp5747402p5767808.html
Sent from the cxf-user mailing list archive at Nabble.com.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: CXF JAXRS Client : HttpAsyncClient and KeepAlive
Hi I've only opened an issue to make sure this issue is being tracked https://issues.apache.org/jira/browse/CXF-6846 Cheers, Sergey On 12/04/16 13:50, chandra shekhar wrote: Hi Sergey, Sorry to bother you. Did you find something ? Regards, Chandra Shekhar -- View this message in context: http://cxf.547215.n5.nabble.com/CXF-JAXRS-Client-HttpAsyncClient-and-KeepAlive-tp5747402p5767808.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: [ 2.7.15] Matrix parameter followed by query parameter are merged?
Just double checked that using WebClient directly (in that test I pointed out to) works as well. So IMHO the problem is somewhere at the lower level. Sergey On 07/04/16 12:24, Sergey Beryozkin wrote: Hi One thing I'd like to point out - CXF checks the matrix parameters on the request URI, obtained from: (HttpServletRequest)req.getRequestURI(); which must not return the query part. So the source of the problem is that either: - the container mistakenly returns the query as part of the request URI - the question mark is mistakenly encoded on the client side, which would result the combined value after being decoded on the client side. Can you please confirm with HTTP trace enabled (TCP trace utility) that a question mark is either encoded or not ? Thanks, Sergey On 06/04/16 17:20, Sergey Beryozkin wrote: Hi Christian I simply updated this test: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/JAXRSClientServerBookTest.java#L1954 I tried "http://localhost:; + PORT + "/bookstore/segment/matrix;first=12;second=3?loc=123" and only with a single matrix param: "http://localhost:; + PORT + "/bookstore/segment/matrix;first=12?loc=123" and then I put a breakpoint in: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/BookStore.java#L764 and checked '?loc=123' was not there. The test runs on top of Jetty Cheers, Sergey On 06/04/16 17:00, Christian Balzer wrote: Hi Sergey, Thank you for looking into this. We are using an internal framework at work that makes use of CXF - which means that it's not that easy to just swap CXF out, unfortunately. Having said that, a colleague started work to replace CXF 2.x by CXF 3.x today. But it'll be a few days before everything will be ported over; so we can't yet say whether CXF 3.x is affected or not. In the meantime, we did create another project from scratch, and implemented an endpoint for GET. Here are the results: 1 - USING CXF CLIENT VIA LOCAL (INTEGRATION TEST) * matrix before query: ;m=1?q=2 * no QUERY_STRING entry in XMLMessage map * result (fails): * matrix param populated with matrix param value concatenated with all of query string: 1?q=2 * query param populated correctly: 2 2 - USING CFX CLIENT VIA HTTP TO SERVER * matrix before query: ;m=1?q=2 * QUERY_STRING present in XMLMessage map * all works 3 - USING CHROME POSTMAN EXTENSION TO SERVER * matrix before query: ;m=1?q=2 * query param value: 2 * marix param vaule: 1 * all works 4 - USING CHROME POSTMAN EXTENSION TO SERVER * query before matrix: ?q=2;m=1 * query string contains the matrix parma: 2;m=1 * query param value: * matrix param value: 2;m=1 * arguably, this one could be considered ok if you argued that the matrix param relates to the query param Running via Chrome Postman extension to the server, here is the TCP dump (I removed proprietary headers): GET /bar/test?q=2;m=1 HTTP/1.1 Host: 10.221.29.152:7001 Connection: keep-alive Accept: application/json Cache-Control: no-cache Content-Type: application/json User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.82 Safari/537.36 Postman-Token: 1bc0c914-80a3-1829-ec9d-8352118792ec Accept-Encoding: gzip, deflate, sdch Accept-Language: en-GB,en;q=0.8,en-US;q=0.6,de;q=0.4 HTTP/1.1 200 OK Cache-Control: private, no-cache, no-store, no-transform Date: Wed, 06 Apr 2016 13:59:38 GMT Content-Length: 0 Operation: (GET)/bar/test X-Powered-By: Servlet/3.0 JSP/2.2 Also, org.apache.cxf.message.Message holds (amongst others): QUERY_STRING -> "q=2;m=1" "HTTP.REQUEST" -> "[ GET /bar/test?q=2;m=1 HTTP/1.1 ... ]" Now, why it fails for LOCAL transport for matrix before query, and for query before matrix via HTTP, we don't quite know. But I think it's fair to say it's probably nothing to do with the encoding of the question mark... (The Message map for test case 1 holds "org.apache.cxf.request.uri" -> "/dates/test;one=1?two=2".) Could you maybe share your test project with us, so we can compare what you are doing, please? :-) Kind regards, Christian On Tue, Apr 5, 2016 at 1:37 PM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi I've experimented with a local test, I can not reproduce it. Can you please add some logging on the client/server, or preferably, pass it via some TCP tracer, I wonder if the question mark ends up being encoded, etc, may be the client (HttpUrlConnection) gets confused about PUT and query parameters. What about GET ? Can you also double check CXF 3.1.x (just to confirm this issue is 2.7.x specific) Cheers, Sergey On 05/04/16 11:35, Christian Balzer wrote: Hi all, We are restricted to the 2.x branch of cxf. When we call our REST endpoint and provide a matrix parameter (;id=123), followed by a query parameter (?loc=abc), cxf populates the String
Re: [ 2.7.15] Matrix parameter followed by query parameter are merged?
Hi One thing I'd like to point out - CXF checks the matrix parameters on the request URI, obtained from: (HttpServletRequest)req.getRequestURI(); which must not return the query part. So the source of the problem is that either: - the container mistakenly returns the query as part of the request URI - the question mark is mistakenly encoded on the client side, which would result the combined value after being decoded on the client side. Can you please confirm with HTTP trace enabled (TCP trace utility) that a question mark is either encoded or not ? Thanks, Sergey On 06/04/16 17:20, Sergey Beryozkin wrote: Hi Christian I simply updated this test: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/JAXRSClientServerBookTest.java#L1954 I tried "http://localhost:; + PORT + "/bookstore/segment/matrix;first=12;second=3?loc=123" and only with a single matrix param: "http://localhost:; + PORT + "/bookstore/segment/matrix;first=12?loc=123" and then I put a breakpoint in: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/BookStore.java#L764 and checked '?loc=123' was not there. The test runs on top of Jetty Cheers, Sergey On 06/04/16 17:00, Christian Balzer wrote: Hi Sergey, Thank you for looking into this. We are using an internal framework at work that makes use of CXF - which means that it's not that easy to just swap CXF out, unfortunately. Having said that, a colleague started work to replace CXF 2.x by CXF 3.x today. But it'll be a few days before everything will be ported over; so we can't yet say whether CXF 3.x is affected or not. In the meantime, we did create another project from scratch, and implemented an endpoint for GET. Here are the results: 1 - USING CXF CLIENT VIA LOCAL (INTEGRATION TEST) * matrix before query: ;m=1?q=2 * no QUERY_STRING entry in XMLMessage map * result (fails): * matrix param populated with matrix param value concatenated with all of query string: 1?q=2 * query param populated correctly: 2 2 - USING CFX CLIENT VIA HTTP TO SERVER * matrix before query: ;m=1?q=2 * QUERY_STRING present in XMLMessage map * all works 3 - USING CHROME POSTMAN EXTENSION TO SERVER * matrix before query: ;m=1?q=2 * query param value: 2 * marix param vaule: 1 * all works 4 - USING CHROME POSTMAN EXTENSION TO SERVER * query before matrix: ?q=2;m=1 * query string contains the matrix parma: 2;m=1 * query param value: * matrix param value: 2;m=1 * arguably, this one could be considered ok if you argued that the matrix param relates to the query param Running via Chrome Postman extension to the server, here is the TCP dump (I removed proprietary headers): GET /bar/test?q=2;m=1 HTTP/1.1 Host: 10.221.29.152:7001 Connection: keep-alive Accept: application/json Cache-Control: no-cache Content-Type: application/json User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.82 Safari/537.36 Postman-Token: 1bc0c914-80a3-1829-ec9d-8352118792ec Accept-Encoding: gzip, deflate, sdch Accept-Language: en-GB,en;q=0.8,en-US;q=0.6,de;q=0.4 HTTP/1.1 200 OK Cache-Control: private, no-cache, no-store, no-transform Date: Wed, 06 Apr 2016 13:59:38 GMT Content-Length: 0 Operation: (GET)/bar/test X-Powered-By: Servlet/3.0 JSP/2.2 Also, org.apache.cxf.message.Message holds (amongst others): QUERY_STRING -> "q=2;m=1" "HTTP.REQUEST" -> "[ GET /bar/test?q=2;m=1 HTTP/1.1 ... ]" Now, why it fails for LOCAL transport for matrix before query, and for query before matrix via HTTP, we don't quite know. But I think it's fair to say it's probably nothing to do with the encoding of the question mark... (The Message map for test case 1 holds "org.apache.cxf.request.uri" -> "/dates/test;one=1?two=2".) Could you maybe share your test project with us, so we can compare what you are doing, please? :-) Kind regards, Christian On Tue, Apr 5, 2016 at 1:37 PM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi I've experimented with a local test, I can not reproduce it. Can you please add some logging on the client/server, or preferably, pass it via some TCP tracer, I wonder if the question mark ends up being encoded, etc, may be the client (HttpUrlConnection) gets confused about PUT and query parameters. What about GET ? Can you also double check CXF 3.1.x (just to confirm this issue is 2.7.x specific) Cheers, Sergey On 05/04/16 11:35, Christian Balzer wrote: Hi all, We are restricted to the 2.x branch of cxf. When we call our REST endpoint and provide a matrix parameter (;id=123), followed by a query parameter (?loc=abc), cxf populates the String argument annotated with the @MatrixParam bind annotation with the merged String of matrix value and query parameter, i.e. 123?loc=abc Any idea why and what we can do to stop that from happening? We are basi
Re: [ 2.7.15] Matrix parameter followed by query parameter are merged?
Hi Christian
I simply updated this test:
https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/JAXRSClientServerBookTest.java#L1954
I tried
"http://localhost:; + PORT +
"/bookstore/segment/matrix;first=12;second=3?loc=123"
and only with a single matrix param:
"http://localhost:; + PORT + "/bookstore/segment/matrix;first=12?loc=123"
and then I put a breakpoint in:
https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/java/org/apache/cxf/systest/jaxrs/BookStore.java#L764
and checked '?loc=123' was not there.
The test runs on top of Jetty
Cheers, Sergey
On 06/04/16 17:00, Christian Balzer wrote:
Hi Sergey,
Thank you for looking into this. We are using an internal framework at
work that makes use of CXF - which means that it's not that easy to
just swap CXF out, unfortunately.
Having said that, a colleague started work to replace CXF 2.x by CXF
3.x today. But it'll be a few days before everything will be ported
over; so we can't yet say whether CXF 3.x is affected or not.
In the meantime, we did create another project from scratch, and
implemented an endpoint for GET.
Here are the results:
1 - USING CXF CLIENT VIA LOCAL (INTEGRATION TEST)
* matrix before query: ;m=1?q=2
* no QUERY_STRING entry in XMLMessage map
* result (fails):
* matrix param populated with matrix param value concatenated with
all of query string: 1?q=2
* query param populated correctly: 2
2 - USING CFX CLIENT VIA HTTP TO SERVER
* matrix before query: ;m=1?q=2
* QUERY_STRING present in XMLMessage map
* all works
3 - USING CHROME POSTMAN EXTENSION TO SERVER
* matrix before query: ;m=1?q=2
* query param value: 2
* marix param vaule: 1
* all works
4 - USING CHROME POSTMAN EXTENSION TO SERVER
* query before matrix: ?q=2;m=1
* query string contains the matrix parma: 2;m=1
* query param value:
* matrix param value: 2;m=1
* arguably, this one could be considered ok if you argued that the
matrix param relates to the query param
Running via Chrome Postman extension to the server, here is the TCP
dump (I removed proprietary headers):
GET /bar/test?q=2;m=1 HTTP/1.1
Host: 10.221.29.152:7001
Connection: keep-alive
Accept: application/json
Cache-Control: no-cache
Content-Type: application/json
User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64)
AppleWebKit/537.36 (KHTML, like Gecko) Chrome/48.0.2564.82
Safari/537.36
Postman-Token: 1bc0c914-80a3-1829-ec9d-8352118792ec
Accept-Encoding: gzip, deflate, sdch
Accept-Language: en-GB,en;q=0.8,en-US;q=0.6,de;q=0.4
HTTP/1.1 200 OK
Cache-Control: private, no-cache, no-store, no-transform
Date: Wed, 06 Apr 2016 13:59:38 GMT
Content-Length: 0
Operation: (GET)/bar/test
X-Powered-By: Servlet/3.0 JSP/2.2
Also, org.apache.cxf.message.Message holds (amongst others):
QUERY_STRING -> "q=2;m=1"
"HTTP.REQUEST" -> "[
GET /bar/test?q=2;m=1 HTTP/1.1
...
]"
Now, why it fails for LOCAL transport for matrix before query, and for
query before matrix via HTTP, we don't quite know.
But I think it's fair to say it's probably nothing to do with the
encoding of the question mark...
(The Message map for test case 1 holds "org.apache.cxf.request.uri" ->
"/dates/test;one=1?two=2".)
Could you maybe share your test project with us, so we can compare
what you are doing, please? :-)
Kind regards,
Christian
On Tue, Apr 5, 2016 at 1:37 PM, Sergey Beryozkin <sberyoz...@gmail.com> wrote:
Hi
I've experimented with a local test, I can not reproduce it.
Can you please add some logging on the client/server, or preferably, pass it
via some TCP tracer, I wonder if the question mark ends up being encoded,
etc, may be the client (HttpUrlConnection) gets confused about PUT and query
parameters.
What about GET ?
Can you also double check CXF 3.1.x (just to confirm this issue is 2.7.x
specific)
Cheers, Sergey
On 05/04/16 11:35, Christian Balzer wrote:
Hi all,
We are restricted to the 2.x branch of cxf.
When we call our REST endpoint and provide a matrix parameter
(;id=123), followed by a query parameter (?loc=abc), cxf populates the
String argument annotated with the @MatrixParam bind annotation with
the merged String of matrix value and query parameter, i.e.
123?loc=abc
Any idea why and what we can do to stop that from happening?
We are basically using the following to fire up the request (via local
protocol) from our integration test:
client.path("foo/987")
.type(MediaType.APPLICATION_JSON)
.accept(MediaType.APPLICATION_JSON)
// adding headers here...
.matrix("id", "123")
.query("loc", "abc")
.put(TEST_PAYLOAD);
Here is the method signature we are using for the endpoint:
@PUT
@Consumes({APPLICATION_JSON})
@Produces({"application/json;charset=UTF-8"})
public Response doStuff(
@BeanParam B b, // pulls out 987
@MatrixParam("id") String id,
Re: [ 2.7.15] Matrix parameter followed by query parameter are merged?
Hi
I've experimented with a local test, I can not reproduce it.
Can you please add some logging on the client/server, or preferably,
pass it via some TCP tracer, I wonder if the question mark ends up being
encoded, etc, may be the client (HttpUrlConnection) gets confused about
PUT and query parameters.
What about GET ?
Can you also double check CXF 3.1.x (just to confirm this issue is 2.7.x
specific)
Cheers, Sergey
On 05/04/16 11:35, Christian Balzer wrote:
Hi all,
We are restricted to the 2.x branch of cxf.
When we call our REST endpoint and provide a matrix parameter
(;id=123), followed by a query parameter (?loc=abc), cxf populates the
String argument annotated with the @MatrixParam bind annotation with
the merged String of matrix value and query parameter, i.e.
123?loc=abc
Any idea why and what we can do to stop that from happening?
We are basically using the following to fire up the request (via local
protocol) from our integration test:
client.path("foo/987")
.type(MediaType.APPLICATION_JSON)
.accept(MediaType.APPLICATION_JSON)
// adding headers here...
.matrix("id", "123")
.query("loc", "abc")
.put(TEST_PAYLOAD);
Here is the method signature we are using for the endpoint:
@PUT
@Consumes({APPLICATION_JSON})
@Produces({"application/json;charset=UTF-8"})
public Response doStuff(
@BeanParam B b, // pulls out 987
@MatrixParam("id") String id,
@QueryParam("loc") String loc,
MySchema mySchema,
@HeaderParam("bar") String header,
@Context UriInfo uriInfo)
throws Exception {
This results in loc being set to "abc" and id to "123?loc=abc"... :-(
Any help is much appreciated!
Kind regards,
Christian
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Dealing with Jettison array serialization issues
Hi Jettison is XMLStreamWriter, no type information is available. If it does accumulate more than one value for a given property it can guess it must be a list, otherwise it does not know if it is a simple or complex property. Sergey On 04/04/16 12:56, Vjacheslav V. Borisov wrote: Hi! Looking at http://cxf.apache.org/docs/jax-rs-data-bindings.html#JAX-RSDataBindings-Jettison paragraph "Dealing with Jettison array serialization issues" does this is only way to configure List<> properties serializtion? May be there is some undocumented property to do this transparently, eg all List<> properties serialize as array, no matter how many elements it has (1 or more)
Re: URL that contains "service" returns CXF Service List in Karaf
"org.apache.cxf.osgi" On 29/03/16 13:41, josephkampf wrote: I'm using Karaf. What would the PID be? -- View this message in context: http://cxf.547215.n5.nabble.com/URL-that-contains-service-returns-CXF-Service-List-in-Karaf-tp5767298p5767333.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: streaming in a large post body
Hi Benson I'd consider using InputStream (or JAXP Source if applicable) as a method parameter or use attachments given that CXF saves some of the data to the temp files... Sergey On 28/03/16 19:07, Benson Margulies wrote: If I want to handle posts with more payload than I want to have in memory all at once (with REST), what's recommended? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: URL that contains "service" returns CXF Service List in Karaf
Hi org.apache.cxf.servlet.service-list-path is a property that you can set in Karaf, to something like "/services-list", etc Thanks, Sergey On 28/03/16 14:13, josephkampf wrote: I have a service that has some extended path information that include the work service Example: http://localhost:8181/cxf/abcd/services The Rest service is abcd and extended path info is /services. I see that this is being done in the web.xml that CXF is using for the CXFServlet. How can I override this in Karaf? Can I configure CXF to point to a different web.xml file? My current version of CXF is 2.7.11, however I will be moving to 3.1.4 soon. So a solution that works for both or a solution for each would be much appreciated. Thanks, Joe Kampf -- View this message in context: http://cxf.547215.n5.nabble.com/URL-that-contains-service-returns-CXF-Service-List-in-Karaf-tp5767298.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: oauth2 server side cxf configuration
Can you clarify please what you'd like to do, as far as enabling OAuth2 is concerned ? Cheers, Sergey -- View this message in context: http://cxf.547215.n5.nabble.com/oauth2-server-side-cxf-configuration-tp5767107p5767152.html Sent from the cxf-user mailing list archive at Nabble.com.
Re: CXF JAXRS Client : HttpAsyncClient and KeepAlive
Well, I suppose you set as per the documentation. I'm not sure why it has no effect with HttpAsyncClient, I'll experiment a bit later Sergey -- View this message in context: http://cxf.547215.n5.nabble.com/CXF-JAXRS-Client-HttpAsyncClient-and-KeepAlive-tp5747402p5767151.html Sent from the cxf-user mailing list archive at Nabble.com.
Re: CXF JAXRS Client : HttpAsyncClient and KeepAlive
Hi How do you set it ? Cheers, Sergey On 23/03/16 08:51, chandra shekhar wrote: Hi Sergey, I set CONNECTION_TTL but I do not see any affect. I see that connection always close in a minute which I see is receiveTimeout default value. Any suggestions here. -- View this message in context: http://cxf.547215.n5.nabble.com/CXF-JAXRS-Client-HttpAsyncClient-and-KeepAlive-tp5747402p5767130.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: catalogLocation in JAXB/JSON Provider and @SchemaValidation
Thanks for the analysis and the patch, I'm traveling so will look into it in detail next week (though might apply the patch earlier), does your patch addresses all of your concerns you raised in this thread ? Cheers, Sergey On 23/03/16 08:21, Vjacheslav V. Borisov wrote: Don't know, how should this fixed (or should not) E.g. , code like this could fix that, but seems ugly: resolvedLocation = catalogResolver.resolvePublic(publicId!=null?publicId:namespaceURI, systemId); Found where it takes place in case of cxf-wadl2java-plugin, please see SchemaCompilerImpl.bind(), it passes namespaceURI as publicId parameter, while SchemaHandler passes publicId as publicId, which is null, and resolve does not takes place. http://grepcode.com/file/repo1.maven.org/maven2/com.sun.xml.bind/jaxb-xjc/2.2.11/com/sun/tools/xjc/api/impl/s2j/SchemaCompilerImpl.java I submitted my patch https://issues.apache.org/jira/browse/CXF-6840 -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Signing and encrypting JSON payload for REST requests over a browser
Hi, On 18/03/16 00:21, Giriraj Bhojak wrote: Thank you Sergey. I went through the spec. It mentions that the spec is not stable yet and is subject to change. Would you know if it is widely used? There are two specs involved here, JOSE and WebCrypto, the former is stable and is already quite widely used, though mostly in OAuth2 flows, but JOSE is independent of OAuth2. WebCrypto is a browser specific mechanism on how to get the keys/etc, the demo worked for me in Firefox/Chrome, not sure about the other browsers, though I might've tried IE too when trying on Windows, do not remember now. I think it is unlikely anything but some minor details will get changed there. If you'd like to start doing signing/encrypting within a script running inside a browser then I guess you have to be prepared at this stage to go some not-very standard-safe path. I was hoping to use one of the JavaScript tools such as jsrrsasign, but looks like it is our of picture. Would you be able to share the source code/API details of the demo that you gave in Apache Con? On the demo page, click at the WebCrypto++ icon and it will bring you to a page with a link to the source code. In my demo I only replaced the server code which validates JWS signatures, the code that signs the data from within a script was the same as in the original demo. I have not experimented with that script, I only wanted to demo the JOSE JWS interoperability between a non-CXF client (the script) and CXF server Could you please expand on the trusted server approach you mentioned in the follow-up? If you can not sign directly within the script then post the data to be signed to the trusted server that will do it for you and return the signed data. HTH, Sergey Thank you for responding to my queries. Thanks, Giriraj Thanks, Giriraj On Mar 17, 2016 6:10 PM, "Sergey Beryozkin" <sberyoz...@gmail.com> wrote: Or a browser may ask a trusted server to help with it, and get this server returning a String representing a JOSE payload, then script then forward it somewhere else... Sergey On 17/03/16 21:35, Sergey Beryozkin wrote: Hi You may be talking about WebCrypto. If you have a CXF client sending JSON, then JWE/JWS protecting it is easy enough, but you have a script running in a browser then this script have no access to the key stores, unless it is a WebCrypto aware browser and most of them are by now AFAIK, See this demo: https://test.webpki.org/WCPPSignatureDemo/signcmd (it says a password is 1234). It shows an interaction between a WebCrypto (https://www.w3.org/TR/WebCryptoAPI/) browser based client and a regular Java HTTP server, the data are signed, using JOSE (JWS Compact) as one option. I actually presented this demo at Apache Con NA 2015, except I replaced the demo server with a CXF JWS-enabled server. Sergey On 17/03/16 15:45, Giriraj Bhojak wrote: Hi, I have been struggling with a basic question related to using signing and encryption for REST services. If the REST call (using JSON) happens over http or https via a browser, how can I ensure that JSON payload is signed and encrypted, just like a SOAP request that is signed and encrypted? Is there a JavaScript component that I can use to implement JOSE for browser based REST requests? Or am I interpreting this in a wrong way? Thanks, Giriraj. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Signing and encrypting JSON payload for REST requests over a browser
Hi You may be talking about WebCrypto. If you have a CXF client sending JSON, then JWE/JWS protecting it is easy enough, but you have a script running in a browser then this script have no access to the key stores, unless it is a WebCrypto aware browser and most of them are by now AFAIK, See this demo: https://test.webpki.org/WCPPSignatureDemo/signcmd (it says a password is 1234). It shows an interaction between a WebCrypto (https://www.w3.org/TR/WebCryptoAPI/) browser based client and a regular Java HTTP server, the data are signed, using JOSE (JWS Compact) as one option. I actually presented this demo at Apache Con NA 2015, except I replaced the demo server with a CXF JWS-enabled server. Sergey On 17/03/16 15:45, Giriraj Bhojak wrote: Hi, I have been struggling with a basic question related to using signing and encryption for REST services. If the REST call (using JSON) happens over http or https via a browser, how can I ensure that JSON payload is signed and encrypted, just like a SOAP request that is signed and encrypted? Is there a JavaScript component that I can use to implement JOSE for browser based REST requests? Or am I interpreting this in a wrong way? Thanks, Giriraj. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Apache CXF 3.0.7 Webclient is taking too long to process the response
Hi Looks like something is seriously wrong with the client configuration, or, perhaps WebClient is targeting a different service endpoint that indeed takes 10 min to return Sergey On 18/03/16 14:44, Khare, Aparna wrote: Dear Colleagues, I'm using Apache cxf 3.0.7 and I use weblient to create the HTTP client webclient.get()to get the response. The response is returned in more than 10 mins whereas in the browser it takes seconds. Thanks, Aparna -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Signing and encrypting JSON payload for REST requests over a browser
Or a browser may ask a trusted server to help with it, and get this server returning a String representing a JOSE payload, then script then forward it somewhere else... Sergey On 17/03/16 21:35, Sergey Beryozkin wrote: Hi You may be talking about WebCrypto. If you have a CXF client sending JSON, then JWE/JWS protecting it is easy enough, but you have a script running in a browser then this script have no access to the key stores, unless it is a WebCrypto aware browser and most of them are by now AFAIK, See this demo: https://test.webpki.org/WCPPSignatureDemo/signcmd (it says a password is 1234). It shows an interaction between a WebCrypto (https://www.w3.org/TR/WebCryptoAPI/) browser based client and a regular Java HTTP server, the data are signed, using JOSE (JWS Compact) as one option. I actually presented this demo at Apache Con NA 2015, except I replaced the demo server with a CXF JWS-enabled server. Sergey On 17/03/16 15:45, Giriraj Bhojak wrote: Hi, I have been struggling with a basic question related to using signing and encryption for REST services. If the REST call (using JSON) happens over http or https via a browser, how can I ensure that JSON payload is signed and encrypted, just like a SOAP request that is signed and encrypted? Is there a JavaScript component that I can use to implement JOSE for browser based REST requests? Or am I interpreting this in a wrong way? Thanks, Giriraj. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: How to manage resource owner login in CXF
Hi On 11/03/16 14:52, matteo wrote: Thank you for your reply! In the end I simply registered a ExceptionMapper and now I redirect the resource owner to our login page (implemented as a wab). The problem I'm still facing is that when the login wab performs the /authorize/ invocation again (this time with the expected basic authentication header in place), the authorization invocation fails because the securityContext.getUserPrincipal() returns null. What is the right way in CXF to correctly populate the SecurityContext with a sound user principal? If you use the servlet security then the container will set up the principal itself, otherwise you need to set a CXF filter, ex, JAX-RS 2.0 filter, which will parse the header and set up SecurityContext Sergey Many thanks again, matte - matteo -- View this message in context: http://cxf.547215.n5.nabble.com/How-to-manage-resource-owner-login-in-CXF-tp5766808p5766837.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: WebClient getting connection already closed
) at org.junit.runners.ParentRunner.access$000(ParentRunner.java:58) at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:268) at org.junit.internal.runners.statements.RunBefores.evaluate(RunBefores.java:26) at org.junit.runners.ParentRunner.run(ParentRunner.java:363) at org.junit.runners.Suite.runChild(Suite.java:128) at org.junit.runners.Suite.runChild(Suite.java:27) at org.junit.runners.ParentRunner$3.run(ParentRunner.java:290) at org.junit.runners.ParentRunner$1.schedule(ParentRunner.java:71) at org.junit.runners.ParentRunner.runChildren(ParentRunner.java:288) at org.junit.runners.ParentRunner.access$000(ParentRunner.java:58) at org.junit.runners.ParentRunner$2.evaluate(ParentRunner.java:268) at org.junit.runners.ParentRunner.run(ParentRunner.java:363) at org.junit.runner.JUnitCore.run(JUnitCore.java:137) at com.intellij.junit4.JUnit4IdeaTestRunner.startRunnerWithArgs(JUnit4IdeaTestRunner.java:69) at com.intellij.rt.execution.junit.JUnitStarter.prepareStreamsAndStart(JUnitStarter.java:234) at com.intellij.rt.execution.junit.JUnitStarter.main(JUnitStarter.java:74) Caused by: java.lang.IllegalStateException: Buffer already closed for writing at org.apache.cxf.transport.http.asyncclient.SharedOutputBuffer.write(SharedOutputBuffer.java:231) at org.apache.cxf.transport.http.asyncclient.AsyncHTTPConduit$AsyncWrappedOutputStream$1.write(AsyncHTTPConduit.java:431) at org.apache.cxf.io.AbstractWrappedOutputStream.write(AbstractWrappedOutputStream.java:51) at org.apache.cxf.io.AbstractThresholdOutputStream.write(AbstractThresholdOutputStream.java:69) at org.apache.cxf.transport.http.HTTPConduit$WrappedOutputStream.close(HTTPConduit.java:1332) ... 59 more -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: How to manage resource owner login in CXF
As far as this interceptor is concerned, it can be configured with custom callback handlers, I don't remember the name of the property right now, but you can set whatever custom handlers are required. Cheers, Sergey On 11/03/16 17:26, matteo wrote: Ok, I narrowed down a little bit my question. I see that it is possible to leverage Jaas to populate the SecurityContext principal via the JAASAuthenticationFilter. Now the point is I would like to manage my users and passwords via datasources as explained in karaf documentation <https://karaf.apache.org/manual/latest/developers-guide/security-framework.html> . The point is that my users' passwords are stored in the database as encrypted tokens. So my previous question becomes: how to specify a javax.security.auth.callback.CallbackHandler to compute received password hash before jaas login? All other steps should be performed transparently by CXF JAASAuthenticationFilter, right? Thank you very much. matteo - matteo -- View this message in context: http://cxf.547215.n5.nabble.com/How-to-manage-resource-owner-login-in-CXF-tp5766808p5766839.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: How to manage resource owner login in CXF
Hi Matteo I've copied those log statements while working with a demo shipped with my company's distribution (you can see a link to that demo if you check 'CXF OAuth2' in Google). In that demo, a custom security filter is protecting the demo OAuth2 web client, it checks if "Authorization: Basic" is there, if not then it will challenge a user and it is managed by the browser, so the first time the user accesses a web client it is asked by the browser to authenticate, and then after the user is redirected to OAuth2 and then back to the client, the browser is remembering the user authenticating and sets this header itself. In many cases one would really need to have some SSO in place, so that a user does not have to sign in into the web client(s) and OAuth2 server separately. The demo also shows one option, SAML Web SSO with Shibboleth. We also have 2 demos shipped with CXF, basic_oidc and big_query, which show OpenIdConnect RP in action. HTH, Sergey On 10/03/16 14:05, matteo wrote: I'm trying to figure out how to deal with resource owner login procedure in CXF OAuth2 implementation. In the docs (http://cxf.apache.org/docs/jax-rs-oauth2.html) it is stated that /The client application asks the current user (the browser) to go to a new address provided by the Location header and the follow-up request to AuthorizationCodeGrantService will look like this: / /Note that the end user needs to authenticate./ Could you please explain how to deal with resource ownser login in order to provide the required header? What kind of cxf handler (if any) should be registered? Is it possible to serve a custom login form in case the AuthorizationCodeGrantService detects that the Authorization header is missing? Many thanks. matteo - matteo -- View this message in context: http://cxf.547215.n5.nabble.com/How-to-manage-resource-owner-login-in-CXF-tp5766808.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Percent encoded string in @QueryParam
Briefly returning to this thread and clarify why I'm asking to check Jersey UriBuilder or perhaps, even better JAX-RS 2.0 Client API with Jersey and the way the certain combinations are treated: - double encoding of URI (Path, Query) parameters. - double encoding of Form parameters. - treatment of '+' in Path and Query You can also experiment with RestEasy. That will help see how different implementations deal with it on the client side. Sergey On 08/03/16 17:58, Sergey Beryozkin wrote: I started another thread on why '+' is encoded because this thread has been all about encoding a '%' symbol. On 08/03/16 16:50, Sergey Beryozkin wrote: I'm not going to repeat most of what I've already said few times in this thread but only state again that CXF proxies depend on JAX-RS UriBuilder and therefore UriBuilder dictates how %-encoded query (and other) parameters get processed. This is a won't fix 'issue'. However I also mentioned that if you validate that Jersey's (i.e RI's) UriBuilder behaves differently then I will act accordingly. Sergey On 08/03/16 16:44, Volkov Sergey wrote: I agree that UriBuilder implements what is states. Such values are allowed in the query representations. The UriBuilder does not know what the intention of the CXF proxy user is, it may have an encoded representation of a given character stored in the DB and wants to pass it as is via the proxy. "+" value also allowed in query representation, but is encoded for some reason by UriBuilder. So there would be problems with stored in db values if they contains "+" I'm not using jerecy, only org.apache.cxf.jaxrs.client.JAXRSClientFactoryBean to create proxies. I found https://issues.apache.org/jira/browse/CXF-6015 and "url.encode.client.parameters" property, but because org.apache.cxf.jaxrs.client.UrlEncodingParamConverter encodes space as "+", and "+" is encoded as "%2B" by UriBuilder it's also not very useful (Sequence like "+ +" would be converted to "%2B%2B%2B" and decoded as "+++") Also it double encodes @FormParam values. Please forget about CXF proxies. It is down to UriBuilder. http://docs.oracle.com/javaee/6/api/javax/ws/rs/core/UriBuilder.html states: 'Percent encoded values are also recognized where allowed and will not be double encoded.' Such values are allowed in the query representations. The UriBuilder does not know what the intention of the CXF proxy user is, it may have an encoded representation of a given character stored in the DB and wants to pass it as is via the proxy. Experiment with Jersey and if they do double encode (I doubt but...) then we will continue, I think I've given you my best explanations so far, lets continue later on when you have the Jersey test results ready. Sergey On 04/03/16 11:16, Volkov Sergey wrote: i.e CXF proxy do not urlencode parameters? (Instead of it use org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded encoding which is not reversible) And the only reason of this encoding is to avoid reports about correctly encoded parameters? As I said, CXF proxy runtime uses JAX-RS UriBuilder. And UriBuilder does not touch valid percent-encoded representations. And this is how the RS client runtime will continue operating. Sergey On 04/03/16 10:45, Volkov Sergey wrote: "100%25BETTER VALID STRING" instead of "100%BETTER VALID STRING" "100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER VALID STRING" - so actually CXF is over-smart now, when it skips all %xx sequence instead of encode % as %25 "100%VALID STRING" at the moment encodes by cxf as "100%25VALID STRING" In the former case it it has no characters that can be used to form the hex %-encoded representations. The bottom line is, you should not expect the runtime auto-encode '%' given that this symbol plays a special role in having correct URIs being created. If CXF will try to me over-smart then the next day someone will report that it started producing "100%25BETTER VALID STRING" instead of "100%BETTER VALID STRING" Sergey On 04/03/16 07:43, Volkov Sergey wrote: Looks like POST request and @FormParam could be used, but I think this behavior of @QueryParam may lead to tricky bugs in many places where it used 2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email] < http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>: Yes, I don't understand why it does not touch this sequences. It's a bit strange because I can successfully transfer "100%VALID STRING" but not "100%BETTER VALID STRING" Both this string not percent encoded, but second contains %BE which just looks like percent encoded. If instead of org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded cxf just urlencode parameter value when building
Re: Why proxies encode '+' query parameters
On 08/03/16 17:16, Sergey Beryozkin wrote: On 08/03/16 16:56, Sergey Beryozkin wrote: "+" value also allowed in query representation, but is encoded for some reason by UriBuilder. Well, this is not CXF specific. So there would be problems with stored in db values if they contains "+" I'm not using jerecy, only org.apache.cxf.jaxrs.client.JAXRSClientFactoryBean to create proxies. I found https://issues.apache.org/jira/browse/CXF-6015 and "url.encode.client.parameters" property, but because org.apache.cxf.jaxrs.client.UrlEncodingParamConverter encodes space as "+", and "+" is encoded as "%2B" by UriBuilder it's also not very useful (Sequence like "+ +" would be converted to "%2B%2B%2B" and decoded as "+++") Look, I accept there could be some strange combination that are expected to go one way to the wire while the users expects something different. We do out best to accommodate various user requirements, and "url.encode.client.parameters" was one of such steps. If using it creates something unexpected in your case then don't use it. If you have a server that wants to really wants to see "+ " after the request has been decoded then encode "+ " as "%2B%29" and be done with it "%2B%20" -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Why proxies encode '+' query parameters
"+" value also allowed in query representation, but is encoded for some reason by UriBuilder. So there would be problems with stored in db values if they contains "+" I'm not using jerecy, only org.apache.cxf.jaxrs.client.JAXRSClientFactoryBean to create proxies. I found https://issues.apache.org/jira/browse/CXF-6015 and "url.encode.client.parameters" property, but because org.apache.cxf.jaxrs.client.UrlEncodingParamConverter encodes space as "+", and "+" is encoded as "%2B" by UriBuilder it's also not very useful (Sequence like "+ +" would be converted to "%2B%2B%2B" and decoded as "+++") Also it double encodes @FormParam values.
Re: Percent encoded string in @QueryParam
I'm not going to repeat most of what I've already said few times in this
thread but only state again that CXF proxies depend on JAX-RS UriBuilder
and therefore UriBuilder dictates how %-encoded query (and other)
parameters get processed.
This is a won't fix 'issue'. However I also mentioned that if you
validate that Jersey's (i.e RI's) UriBuilder behaves differently then I
will act accordingly.
Sergey
On 08/03/16 16:44, Volkov Sergey wrote:
I agree that UriBuilder implements what is states.
Such values are allowed in the query representations. The UriBuilder
does not know what the intention of the CXF proxy user is, it may have
an encoded representation of a given character stored in the DB and
wants to pass it as is via the proxy.
"+" value also allowed in query representation, but is encoded for some
reason by UriBuilder. So there would be problems with stored in db values
if they contains "+"
I'm not using jerecy, only
org.apache.cxf.jaxrs.client.JAXRSClientFactoryBean to create proxies.
I found https://issues.apache.org/jira/browse/CXF-6015 and
"url.encode.client.parameters" property, but because
org.apache.cxf.jaxrs.client.UrlEncodingParamConverter encodes space as "+",
and "+" is encoded as "%2B" by UriBuilder it's also not very useful
(Sequence like "+ +" would be converted to "%2B%2B%2B" and decoded as
"+++") Also it double encodes @FormParam values.
Please forget about CXF proxies. It is down to UriBuilder.
http://docs.oracle.com/javaee/6/api/javax/ws/rs/core/UriBuilder.html states:
'Percent encoded values are also recognized where allowed and will not
be double encoded.'
Such values are allowed in the query representations. The UriBuilder
does not know what the intention of the CXF proxy user is, it may have
an encoded representation of a given character stored in the DB and
wants to pass it as is via the proxy.
Experiment with Jersey and if they do double encode (I doubt but...)
then we will continue, I think I've given you my best explanations so
far, lets continue later on when you have the Jersey test results ready.
Sergey
On 04/03/16 11:16, Volkov Sergey wrote:
i.e CXF proxy do not urlencode parameters? (Instead of it use
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded encoding
which
is not reversible)
And the only reason of this encoding is to avoid reports about correctly
encoded parameters?
As I said, CXF proxy runtime uses JAX-RS UriBuilder.
And UriBuilder does not touch valid percent-encoded representations.
And this is how the RS client runtime will continue operating.
Sergey
On 04/03/16 10:45, Volkov Sergey wrote:
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
"100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER
VALID
STRING" - so actually CXF is over-smart now, when it skips all %xx
sequence
instead of encode % as %25
"100%VALID STRING" at the moment encodes by cxf as "100%25VALID
STRING"
In the former case it it has no characters that can be used to form
the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs
being
created.
If CXF will try to me over-smart then the next day someone will
report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think
this
behavior of @QueryParam may lead to tricky bugs in many places where
it
used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email]
<
http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID
STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which
just
looks like percent encoded.
If instead of
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both
examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to
the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic s
Re: Is there anything asynchronous about starting a server?
Interesting, I guess more analysis may be needed.
Perhaps Jetty does take few seconds to get 'really' ready after
sf.create() returns, may be it is Jetty version specific ?
Sergey
On 06/03/16 12:51, Benson Margulies wrote:
On Sun, Mar 6, 2016 at 7:08 AM, Sergey Beryozkin <sberyoz...@gmail.com> wrote:
May be the service has not been started (exception is lost, not sure) or
address is wrong ?
A moment later all is well.
Sergey
On 05/03/16 00:20, Benson Margulies wrote:
The code below runs in Karaf. Is there any chance of a caller getting:
No service was found.
after this has executed?
bus = BusFactory.newInstance().createBus();
// set the class loader so that the TCCL is reliably 'us' when
we get called.
bus.setExtension(bundleContext.getBundle().adapt(BundleWiring.class).getClassLoader(),
ClassLoader.class);
JAXRSServerFactoryBean sf = new JAXRSServerFactoryBean();
sf.setBus(bus);
sf.setServiceBean(this);
sf.setAddress("/worker");
server = sf.create();
Re: How CXF helps to redirect to non-matching http request to a static HTML page
Hi If CXFServlet has a wildcard pattern then another idea might be to catch 404 with ExceptionMapper and then redirect with the help of RequestDispatcherProvider (ex, the mapper returns Error bean in Response and the provider is configured to map Error to error.html) - in this case CXFServlet will need to be additionally configured to redirect requests to say error.html to a 'default' servlet. Sergey On 05/03/16 17:15, Jose María Zaragoza wrote: Hi You could use a custom error page for 404 error code Look at error-page and error code in web.xml 2016-03-05 6:23 GMT+01:00 Sumit Arora <sumit1...@gmail.com>: As CXF provides configuration to set default page from : static-welcome-file /main.html This is working fine. My question : *How to redirect to non-matching http request to a static HTML page ?* 'Sumit -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Percent encoded string in @QueryParam
Please forget about CXF proxies. It is down to UriBuilder.
http://docs.oracle.com/javaee/6/api/javax/ws/rs/core/UriBuilder.html states:
'Percent encoded values are also recognized where allowed and will not
be double encoded.'
Such values are allowed in the query representations. The UriBuilder
does not know what the intention of the CXF proxy user is, it may have
an encoded representation of a given character stored in the DB and
wants to pass it as is via the proxy.
Experiment with Jersey and if they do double encode (I doubt but...)
then we will continue, I think I've given you my best explanations so
far, lets continue later on when you have the Jersey test results ready.
Sergey
On 04/03/16 11:16, Volkov Sergey wrote:
i.e CXF proxy do not urlencode parameters? (Instead of it use
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded encoding which
is not reversible)
And the only reason of this encoding is to avoid reports about correctly
encoded parameters?
As I said, CXF proxy runtime uses JAX-RS UriBuilder.
And UriBuilder does not touch valid percent-encoded representations.
And this is how the RS client runtime will continue operating.
Sergey
On 04/03/16 10:45, Volkov Sergey wrote:
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
"100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER
VALID
STRING" - so actually CXF is over-smart now, when it skips all %xx
sequence
instead of encode % as %25
"100%VALID STRING" at the moment encodes by cxf as "100%25VALID STRING"
In the former case it it has no characters that can be used to form the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs
being
created.
If CXF will try to me over-smart then the next day someone will report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think this
behavior of @QueryParam may lead to tricky bugs in many places where
it
used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email]
<
http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID
STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which
just
looks like percent encoded.
If instead of
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both
examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to
the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but
when
i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82"
and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic
symbol
either as is or already percent encoded and in both case, as you
say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
I see a test like this:
@Test
public void testClonePctEncoded() throws Exception {
URI uri = new URI("http://bar;);
URI newUri = new UriBuilderImpl(uri)
.path("{a}").path("{b}")
.matrixParam("m", "m1 ", "m2+%20")
.queryParam("q", "q1 ",
"q2+q3%20").clone().buildFromEncoded("a+ ", "b%2B%20 ");
assertEquals("URI is not built correctly",
"http://bar/a+%20/b%2B%20%20;m=m1%20;m=m2+%20?q=q1+=q2%2Bq3%20;,
newUri.toString());
}
Note a 'q' query parameter has 2 values, one of them is
"q2+q3%20" so here we have a valid %-encoded representation.
The test asserts it is encoded as:
"q2%2Bq3%20"
i.e it stays untouched.
I nearly 100% sure this test was added as part of the TCK work, but do
not hesitate to test Jersey further if you'd like
Sergey
On 04/03/16 11:11, Sergey Beryozkin wrote:
Can you experiment with JAX-RS RI UriBuilder and double check that
passing to it a query parameter like "%DE" will not encode it ?
I'm pretty sure they won't. I have done a lot of work with UriBuilder to
make sure it passed JAX-RS 1.1 and the early 2.0 TCK that was available
to Apache. So I'd be surprised if I had to fix UriBuilder further given
that a lot of TCK tests are targeted at UriBuilder.
If Jersey does encode it then I'll review it and see what may need to be
done
Sergey
On 04/03/16 10:56, Sergey Beryozkin wrote:
As I said, CXF proxy runtime uses JAX-RS UriBuilder.
And UriBuilder does not touch valid percent-encoded representations.
And this is how the RS client runtime will continue operating.
Sergey
On 04/03/16 10:45, Volkov Sergey wrote:
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
"100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER
VALID
STRING" - so actually CXF is over-smart now, when it skips all %xx
sequence
instead of encode % as %25
"100%VALID STRING" at the moment encodes by cxf as "100%25VALID STRING"
In the former case it it has no characters that can be used to form the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs
being
created.
If CXF will try to me over-smart then the next day someone will report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think this
behavior of @QueryParam may lead to tricky bugs in many places
where it
used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email]
<http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID
STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which
just
looks like percent encoded.
If instead of
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both
examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to
the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but
when
i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82"
and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic
symbol
either as is or already percent encoded and in both case, as you
say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
Can you experiment with JAX-RS RI UriBuilder and double check that
passing to it a query parameter like "%DE" will not encode it ?
I'm pretty sure they won't. I have done a lot of work with UriBuilder to
make sure it passed JAX-RS 1.1 and the early 2.0 TCK that was available
to Apache. So I'd be surprised if I had to fix UriBuilder further given
that a lot of TCK tests are targeted at UriBuilder.
If Jersey does encode it then I'll review it and see what may need to be
done
Sergey
On 04/03/16 10:56, Sergey Beryozkin wrote:
As I said, CXF proxy runtime uses JAX-RS UriBuilder.
And UriBuilder does not touch valid percent-encoded representations.
And this is how the RS client runtime will continue operating.
Sergey
On 04/03/16 10:45, Volkov Sergey wrote:
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
"100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER VALID
STRING" - so actually CXF is over-smart now, when it skips all %xx
sequence
instead of encode % as %25
"100%VALID STRING" at the moment encodes by cxf as "100%25VALID STRING"
In the former case it it has no characters that can be used to form the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs being
created.
If CXF will try to me over-smart then the next day someone will report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think this
behavior of @QueryParam may lead to tricky bugs in many places where it
used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email]
<http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID
STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which
just
looks like percent encoded.
If instead of
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both
examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to
the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but
when
i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82"
and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic symbol
either as is or already percent encoded and in both case, as you say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
As I said, CXF proxy runtime uses JAX-RS UriBuilder.
And UriBuilder does not touch valid percent-encoded representations.
And this is how the RS client runtime will continue operating.
Sergey
On 04/03/16 10:45, Volkov Sergey wrote:
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
"100%25BETTER VALID STRING" is correct way to urlencode "100%BETTER VALID
STRING" - so actually CXF is over-smart now, when it skips all %xx sequence
instead of encode % as %25
"100%VALID STRING" at the moment encodes by cxf as "100%25VALID STRING"
In the former case it it has no characters that can be used to form the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs being
created.
If CXF will try to me over-smart then the next day someone will report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think this
behavior of @QueryParam may lead to tricky bugs in many places where it
used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <[hidden email]
<http://cxf.547215.n5.nabble.com/user/SendEmail.jtp?type=node=5766616=0>>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID
STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which
just
looks like percent encoded.
If instead of
org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both
examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but
when
i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82" and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic symbol
either as is or already percent encoded and in both case, as you say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
In the former case it it has no characters that can be used to form the
hex %-encoded representations.
The bottom line is, you should not expect the runtime auto-encode '%'
given that this symbol plays a special role in having correct URIs being
created.
If CXF will try to me over-smart then the next day someone will report
that it started producing
"100%25BETTER VALID STRING"
instead of
"100%BETTER VALID STRING"
Sergey
On 04/03/16 07:43, Volkov Sergey wrote:
Looks like POST request and @FormParam could be used, but I think this
behavior of @QueryParam may lead to tricky bugs in many places where it used
2016-03-04 10:37 GMT+03:00 Volkov Sergey <sergey.volko...@gmail.com>:
Yes, I don't understand why it does not touch this sequences.
It's a bit strange because I can successfully transfer "100%VALID STRING"
but not "100%BETTER VALID STRING"
Both this string not percent encoded, but second contains %BE which just
looks like percent encoded.
If instead of org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded
cxf just urlencode parameter value when building query string both examples
would be transferred correctly.
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but when
i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82" and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic symbol
either as is or already percent encoded and in both case, as you say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
If you are saying that
when you pass something like "%D1%82", or to make it clearer,
"%D1" and expect CXF do "%25D1", then no, it won't work.
UriBuilder does not touch percent-encoded sequences %XX.
May be try building the query with UriBuilder before passing it to the
proxy, something like
UriBuilder.fromUri("{a}%D21").build("%").toString()
Though not sure it will be cheaper.
Sergey
On 03/03/16 21:27, Sergey Beryozkin wrote:
From the original post in this thread:
When I pass string "т" (Cyrillic symbol) to generated client in
encodes
to
"%D1%82" in url and decodes as "т" on server what is fine, but when i
pass
"%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82" and
decodes
also to "т".
I honestly don't get what the problem is. You pass a Cyrillic symbol
either as is or already percent encoded and in both case, as you say
above, you get this Cyrillic symbol.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Percent encoded string in @QueryParam
I am a bit confused. If you pass "%25D1%2582" and have @Encoded then you will have "%25D1%2582" made available to the service. Sergey On 03/03/16 18:09, Volkov Sergey wrote: So currently there no way to pass both Cyrillic "т" as "т" (coded as "%D1%82" in url ) and "%D1%82" as "%D1%82" (coded as "%25D1%2582") from client to server using proxy? Should I create issue about that? 2016-03-03 20:52 GMT+03:00 Sergey Volkov <ser...@yandex-team.ru>: if you use Http centric API then indeed you can have @Encode on service interfaces but pass %D1%82 directly to 2.0 Client API or WebClient. If you have a proxy: it actually does not do @Encode processing either - the reason is that it is irrelevant for the proxy, it uses UriBuilder.buildFromEncoded() and must make sure the URI is correctly built. If it sees %-encoded values it does not touch them, otherwise it will encode them if needed. @Encoded is really for the server - as the processing requirements may require the value being either decoded or encoded. So adding @Encoded to the interface should do it... Sergey On 03/03/16 15:43, Volkov Sergey wrote: Hi! There was issue https://issues.apache.org/jira/browse/CXF-6122 that + was not working in QueryParam I have quite similar issue - but I can't pass percent encoded string in query params: When I pass string "т" (Cyrillic symbol) to generated client in encodes to "%D1%82" in url and decodes as "т" on server what is fine, but when i pass "%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82" and decodes also to "т". As far as I see this happens because org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded don't encode characters that looks like percent encoded. Is it bug? Is there any way to transfer %D1%82 trough @QueryParam? There is annotation javax.ws.rs.Encoded which as said in javadoc prevent @QueryParams from encoding, but with in that case I have to encode it manually in all places where generated method is used. -- Сергей Волков serg-v@ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Percent encoded string in @QueryParam
if you use Http centric API then indeed you can have @Encode on service interfaces but pass %D1%82 directly to 2.0 Client API or WebClient. If you have a proxy: it actually does not do @Encode processing either - the reason is that it is irrelevant for the proxy, it uses UriBuilder.buildFromEncoded() and must make sure the URI is correctly built. If it sees %-encoded values it does not touch them, otherwise it will encode them if needed. @Encoded is really for the server - as the processing requirements may require the value being either decoded or encoded. So adding @Encoded to the interface should do it... Sergey On 03/03/16 15:43, Volkov Sergey wrote: Hi! There was issue https://issues.apache.org/jira/browse/CXF-6122 that + was not working in QueryParam I have quite similar issue - but I can't pass percent encoded string in query params: When I pass string "т" (Cyrillic symbol) to generated client in encodes to "%D1%82" in url and decodes as "т" on server what is fine, but when i pass "%D1%82" rather then "%25D1%2582" in encodes also to "%D1%82" and decodes also to "т". As far as I see this happens because org.apache.cxf.jaxrs.utils.HttpUtils#encodePartiallyEncoded don't encode characters that looks like percent encoded. Is it bug? Is there any way to transfer %D1%82 trough @QueryParam? There is annotation javax.ws.rs.Encoded which as said in javadoc prevent @QueryParams from encoding, but with in that case I have to encode it manually in all places where generated method is used.
Re: Asking for comments about cxf-rs with oneway jms transport
Hi Tobi Using CXFRS client with JMS is indeed uncommon with the main reason being is that of course CXFRS is primarily meant to work with the HTTP targets. But with CXF being so flexible it ended up supporting non-HTTP targets too with some minor updates. So, ultimately, as far as the task of delivering a message over JMS is concerned, one has a choice to use JMS directly. Or, with CXF, use either SOAP (not unusual) or as in this case RS clients - where the client works with an Http centric API but ends up sending a message over JMS. If you use the RS proxies (as opposed to WebClient) then it will be less obvious what transport is being used, and can be a reasonable compromise as far as the 'mis-use' of the API is concerned Thanks, Sergey On 01/03/16 13:47, Tobias Schöneberg wrote: Hi, this morning I created this issue (which is already solved, thx again): https://issues.apache.org/jira/browse/CXF-6810. Note that in a comment, Sergey Beryozkin writes "So now we have WebClient and proxies being able to use HTTP Api to send the messages over JMS which is unusual ;-)" Because I'm new to cxf, I would like to ask the community whether they think I'm on the right track with my approach. My usage scenario is that I have a "frontend" web-UI server running in the cloud, that needs to exchange data with the backend ERP system. The internet link between those two is known to be a bit shaky. Now, what i had in mind for the solution is to run one jms-broker on the web-UI server and one near the ERP system. Those two brokers shall forward messages between each other and communicate via ssl. Both the web-UI and the ERP-system shall provide some services via cxf-rs. Also, they both share the same service-interfaces and serializable POJOs. The respective clients shall be able to make the invocation one-way, i.e. without having to wait for a result. They just shall obtain a client-proxy and make their call. The call shall directly return with a (http-)result of 202, as described here in the docs: http://cxf.apache.org/docs/jax-rs-advanced-features.html#JAX-RSAdvancedFeatures-Onewayinvocations. If the internet link is currently down, that's not the client's problem, but the two brokers need to retry and reestablish their connection. I think that if i used the http-transport, the client would have to check the if the request worked and explcityl retry if not. Note that if the server-endpoint of the other party can't process the message for whatever reason, it's also not the client's problem, but the server-endpoint needs to store the unprocessable message and notify someone about the problem. I think that my scenario is not that uncommon, but aparently using cxf-oneway-jms is uncommon. So I wonder, is this a reasonable approach? Does cxf provide better tools for this? Or maybe cxf is generally used for scenarios different than this one.? Best regards Tobi -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Registering a message body reader
Hi Benson WebClient.create(address, Collections.singletonList(new MyProvider())) you can also inject WebClient as jaxrs:client (set serviceClass="org.apache.cxf.jaxrs.client.WebClient") - in this case you can have jaxrs:client/@basePackages set to some package starting from which Spring will attempt to auto-discover it - though no way to do it with Blueprint just yet Cheers, Sergey On 26/02/16 20:03, Benson Margulies wrote: We have our very own, registered, content type: model/vnd.rosette.annotated-data-model How do I teach the CXF WebClient to call some class of ours when confronted with it? (It's really just json in disguise.) -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: JAX-RS CookieHeaderProvider with httpOnly cookie doesn't work
Hi Yes, you send a NewCookie but attempt to read it later on as Cookie. You should use NewCookie.valueOf(newCookieValue) to recreate it. Can you please use one of Cookie constructors to convert NewCookie to Cookie ? Or simply take the relevant value from this new cookie and send the name and value only which is what is expected really by the Cookie receiver. Sergey On 24/02/16 17:58, cc75005++ wrote: Sorry, now I understand your question.. please find,I think, the bad code that I use to create a new webclient to make the request : Thanks for your time -- View this message in context: http://cxf.547215.n5.nabble.com/JAX-RS-CookieHeaderProvider-with-httpOnly-cookie-doesn-t-work-tp5766083p5766290.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: No support for FailoverFeature in CXF Rest Client implementation (version 3.1.0)
Hi Please create a CXF JIRA issue, not sure how to qualify it, 'Task' I guess. As I said though it is unlikely to be addressed in the short term unless one of colleagues or yourself will look at it. Cheers, Sergey On 24/02/16 16:36, ranadeep.sha...@gmail.com wrote: Hi Sergey, Thanks for the reply. Is there any place where I can raise a request for this feature implementation ? That way, I believe all CXF consumers including myself can keep a track of it. Regards, Ranadeep. -- View this message in context: http://cxf.547215.n5.nabble.com/No-support-for-FailoverFeature-in-CXF-Rest-Client-implementation-version-3-1-0-tp5765813p5766277.html Sent from the cxf-user mailing list archive at Nabble.com.
Re: Proxy-based API genaration (two or more interaces)
Hi On 24/02/16 13:42, Vjacheslav V. Borisov wrote: Hi! Currently it is possible to write server side resource implementing two interfaces at the same time Eg,. public class SomeResource implements OneResourceInterface, TwoResourceInterface does it possible to generate Proxy-based client based on more that one interface? http://cxf.apache.org/docs/jax-rs-client-api.html#JAX-RSClientAPI-Proxy-basedAPI Currently I am see only possibility to generate two ProxyBased client interfaces an combine them in client side? in class like public class SomeResourceClient implements OneResourceInterface, TwoResourceInterface Yes, this is how it can be done, please make sure cglib-nodep dependency is on the classpath Cheers, Sergey -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: No support for FailoverFeature in CXF Rest Client implementation (version 3.1.0)
Hi Ranadeep,
On 24/02/16 01:35, ranadeep.sha...@gmail.com wrote:
Hi Sergey,
Thanks for the response. We have a requirement where we are planning to
preserve the usage of standard API going forward so that the functionality
is not impacted, no matter what the underlying implementation is.
I have some follow-up questions, below.
1) Is there any roadmap of releasing a future version of CXF Clustering
with full support for JAX-RS 2.0 API including the Failover retry features.
If so, when can we expect that to happen?
No specific work is being planned, but I suppose that can be done
eventually.
2) Is the CXF implementation of Failover Retry (for both SOAP client and
REST client) thread-safe?
I haven't seen any thread-safety related issues reported so far
Cheers, Sergey
Regards,
Ranadeep.
Regards,
Ranadeep.
On Wed, Feb 10, 2016 at 5:42 PM, Sergey Beryozkin [via CXF] <
ml-node+s547215n5765816...@n5.nabble.com> wrote:
JAX-RS 2.0 API does not work with a CXF specific failover feature,
definitely not something I've had a chance to look into.
We'd need to create a JAX-RS 2.0 Feature wrapper but what is the point
of such a JAX-RS Feature can not be portable.
Consider using WebClient with FailoverFeature.
Sergey
On 10/02/16 21:57, [hidden email]
<http:///user/SendEmail.jtp?type=node=5765816=0> wrote:
I had tried some heavy exploration on enabling Failover feature, as per
documentation provided by Apache CXF. While the one for SOAP client
works
like a charm, the other one for REST client is not so with the way I am
trying to instantiate the client. My requirement of instantiating the
REST
client is via JAXRS 2.0.1 API, as shown in following sample code.NOTE :
The
following endpoint URL is a valid one, except that the port number has
been
switched from 8080 (working one) to 8081 (non-working one), since I need
to
simulate the invocation failure retry behavior.Client client =
ClientBuilder.newClient(); WebTarget target =
client.target("http://localhost:8081/my-app/ws/ers/customers;);
Invocation.Builder builder =
target.request(MediaType.APPLICATION_JSON_TYPE); Response response =
builder.get();I have found that the Failover feature is not working when
the
above last line of code is invoked.Now, here comes the interesting part.
I
have found only 2 classes in cxf-rt-rs-client-3.1.0-sources.jar where
the
initialization of Failover feature in a
org.apache.cxf.jaxrs.client.JAXRSClientFactoryBean instance happens. And
my
requirement is to have one of these methods (or, a similar one where
bean.setFeatures(features) is invoked ) being called.Method in
org.apache.cxf.jaxrs.client.JAXRSClientFactory class.public static T
create(String baseAddress,Class cls, List
providers, List features,
String configLocation) {JAXRSClientFactoryBean bean =
getBean(baseAddress, cls, configLocation);
bean.setProviders(providers);
bean.setFeatures(features);return bean.create(cls);}Method in
org.apache.cxf.jaxrs.client.WebClient class.public static WebClient
create(String baseAddress,List providers,
List features, String
configLocation) {JAXRSClientFactoryBean bean = getBean(baseAddress,
configLocation);bean.setProviders(providers);
bean.setFeatures(features);return bean.createWebClient();}However,
when
my custom client code above is executed, neither of the above 2 methods
have
been invoked. Instead, the following method in
org.apache.cxf.jaxrs.client.spec.ClientImpl is invoked.private void
initTargetClientIfNeeded() {URI uri = this.uriBuilder.build(new
Object[0]);if(this.targetClient == null) {
JAXRSClientFactoryBean
bean = new JAXRSClientFactoryBean();
bean.setAddress(uri.toString());
this.targetClient = bean.createWebClient();
ClientImpl.this.baseClients.add(this.targetClient);} else
if(!this.targetClient.getCurrentURI().equals(uri)) {
this.targetClient.to(uri.toString(), false);} }Does this imply that
Failover Feature is NOT supported for CXF version 3.1.0 ?If so, which
upcoming version of CXF will support my above requirement ?
--
View this message in context:
http://cxf.547215.n5.nabble.com/No-support-for-FailoverFeature-in-CXF-Rest-Client-implementation-version-3-1-0-tp5765813.html
Sent from the cxf-user mailing list archive at Nabble.com.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
--
If you reply to this email, your message will be added to the discussion
below:
http://cxf.547215.n5.nabble.com/No-support-for-FailoverFeature-in-CXF-Rest-Client-implementation-version-3-1-0-tp5765813p5765816.html
To unsubscribe from No support for FailoverFeature in CXF Rest Client
implementation (version 3.1.0), click here
<http://cxf.547215.n5.nabble.com/template/NamlServlet.jtp?macro=unsubscribe_by_code=5765813=cmFuYWRlZXAuc2hhcm1hQGdtYWlsLmNvbXw1NzY1ODEzfC0xOTAyNzcwMDYz
Re: NullPointerException in OAuthRequestFilter
Hi Thanks for being so appreciative, but I'm only trying to provide whatever info I can, we work with all the users this way :-) I honestly do not know where are you getting these dependencies from. And if you wish to run your application in Glassfish you do not have to start going the embedded Jetty route, simply create a war, and in web.xml - declare CXFNonSpringJaxrsServlet and register your custom JAX-RS Application with it, something like this: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/resources/jaxrs_non_spring/WEB-INF/web.xml#L122 HTH, Sergey On 24/02/16 01:04, LaVloZ . wrote: Hii, Thank you again for your time, i tried to use JAXRSServerFactoryBean to create a new JAX-RS server so i added it Jetty transport jar to my maven, but CXF seems to need Google Guava and Glassfish HK2, so i got some exceptions in the first time until i added manualy Guava and HK2 in maven, why CXF doesn't declare these two dependencies althought it need them? Thank you again and again, very nice from you for the support :) 2016-02-23 19:13 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi, in most cases you'd like to configure it somehow, so may be try to create JAX-RS Applications where it is all configured in the code and register these applications with CXFNonSpringJaxrsServlet Sergey On 23/02/16 18:08, LaVloZ . wrote: Hi, Thank you for your time :), i have a little question, i see that the doc uses Spring, can we use CXF and OAuth implementation totaly without Spring?? :) Thanks 2016-02-23 10:33 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi, well, Jersey is of course a quality JAX-RS RI, but indeed, given its close integration with Glassfish, those users who may want to run CXF JAX-RS with Glassfish will need to experiment and try to disable Jersey somehow :-) Sergey On 22/02/16 22:04, LaVloZ . wrote: I didn't recieve an anwser mail for my original question so i resent the same question, then i saw your answer in my SO :) And as you said, Jersey is the problem, you have saved us alots of time. Thank you so much :) 2016-02-22 18:35 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi I replied to your original question a week or so ago, can you please check it ? Jersey is loaded is there so I think you may need to disable it somehow Thanks, Sergey On 22/02/16 17:24, LaVloZ . wrote: Hello all i have more than one week tryign to figure out the problem, but i didn't find it, googling, stackoverflow ..., but nothing. The problem is that the web services work fine but when i add the jar for OAuth support using Maven i get NullPointerException raised inside OAuthRequestFilter : this is the link of the line on GitHub : https://github.com/apache/cxf/blob/7d1890510a85d4fd7e70faebc56d6685f103621d/rt/rs/security/oauth-parent/oauth2/src/main/java/org/apache/cxf/rs/security/oauth2/filters/OAuthRequestFilter.java#L82 That JAXRSUtils.getCurrentMessage() returns null reference so that null cuz a NullPointerException latter in code. i'm using CXF 3.1.5 and tried it on Glassfish 4.1.1 and TomEE 7.0.63 on Oracle JDK 1.8.0_66 Thank you all. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: JAX-RS CookieHeaderProvider with httpOnly cookie doesn't work
Hi I'm sorry I still do not understand. Cookie header (as opposed to Set-Cookie) is not supposed to include HttpOnly, this is why I'm curious, how does this Cookie header ends up with HttpOnly ? Thanks, Sergey On 23/02/16 20:46, cc75005++ wrote: Hi Sergey, Thank you for your response. In fact the httpOnly flag is done by my authentification rest service that use NewCookie object. So the question should be "how can I have access to a NewCookie object in my ContainerRequestFilter ? or maybe I've to use another class to filter my request ? regards -- View this message in context: http://cxf.547215.n5.nabble.com/JAX-RS-CookieHeaderProvider-with-httpOnly-cookie-doesn-t-work-tp5766083p5766231.html Sent from the cxf-user mailing list archive at Nabble.com.
Re: NullPointerException in OAuthRequestFilter
Hi, in most cases you'd like to configure it somehow, so may be try to create JAX-RS Applications where it is all configured in the code and register these applications with CXFNonSpringJaxrsServlet Sergey On 23/02/16 18:08, LaVloZ . wrote: Hi, Thank you for your time :), i have a little question, i see that the doc uses Spring, can we use CXF and OAuth implementation totaly without Spring?? :) Thanks 2016-02-23 10:33 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi, well, Jersey is of course a quality JAX-RS RI, but indeed, given its close integration with Glassfish, those users who may want to run CXF JAX-RS with Glassfish will need to experiment and try to disable Jersey somehow :-) Sergey On 22/02/16 22:04, LaVloZ . wrote: I didn't recieve an anwser mail for my original question so i resent the same question, then i saw your answer in my SO :) And as you said, Jersey is the problem, you have saved us alots of time. Thank you so much :) 2016-02-22 18:35 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi I replied to your original question a week or so ago, can you please check it ? Jersey is loaded is there so I think you may need to disable it somehow Thanks, Sergey On 22/02/16 17:24, LaVloZ . wrote: Hello all i have more than one week tryign to figure out the problem, but i didn't find it, googling, stackoverflow ..., but nothing. The problem is that the web services work fine but when i add the jar for OAuth support using Maven i get NullPointerException raised inside OAuthRequestFilter : this is the link of the line on GitHub : https://github.com/apache/cxf/blob/7d1890510a85d4fd7e70faebc56d6685f103621d/rt/rs/security/oauth-parent/oauth2/src/main/java/org/apache/cxf/rs/security/oauth2/filters/OAuthRequestFilter.java#L82 That JAXRSUtils.getCurrentMessage() returns null reference so that null cuz a NullPointerException latter in code. i'm using CXF 3.1.5 and tried it on Glassfish 4.1.1 and TomEE 7.0.63 on Oracle JDK 1.8.0_66 Thank you all. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Adding NCSARequestLog via JAXRSServerFactoryBean
Hi
This is a fragment from a test:
ClassPathXmlApplicationContext ctx =
new ClassPathXmlApplicationContext(new String[]
{"/org/apache/cxf/jaxrs/spring/servers2.xml"});
Object serverBean = ctx.getBean("server1");
assertNotNull(serverBean);
JAXRSServerFactoryBean factoryBean =
(JAXRSServerFactoryBean)serverBean;
where 'server1' is a jaxrs:server id
and then it is only
factoryBean.create().
HTH, Sergey
On 23/02/16 13:00, matthew.pres...@thomsonreuters.com wrote:
Sorry if this question is trivial, but if I use that configuration, where do I
put the xml and how do I need to change my code to launch the server?
Thanks,
Matt
-Original Message-
From: Freeman Fang [mailto:freeman.f...@gmail.com]
Sent: 23 February 2016 12:34
To: users@cxf.apache.org
Subject: Re: Adding NCSARequestLog via JAXRSServerFactoryBean
Yeah, the configuration way works, something like
On Feb 23, 2016, at 8:13 PM, Sergey Beryozkin <sberyoz...@gmail.com> wrote:
Hi
I guess you may want to write a Jetty specific code directly, may be as in this
demo server:
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_apache
_cxf_blob_master_distribution_src_main_release_samples_jax-5Frs_search
_src_main_java_demo_jaxrs_search_server_Server.java=CwIFaQ=4ZIZThy
kDLcoWk-GVjSLm9hvvvzvGv0FLoWSRuCSs5Q=ykXcTN60h3-cXsX5t7kv8zVQwTSxpeA
xKWimPUn10ALTg1Uk6Ao2R1mxc0lzB5Ia=tCDKXnNyVwy8AwK0XtEPUZjGzz6XSLoDWQ
YbA4oG-NM=_0kGFcXUU-GIcEYu72VvjMaJimjoof7nCUjyHemzUUA=
Not sure how to do it with the code below, may be
Server server = sf.create()
and then get a Bus from it, and then
DestinationRegistry reg = bus.getExtension(DestinationRegistry.class)
and then get JettyHttpDestination and then somehow add a new handler to it...
Finally, may be you can create an httpj element and jaxrs:endpoint in
Spring/Blueprint and activate the context:
https://urldefense.proofpoint.com/v2/url?u=http-3A__cxf.apache.org_doc
s_jetty-2Dconfiguration.html=CwIFaQ=4ZIZThykDLcoWk-GVjSLm9hvvvzvGv
0FLoWSRuCSs5Q=ykXcTN60h3-cXsX5t7kv8zVQwTSxpeAxKWimPUn10ALTg1Uk6Ao2R1
mxc0lzB5Ia=tCDKXnNyVwy8AwK0XtEPUZjGzz6XSLoDWQYbA4oG-NM=9K8w1aY1JRY
xpRo7sg1V3BcTWkVKZyB7dLLFMi1MZCw=
HTH, Sergey
On 23/02/16 11:56, matthew.pres...@thomsonreuters.com wrote:
Hi,
I'm using cxf to start a simple JAX-RS server using the following code (from
the documentation):
JAXRSServerFactoryBean sf = new JAXRSServerFactoryBean();
sf.setResourceClasses(clazz);
sf.setResourceProvider(clazz, new SingletonResourceProvider(this));
sf.setAddress("https://urldefense.proofpoint.com/v2/url?u=http-3A__lo
calhost-3A5000=CwIFaQ=4ZIZThykDLcoWk-GVjSLm9hvvvzvGv0FLoWSRuCSs5Q
=ykXcTN60h3-cXsX5t7kv8zVQwTSxpeAxKWimPUn10ALTg1Uk6Ao2R1mxc0lzB5Ia
=tCDKXnNyVwy8AwK0XtEPUZjGzz6XSLoDWQYbA4oG-NM=MN2PXIbv9RImP4of7wb92Q
3p7Mr6N-Zmk7nDaMryD8k= ");
sf.create();
How can I modify this code to add an NCSARequestLog to the embedded jetty
instance? Is this possible?
Thanks,
Matt
This e-mail is for the sole use of the intended recipient and
contains information that may be privileged and/or confidential. If
you are not an intended recipient, please notify the sender by return e-mail and
delete this e-mail and any attachments. Certain required legal entity disclosures can
be accessed on our website.<http://site.thomsonreuters.com/site/disclosures/>
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Adding NCSARequestLog via JAXRSServerFactoryBean
Hi
I guess you may want to write a Jetty specific code directly, may be as
in this demo server:
https://github.com/apache/cxf/blob/master/distribution/src/main/release/samples/jax_rs/search/src/main/java/demo/jaxrs/search/server/Server.java
Not sure how to do it with the code below,
may be
Server server = sf.create()
and then get a Bus from it, and then
DestinationRegistry reg = bus.getExtension(DestinationRegistry.class)
and then get JettyHttpDestination and then somehow add a new handler to
it...
Finally, may be you can create an httpj element and jaxrs:endpoint in
Spring/Blueprint and activate the context:
http://cxf.apache.org/docs/jetty-configuration.html
HTH, Sergey
On 23/02/16 11:56, matthew.pres...@thomsonreuters.com wrote:
Hi,
I'm using cxf to start a simple JAX-RS server using the following code (from
the documentation):
JAXRSServerFactoryBean sf = new JAXRSServerFactoryBean();
sf.setResourceClasses(clazz);
sf.setResourceProvider(clazz, new SingletonResourceProvider(this));
sf.setAddress("http://localhost:5000;);
sf.create();
How can I modify this code to add an NCSARequestLog to the embedded jetty
instance? Is this possible?
Thanks,
Matt
This e-mail is for the sole use of the intended recipient and contains information
that may be privileged and/or confidential. If you are not an intended recipient,
please notify the sender by return e-mail and delete this e-mail and any attachments.
Certain required legal entity disclosures can be accessed on our
website.
Re: Support both Bearer and JWT tokens
On 23/02/16 10:27, Sergey Beryozkin wrote: Hi On 23/02/16 07:24, Rajan1311 wrote: Hi Sergey, I think I have understood what you mean, so the JWT has nothing to do with the OAuth flows. I did not quite imply that JWT was not related to OAuth flows, rather I said the JWTAuthenticationFilter you were thinking of using was not related. Or to be more precise, it can be used as a filter for authenticating the OAuth2 clients but only when these clients use JWT for a pure authentication when accessing OAuth2 AccessTokenService, example, when using a client credentials or authorization code grant... Sergey In my case though, when a client server needs access to my API, he does so with the access token obtained via the client credentials grant. Each token (default bearer type) has registered scopes which govern the parts of the API the client can access. According to the documentation, its the OAuthRequestFilter and OAuthScopesFilter that make this restriction happen. I am using the @Scopes annotation my controllers. Sure, you might also want to consider using OAuthRequestFilter only with the requiredScopes property set, unless you'd like to enforce the scopes declaratively on per-method level Now, the Client can also make requests with a JWT(issued by the client's server). The client signs the JWT with their client secret and sends me (API server) the JWT along with the client id (so that I can verify it with the corresponding secret). The JWT claims has a list of scopes along with some other identification info. This is where I need some sort of convergence between the 2 flows. Once I have the scopes, I want to process them in such a way that I use the scopes in the access token (via client credentials). So, for this to happen, I think I would need some sort of modification in my OAuthRequestFilter, which would also accept JWT tokens, validate them and extract the scopes. Then, it should work seamlessly work with the OAuthScopesFilter. I hope you understood what I mean. I am very new to this, so would be great if you could just point me in the right direction. I am thinking of implementing a custom Filter to replace the OAuthRequestFilter, but I am not too sure on how to proceed. Hmm, well, the client gets the access token from OAuth2 Access Token Service, and then the client uses this token to access some users' API server. But what you are describing is that the client creates a JWT token itself (may be it gets this token from STS such as CXF STS that can issue JWT assertions), sets a client id on it and then wants to use it as an access token ? if so then I don't think it is possible, all the client has is a set of JWT claims representing the fact of the successful client authentication with STS/etc, and by adding a client id to such claims does not turn them into an access token, but the client can use them to claim a new access token and then use it. Cheers, Sergey Thanks and Regards, Rajan -- View this message in context: http://cxf.547215.n5.nabble.com/Support-both-Bearer-and-JWT-tokens-tp5766050p5766172.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Support both Bearer and JWT tokens
Hi On 23/02/16 07:24, Rajan1311 wrote: Hi Sergey, I think I have understood what you mean, so the JWT has nothing to do with the OAuth flows. I did not quite imply that JWT was not related to OAuth flows, rather I said the JWTAuthenticationFilter you were thinking of using was not related. In my case though, when a client server needs access to my API, he does so with the access token obtained via the client credentials grant. Each token (default bearer type) has registered scopes which govern the parts of the API the client can access. According to the documentation, its the OAuthRequestFilter and OAuthScopesFilter that make this restriction happen. I am using the @Scopes annotation my controllers. Sure, you might also want to consider using OAuthRequestFilter only with the requiredScopes property set, unless you'd like to enforce the scopes declaratively on per-method level Now, the Client can also make requests with a JWT(issued by the client's server). The client signs the JWT with their client secret and sends me (API server) the JWT along with the client id (so that I can verify it with the corresponding secret). The JWT claims has a list of scopes along with some other identification info. This is where I need some sort of convergence between the 2 flows. Once I have the scopes, I want to process them in such a way that I use the scopes in the access token (via client credentials). So, for this to happen, I think I would need some sort of modification in my OAuthRequestFilter, which would also accept JWT tokens, validate them and extract the scopes. Then, it should work seamlessly work with the OAuthScopesFilter. I hope you understood what I mean. I am very new to this, so would be great if you could just point me in the right direction. I am thinking of implementing a custom Filter to replace the OAuthRequestFilter, but I am not too sure on how to proceed. Hmm, well, the client gets the access token from OAuth2 Access Token Service, and then the client uses this token to access some users' API server. But what you are describing is that the client creates a JWT token itself (may be it gets this token from STS such as CXF STS that can issue JWT assertions), sets a client id on it and then wants to use it as an access token ? if so then I don't think it is possible, all the client has is a set of JWT claims representing the fact of the successful client authentication with STS/etc, and by adding a client id to such claims does not turn them into an access token, but the client can use them to claim a new access token and then use it. Cheers, Sergey Thanks and Regards, Rajan -- View this message in context: http://cxf.547215.n5.nabble.com/Support-both-Bearer-and-JWT-tokens-tp5766050p5766172.html Sent from the cxf-user mailing list archive at Nabble.com.
Re: NullPointerException in OAuthRequestFilter
Hi, well, Jersey is of course a quality JAX-RS RI, but indeed, given its close integration with Glassfish, those users who may want to run CXF JAX-RS with Glassfish will need to experiment and try to disable Jersey somehow :-) Sergey On 22/02/16 22:04, LaVloZ . wrote: I didn't recieve an anwser mail for my original question so i resent the same question, then i saw your answer in my SO :) And as you said, Jersey is the problem, you have saved us alots of time. Thank you so much :) 2016-02-22 18:35 GMT+01:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi I replied to your original question a week or so ago, can you please check it ? Jersey is loaded is there so I think you may need to disable it somehow Thanks, Sergey On 22/02/16 17:24, LaVloZ . wrote: Hello all i have more than one week tryign to figure out the problem, but i didn't find it, googling, stackoverflow ..., but nothing. The problem is that the web services work fine but when i add the jar for OAuth support using Maven i get NullPointerException raised inside OAuthRequestFilter : this is the link of the line on GitHub : https://github.com/apache/cxf/blob/7d1890510a85d4fd7e70faebc56d6685f103621d/rt/rs/security/oauth-parent/oauth2/src/main/java/org/apache/cxf/rs/security/oauth2/filters/OAuthRequestFilter.java#L82 That JAXRSUtils.getCurrentMessage() returns null reference so that null cuz a NullPointerException latter in code. i'm using CXF 3.1.5 and tried it on Glassfish 4.1.1 and TomEE 7.0.63 on Oracle JDK 1.8.0_66 Thank you all. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: conflicts with the registered path
Not really - can you debug JettyHTTPServerEngine and check if this property is actually recognized at runtime. FYI, I'm only assuming that update I was referring to is relevant, might not be. So start with confirming this property is indeed seen as being set, in JettyHTTPServerEngine.shouldCheckUrl(). Cheers, Sergey On 22/02/16 18:09, ravi473 wrote: Thanks for reply . I have added the system property as you suggest but there is no difference .Can you please suggest me if have any other way to get out of the problem. Cheers, ravi -- View this message in context: http://cxf.547215.n5.nabble.com/conflicts-with-the-registered-path-tp5766073p5766158.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: NullPointerException in OAuthRequestFilter
Hi I replied to your original question a week or so ago, can you please check it ? Jersey is loaded is there so I think you may need to disable it somehow Thanks, Sergey On 22/02/16 17:24, LaVloZ . wrote: Hello all i have more than one week tryign to figure out the problem, but i didn't find it, googling, stackoverflow ..., but nothing. The problem is that the web services work fine but when i add the jar for OAuth support using Maven i get NullPointerException raised inside OAuthRequestFilter : this is the link of the line on GitHub : https://github.com/apache/cxf/blob/7d1890510a85d4fd7e70faebc56d6685f103621d/rt/rs/security/oauth-parent/oauth2/src/main/java/org/apache/cxf/rs/security/oauth2/filters/OAuthRequestFilter.java#L82 That JAXRSUtils.getCurrentMessage() returns null reference so that null cuz a NullPointerException latter in code. i'm using CXF 3.1.5 and tried it on Glassfish 4.1.1 and TomEE 7.0.63 on Oracle JDK 1.8.0_66 Thank you all. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: conflicts with the registered path
Hi I recall there was some work added to avoid some side-effects related to running multiple CXF Jetty endpoint on the overlapping paths, but then it caused issues for some existing deployments (like yours) so a property "org.apache.cxf.transports.http_jetty.DontCheckUrl" was added. It can be set to 'true' as a system or bus property. Try it please Sergey On 19/02/16 21:46, ravi473 wrote: We are starting multiple web service endpoints with same base name using code below. For instance, http://localhost:9001/org/dep and http://localhost:9001/org/dep/manage . It used to work till 2.1.7. After we upgraded CXF to 3.1.4 and it is failing with below exception . Caused by: org.apache.cxf.interceptor.Fault: Could not add cxf jetty handler for url http://localhost:8585/rs/jde/jxml to Jetty server, as the path /rs/jde/jxml is still in use. at org.apache.cxf.transport.http_jetty.JettyHTTPServerEngine.checkRegistedContext(JettyHTTPServerEngine.java:317) at org.apache.cxf.transport.http_jetty.JettyHTTPServerEngine.addServant(JettyHTTPServerEngine.java:386) at org.apache.cxf.transport.http_jetty.JettyHTTPDestination.activate(JettyHTTPDestination.java:175) at org.apache.cxf.transport.AbstractObservable.setMessageObserver(AbstractObservable.java:53) at org.apache.cxf.binding.AbstractBindingFactory.addListener(AbstractBindingFactory.java:95) at org.apache.cxf.jaxrs.JAXRSBindingFactory.addListener(JAXRSBindingFactory.java:88) at org.apache.cxf.endpoint.ServerImpl.start(ServerImpl.java:123) at org.apache.cxf.jaxrs.JAXRSServerFactoryBean.create(JAXRSServerFactoryBean.java:206) -- View this message in context: http://cxf.547215.n5.nabble.com/conflicts-with-the-registered-path-tp5766073.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: JAX-RS CookieHeaderProvider with httpOnly cookie doesn't work
Hi
Can Cookie contain an attribute such as HttpOnly ? I'm not sure.
NewCookie has an isHttpOnly method - but not Cookie.
How is Set-Cookie processed on the client side ? By a browser or somehow
else ? I'm curious how HttpOnly is added to Cookie.
I guess if you have a Cookie with unsupported properties then the clean
workaround is to access Cookie header directly,
with
requestContext.getHeaders().getFirst("Cookie");
and then do a manual parsing
Thanks, Sergey
On 20/02/16 22:24, cc75005++ wrote:
Hello,
With Tomee 7.00-M1 (cxf 3.1.3) , If I send a request with an httponly cookie
on my rest service, I receive an IllegalArgumentException("Cookie is
malformed : " + c) when I call requestContext.getCookies().
Please find my container request filter implementation :
The CXF-6451 <https://issues.apache.org/jira/browse/CXF-6451> fix the bug
on the NewCookieHeaderProvider.java on 3.1.2 version but not for
CookieHeaderProvider class.
Any help would be much appreciated to find a clean workaround.
Regards
--
View this message in context:
http://cxf.547215.n5.nabble.com/JAX-RS-CookieHeaderProvider-with-httpOnly-cookie-doesn-t-work-tp5766083.html
Sent from the cxf-user mailing list archive at Nabble.com.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Support both Bearer and JWT tokens
Hi Thanks for experimenting with the JWT authentication filter, but it is not exactly tailored for the use in OAuth2 flows, let me clarify. JWT (claims) is one way to have a given access token represented *internally*. An access token can be a DB pointer to some custom table, or may be an encrypted set of JWT claims. It is still a Bearer token though. Of course, other types of tokens are possible, such as PoP tokens which have been standardized recently, in CXF we currently support Hawk schemes, but even in those schemes it can be JWT claims that are used internally. So, speaking of the access tokens, they are typically Bearer (or PoP/etc) access tokens, whether JWT claims are used to represent them or not is somewhat orthogonal. We have some code in place though to support JWT assertions which represent a given client's authentication credentials and can be used, but these JWT assertions are not the tokens, the end result there is that the OAuth2 server will still return a Bearer/PoP/etc token http://cxf.apache.org/docs/jaxrs-oauth2-assertions.html JwtAuthenticationFilter is meant to be used in regular client server communications, the JWT scheme there is not related to OAuth2 flows, for example, the client obtains a JWT assertion, and uses it to authenticate, and may be also link it to the JWS/JWE secured payload... Does it help ? Sergey On 19/02/16 04:24, Rajan1311 wrote: Hi, I am still pretty new at this, so might have not understood this correctly. I am trying to protect my API via ouath. I basically need 2 types of flow, one is a client credentials, where a client can dynamically register and then access their resources. For this, I have followed the documentation and have successfully implemented the same. I am using a bearer token for this. Next I would need a JWT which my clients would generate and then use them. I have created an SDK which is used by my clients to create JWTs. Based on the information within the JWT, the bearer of this token can get access to the clients resource. So, I need to support 2 types of tokens - Bearer and JWT tokens. I did a little bit of searching but was left confused. I tried using the org.apache.cxf.rs.security.jose.jaxrs.JwtAuthenticationFilter (registered it as a filter in the service I want to protect), which checks for the validity of the JWT token, but it would block requests made with the bearer token. I am not sure how to proceed from here. Should I create a custom token validator for my OAuth filter ? Thanks Rajan -- View this message in context: http://cxf.547215.n5.nabble.com/Support-both-Bearer-and-JWT-tokens-tp5766050.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: WebClient - Content-Type Override?
Hi
I guess the workaround is required given that 2.7.x is not being
released any longer
Cheers, Sergey
On 16/02/16 18:08, Andrew Clark wrote:
Hi Sergey
Unfortunately we are currently restricted to using 2.x, we are currently on
version 2.7.15.
Sent from my iPhone
On 16 Feb 2016, at 17:49, "Sergey Beryozkin" <sberyoz...@gmail.com> wrote:
Hi Christian
Starting with CXF 3.1.5 (or may be even 3.1.4) you should be able to set a
custom Content-Type and it will go on the wire
Cheers, Sergey
On 16/02/16 17:43, Christian wrote:
Hi Sergey,
We are facing a similar issue; we are trying to send a GET with a
Content-Type header, as the service called requires it.
Just wondering if the workaround with the interceptor is still required?
We are trying to set a particular Content-Type (application/json), but it
gets stripped out at some point before the request gets sent, and the call
fails. Is that cxf's doing?
Regards,
Christian
Sergey Beryozkin wrote
Ok, thanks for this info.
You've found the workaround which is good, another option is to
register a custom CXF out interceptor which would reset Content-Type
to the required value, it can be registered with
JAXRSClientFactoryBean:
JAXRSClientFactoryBean bean = new JAXRSClientFactoryBean();
bean.setAddress(address);
bean.getOutInterceptors().add(new CustomOutInterceptor());
bean.createWebClient();
I'll try fix it as well
Sergey
On Sun, Jul 31, 2011 at 10:17 PM, Jeff Wang
jeff.wang@
wrote:
We're trying to use Amazon's query string authentication. We can set
Content-Type to "", or heck, to any other string. It's just that the
server uses the content type as part of its signature calculation. So
when we generate the signature, we expected to set it with some content
type (tried it with either "" or an actual type, it doesn't matter.) If
it gets overridden, like what happens with webClient.get(), then the
signatures will be wrong and auth will get rejected. The actual header
has no meaning in a GET, but we do need to control what it is so that the
signatures match.
Jeff
-----Original Message-
From: Sergey Beryozkin [mailto:
sberyozkin@
]
Sent: Sun 7/31/2011 11:00 AM
To:
users@.apache
Subject: Re: WebClient - Content-Type Override?
Hi,
Does the server expect Content-Type being set even in case of GET
requests ?
Cheers, Sergey
On 28/07/11 22:45, Jeff Wang wrote:
Looking through the code, it seems that if the body is null (which has
to be the case considering it's a get, then the content type is
overwritten with wildcards.
Using client.invoke("GET",""); solved the problem for me.
Jeff
-Original Message-
From: Jeff Wang [mailto:
jeff.wang@
]
Sent: Wednesday, July 27, 2011 5:29 PM
To:
users@.apache
Subject: WebClient - Content-Type Override?
I'm trying to act as a proxy to a third party webservice, and need to
transform a REST request. One issue that I have is that the
Content-Type header seems to get overridden no matter what I do. The
code is actually very simple (context is an @Context MessageContext
variable):
WebClient client = WebClient.create(url)
.header("real-header-removed",
"auth-string-removed")
//.header("Content-Type",
context.getHttpHeaders().getMediaType().toString());
.type(context.getHttpHeaders().getMediaType());
Response resp = client.get();
return (InputStream)resp.getEntity();
Neither the .header nor the .type worked. The tcpmon output of the
request is:
Content-Type: */*
real-header-removed: auth-string-removed
Accept: application/xml
User-Agent: Apache CXF 2.3.5
Cache-Control: no-cache
Pragma: no-cache
How do I avoid the content Type override (or rather, why is it
happening?)
Btw, while looking for answers, I found my CXF server side answers
relatively easily either via google or the archive. But client
questions seems to be rather rare, and mostly on the SOAP calls. Are
most people using other packages?
thanks
Jeff
--
Sergey Beryozkin
http://sberyozkin.blogspot.com
Talend - http://www.talend.com
--
View this message in context:
http://cxf.547215.n5.nabble.com/WebClient-Content-Type-Override-tp4640790p5765944.html
Sent from the cxf-user mailing list archive at Nabble.com.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: WebClient - Content-Type Override?
Hi Christian
Starting with CXF 3.1.5 (or may be even 3.1.4) you should be able to set
a custom Content-Type and it will go on the wire
Cheers, Sergey
On 16/02/16 17:43, Christian wrote:
Hi Sergey,
We are facing a similar issue; we are trying to send a GET with a
Content-Type header, as the service called requires it.
Just wondering if the workaround with the interceptor is still required?
We are trying to set a particular Content-Type (application/json), but it
gets stripped out at some point before the request gets sent, and the call
fails. Is that cxf's doing?
Regards,
Christian
Sergey Beryozkin wrote
Ok, thanks for this info.
You've found the workaround which is good, another option is to
register a custom CXF out interceptor which would reset Content-Type
to the required value, it can be registered with
JAXRSClientFactoryBean:
JAXRSClientFactoryBean bean = new JAXRSClientFactoryBean();
bean.setAddress(address);
bean.getOutInterceptors().add(new CustomOutInterceptor());
bean.createWebClient();
I'll try fix it as well
Sergey
On Sun, Jul 31, 2011 at 10:17 PM, Jeff Wang
jeff.wang@
wrote:
We're trying to use Amazon's query string authentication. We can set
Content-Type to "", or heck, to any other string. It's just that the
server uses the content type as part of its signature calculation. So
when we generate the signature, we expected to set it with some content
type (tried it with either "" or an actual type, it doesn't matter.) If
it gets overridden, like what happens with webClient.get(), then the
signatures will be wrong and auth will get rejected. The actual header
has no meaning in a GET, but we do need to control what it is so that the
signatures match.
Jeff
-Original Message-
From: Sergey Beryozkin [mailto:
sberyozkin@
]
Sent: Sun 7/31/2011 11:00 AM
To:
users@.apache
Subject: Re: WebClient - Content-Type Override?
Hi,
Does the server expect Content-Type being set even in case of GET
requests ?
Cheers, Sergey
On 28/07/11 22:45, Jeff Wang wrote:
Looking through the code, it seems that if the body is null (which has
to be the case considering it's a get, then the content type is
overwritten with wildcards.
Using client.invoke("GET",""); solved the problem for me.
Jeff
-Original Message-
From: Jeff Wang [mailto:
jeff.wang@
]
Sent: Wednesday, July 27, 2011 5:29 PM
To:
users@.apache
Subject: WebClient - Content-Type Override?
I'm trying to act as a proxy to a third party webservice, and need to
transform a REST request. One issue that I have is that the
Content-Type header seems to get overridden no matter what I do. The
code is actually very simple (context is an @Context MessageContext
variable):
WebClient client = WebClient.create(url)
.header("real-header-removed",
"auth-string-removed")
//.header("Content-Type",
context.getHttpHeaders().getMediaType().toString());
.type(context.getHttpHeaders().getMediaType());
Response resp = client.get();
return (InputStream)resp.getEntity();
Neither the .header nor the .type worked. The tcpmon output of the
request is:
Content-Type: */*
real-header-removed: auth-string-removed
Accept: application/xml
User-Agent: Apache CXF 2.3.5
Cache-Control: no-cache
Pragma: no-cache
How do I avoid the content Type override (or rather, why is it
happening?)
Btw, while looking for answers, I found my CXF server side answers
relatively easily either via google or the archive. But client
questions seems to be rather rare, and mostly on the SOAP calls. Are
most people using other packages?
thanks
Jeff
--
Sergey Beryozkin
http://sberyozkin.blogspot.com
Talend - http://www.talend.com
--
View this message in context:
http://cxf.547215.n5.nabble.com/WebClient-Content-Type-Override-tp4640790p5765944.html
Sent from the cxf-user mailing list archive at Nabble.com.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: NullPointerException in OAuthRequestFilter
Hi Jersey is loaded and it is auto-scanning the providers. Are you using Jersey but only would like to use CXF OAuth2 ? if only CXF is meant to be used then you need to disable Jersey somehow, some info is here: http://cxf.apache.org/docs/jax-rs-deployment.html#JAX-RSDeployment-Glassfish but it may be incomplete I will reply to your question on Stackoverflow too once we understand where the problem is Cheers, Sergey On 16/02/16 11:48, LaVloZ . wrote: Hello all, how are you?? i've already posted this question on StackOverflow but it seems no one has respond me, so please visit this link to see my question :) http://stackoverflow.com/questions/35420727/cxf-nullpointerexception-in-oauthrequestfilter Thank you :)
Re: CXF ignoring custom headers
Hi, see https://issues.apache.org/jira/browse/CXF-6611 Sergey On 09/02/16 09:03, Zizou84 wrote: Hello I am using CXF 3.1.3 (jaxrs). My custom header is ignored org.apache.cxf.jaxrs.client.WebClient webClient = factoryBean.createWebClient(); // set connection properties org.apache.cxf.jaxrs.client.ClientConfiguration clientConfig = WebClient.getConfig(webClient); webClient.header(HttpHeaders.HOST, "mycustomHeader"); in the server side, the HttpHeaders.HOST header is the default one!! How to set this custom header -- View this message in context: http://cxf.547215.n5.nabble.com/CXF-ignoring-custom-headers-tp5765711.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Sub-Agent in logs telling the version of apache cxf
Hi I just tried, the header goes out when it is set on HttpConduit but it is named as 'BrowserType' which looks like a typo because it is documented to be User-Agent, and I can find no links to BrowserType HTTP header (does not exist), I'll fix it shortly Thanks, Sergey On 09/02/16 15:20, Khare, Aparna wrote: Thanks a lot Sergey. I tried both the options Setting browser type did not work but setting the Sub-Agent in header worked Thanks for the help, Aparna -Original Message- From: Sergey Beryozkin [mailto:sberyoz...@gmail.com] Sent: Sunday, February 7, 2016 9:46 PM To: users@cxf.apache.org Subject: Re: Sub-Agent in logs telling the version of apache cxf Hi You can customize it, set a User-Agent header when the clients are created, ex, in jaxrs-client:client/jaxrs-client:headers, or in JAXRSClientFactoryBean, or directly on WebClient Another option is to set it up on HttpConduit, the property is called 'BrowserType' there. HTH, Sergey On 06/02/16 09:32, Khare, Aparna wrote: Dear Colleagues, We are using apache CXF 3.0.7 and use Webclient to make an outbound HTTP call. We connect to the external endpoint from our systems .The external system would log from which agent the call is being intercepted. When a cxf client makes a call the logs are writtern with CXF agent and the version which is a security concern.Can you tell me how to override the agent information. Best Regards, Aparna -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Sub-Agent in logs telling the version of apache cxf
Hi You can customize it, set a User-Agent header when the clients are created, ex, in jaxrs-client:client/jaxrs-client:headers, or in JAXRSClientFactoryBean, or directly on WebClient Another option is to set it up on HttpConduit, the property is called 'BrowserType' there. HTH, Sergey On 06/02/16 09:32, Khare, Aparna wrote: Dear Colleagues, We are using apache CXF 3.0.7 and use Webclient to make an outbound HTTP call. We connect to the external endpoint from our systems .The external system would log from which agent the call is being intercepted. When a cxf client makes a call the logs are writtern with CXF agent and the version which is a security concern.Can you tell me how to override the agent information. Best Regards, Aparna -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Limit connection/threads for CXF endpoint?
Hi. Dan did a nice demo, see a 'throttling' demo in the distribution, it uses a Codahale library to do the advanced rate invocation checks, right now it can delay the overflowing responses using the suspended invocations mechanism, from 3.1.6 one would be able to abort such requests with whatever HTTP error is preferred, 429/503, etc (Dan, please review the minor changes I just did...) Please experiment with the demo, Sergey On 01/02/16 11:12, Sergey Maslov wrote: Hi! I have web application that provides several soap service endpoints and consumes several services as clients. One of endpoints receives too many requests so it consumes much cpu resources and the rest of application services freeze. We also get "Executor queue is full, run the oneway invocation task in caller thread" message, but increasing executor queue doesn't help as the reason is heavy load. Now we think of application redesign and moving to microservices, but it is not easy deal. But we want to have some fast solution for right now. For example, is it possible to limit threads or connections for CXF endpoint calls? The desired behavior that if some external service calls our endpoint and the connection size extends the limit, the caller receives connection timeout or some other connection status. And our application remains stable. Thank you in advance. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Accessing init-param in spring with servlet transport
Hi Paul
Thanks for the explanation, I guess it does not work because CXFServlet
parameters are not visible to a web.xml-'global' application context
which can be shared between multiple servlets, so injecting a servlet
specific parameters into this context would be wrong.
Try using a CXFServlet config-location init param instead of a context
contextConfigLocation to refer to the application context, perhaps it
will make the init parameters visible
Cheers, Sergey
Try setting the application context
On 26/01/16 10:34, Paul Millar wrote:
Hi Sergey,
On 19/01/16 12:58, Sergey Beryozkin wrote:
> Sorry for a delay.
Sorry for my delay!
I'm not sure I understand, so you'd like to have some context-params and
CXFServlet specific init-params set in a given web.xml and make them
available to the application context ?
Can you please show some configuration fragments which will help to
understand what exactly are you trying to achieve ?
A little bit of background.
I've "inherited" a rather basic servlet that I'm trying to bring into a
reasonable shape. It provides support for a network protocol, using a
directory to store some state.
Currently, the directory it uses is hardcoded as '/data'. I would like
to expose this as a configuration option for two reasons:
a. Using '/data' might not be convenient for all users.
b. I'm adding integration testing as part of the maven build process
and I want to run the servlet (from within maven) but configuring
it to use a path within the maven-project's 'target' directory.
I believe init-param is the implementation-agnostic way of configuring a
servlet. Given I want to expose a single string argument, this seemed
appropriate.
I would also like to avoid introducing CXF-specific code in the
application: the servlet should work with any JAX-RS implementation.
Therefore, it looked like injecting the init-param as a spring argument
is the way forward and an initial scan of CXF code-base suggests that
CXF servlet transport does inject the key-value pairs from init-param
arguments into a named Spring bean; however, when I try this it doesn't
work.
Here are some details. The web.xml:
http://java.sun.com/xml/ns/javaee;
xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance;
xsi:schemaLocation="http://java.sun.com/xml/ns/javaee
http://java.sun.com/xml/ns/javaee/web-app_2_5.xsd;>
contextConfigLocation
classpath:applicationContext.xml
org.springframework.web.context.ContextLoaderListener
CXFServlet
org.apache.cxf.transport.servlet.CXFServlet
filesystemBaseDirectory
/home/paul/test-dir
1
CXFServlet
/*
Some fragments from applicationContext.xml, omitting some unrelated parts.
With the above configuration, the SpEL reference
#{contextParameters['filesystemBaseDirectory']} expands to null.
If I add context-param declaration to web.xml then the SpEL expression
expands to the correct value.
Any ideas how to get init-param to work?
Cheers,
Paul.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: AW: CXF 3.0.7 JAXRS POST application/x-www-form-urlencoded
Hi
I guess you can not make it work reliably with @BeanParam bean.
First of all if a given method has a JAX-RS annotation then the
superclasses are not checked, this is per the annotation inheritance
spec restriction. And perhaps you can get the ordering of methods being
different with various JDKs.
Perhaps you can make it work with something like:
@Consumes("application/x-www-form-urlencoded")
post(@Context UriInfo ui, MultivaluedMap<String, String> formPayload) {
String a = getParam(ui, formPayload, "a");
}
String get(UriInfo ui, MultivaluedMap<String, String> formPayload,
String paramName) {
String value = ui.getQueryParameters().getFirst(paramName);
return value != null ? value : formPayload.getFirst(paramName);
}
Sergey
On 26/01/16 08:36, Julien Charon wrote:
Hi,
Recently, I had time to have a closer look at this. I first analysed the
behaviour of the old implementation using
(Http)ServletRequest.getParameter(String).
I observed that it is possible to mix both parameters provided in the URL and
in the body of the request, whereas the parameters in the URL are preferred to
the ones in the body.
For example, assuming a service method with 3 parameters param1, param2 and
param3:
URL: http://server1/service/method?param1=urlValue1=urlValue2
Body: param2=bodyValue2=bodyValue3
Then (Http)ServletRequest.getParameter(String) will return: param1=urlValue1,
param2=urlValue2 and param3=bodyValue3
So I tried to create a bean param class for a GET method containing @QueryParam
annotations on the setters. In a second step, I created a subclass of the bean
param class to use it for a POST method with application/x-www-form-urlencoded
by overriding the setters and annotating them with @FormParam using the same
parameter names. Now this is working fine as long as parameters are not mixed
like I described it above. I think this is logical because I overwrote the
setters and replaced the annotation. So I tried to create new setter methods
with a suffix, annotated those with @FormParam and make them call the setter of
the superclass. Unfortunately, this doesn't work.
Does someone have an idea how I can achieve the behaviour of the old
implementation?
Viele Grüße,
Julien
Avitech GmbH
Engineering AxL
Tel.: +49 (0)7541/282-177
Fax: +49 (0)7541/282-199
e-mail: julien.cha...@avitech.aero
Avitech GmbH
Principal Office: Bahnhofplatz 1 | 88045 Friedrichshafen | Germany
Court Registration: Amtsgericht Ulm | HRB 728293
Geschäftsführer/Managing Director: Antonio Maria Gonzalez Gorostiza
http://avitech.aero
This message may contain confidential information and is intended only for the
individual named. If you are not the named addressee you should not
disseminate, distribute or copy this e-mail. Please notify the sender
immediately by e-mail if you have received this e-mail by mistake and delete
this e-mail from your system.
-Ursprüngliche Nachricht-
Von: Julien Charon [mailto:julien.cha...@avitech.aero]
Gesendet: Donnerstag, 10. Dezember 2015 15:51
An: users@cxf.apache.org
Betreff: AW: CXF 3.0.7 JAXRS POST application/x-www-form-urlencoded
Thank you. So I will wait for 3.0.8.
However, I noticed another behaviour related to using HttpServletRequest but
with content type text/xml. If javax.servlet.ServletRequest.getInputStream() is
called, an InputStream is returned that contains no data, i.e. I can't access
the body of the POST request. Anything else I can do to avoid that than getting
the body as method parameter?
Best regards,
Julien
Avitech GmbH
Engineering AxL
Tel.: +49 (0)7541/282-177
Fax: +49 (0)7541/282-199
e-mail: julien.cha...@avitech.aero
Avitech GmbH
Principal Office: Bahnhofplatz 1 | 88045 Friedrichshafen | Germany Court
Registration: Amtsgericht Ulm | HRB 728293 Geschäftsführer/Managing Director:
Antonio Maria Gonzalez Gorostiza http://avitech.aero
This message may contain confidential information and is intended only for the
individual named. If you are not the named addressee you should not
disseminate, distribute or copy this e-mail. Please notify the sender
immediately by e-mail if you have received this e-mail by mistake and delete
this e-mail from your system.
-Ursprüngliche Nachricht-
Von: Sergey Beryozkin [mailto:sberyoz...@gmail.com]
Gesendet: Sonntag, 6. Dezember 2015 18:47
An: users@cxf.apache.org
Betreff: Re: CXF 3.0.7 JAXRS POST application/x-www-form-urlencoded
FYI:
https://issues.apache.org/jira/browse/CXF-6679
Sergey
On 04/12/15 11:59, Sergey Beryozkin wrote:
Hi
Thanks for experimenting with CXF,
https://issues.apache.org/jira/browse/CXF-6679
is open and I'm honestly not sure why it is not working because CXF
itself does not consume that data, so if it works directly with the
servlet then it should work inside the JAX-RS code.
Unless CXF does some internal caching
Re: How to limit concurrent requests served by resource?
I might have some time to look at https://issues.apache.org/jira/browse/CXF-6610 before 3.1.5 gets released if not then we will do for 3.1.6 It should be a very basic update, say if an optional property "abortWhenThresholdIsReached" is set then a specific exception is thrown, and either JAXWS or JAXRS mappers can then map it to 503 Sergey On 18/01/16 11:08, Sergey Beryozkin wrote: Hi, jaxrs:server can have features or native CXF interceptors registered within jaxrs:features and jaxrs:inInterceptors (or jaxrs:outInterceptors), so that demo should work the same way with JAXRS servers it works with JAXWS. Unfortunately I haven't experimented much myself yet, though as I said one thing we noticed that supporting optionally returning 503 instead of suspending the overflowing client requests will be needed - that should be not be difficult to do. Cheers, Sergey On 18/01/16 10:54, Vjacheslav V. Borisov wrote: Sergey, how should we adopt this approach to jax-rs services? Which is there jax-rs analogs to AbstractPhaseInterceptor ? 2016-01-15 16:19 GMT+04:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi CXF ships its own throttling feature, Dan did it, see https://fisheye6.atlassian.com/browse/cxf/distribution/src/main/release/samples/throttling The threads will be temporarily suspended If the threshold is reached. One thing that still needs to be done is to support actually returning 503 (if preferred) as opposed to suspending the clients. Sergey On 15/01/16 11:12, Vjacheslav V. Borisov wrote: Hi, How to limit concurrent requests served by specific resource, or by application iteself? Found this thread http://stackoverflow.com/questions/5563442/how-to-set-limit-to-the-number-of-concurrent-request-in-servlet Should i do something like showed in accepted answer, or cxf has his limit configuration ? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: WADL and Swagger
Hi wadl-to-java generates a Java code and optionally converts WADL docs into JavaDocs too - and with the latest improvements to the feature done by Francesco the source (+ docs) will be converted into Swagger JSON, without necessarily adding Swagger annotations... As a side note, if the only reason you'd like to do Swagger is to get some UI then note WADLGenerator can be configured with and XSLT source reference, an XSLT conversion (like the one done in Syncope :-)) can make a great difference :-) Sergey On 21/01/16 12:05, Francesco Chicchiriccò wrote: On 21/01/2016 11:34, Gary Gregory wrote: Hi All: I see that CXF includes supports for both WADL [1] and Swagger [2]. Is to possible to programatically use CXF to convert a WADL file to a Swagger file? Perhaps by first parsing WADL to some CXF internal format and then back out to Swagger? Hi Gary, don't know about WADL -> internal (maybe Sergey), but the internal -> Swagger part is what actually performed by [3] as triggered by [4] (references shown for 3.1.X). Regards. [1] https://cxf.apache.org/docs/jaxrs-services-description.html#JAXRSServicesDescription-WADLfirstDevelopment [2] https://cxf.apache.org/docs/swagger2feature.html [3] https://github.com/apache/cxf/blob/3.1.x-fixes/rt/rs/description/src/main/java/org/apache/cxf/jaxrs/swagger/Swagger2Serializers.java [4] https://github.com/apache/cxf/blob/3.1.x-fixes/rt/rs/description/src/main/java/org/apache/cxf/jaxrs/swagger/Swagger2Feature.java#L92
Re: How to generate episode file using wadl2java ?
Hi Sure, have a look please at org.apache.cxf.catalog.OASISCatalogManager, this is in the cxf-core module. I see there it checks "OASISCatalogManager.catalog.debug.level" system property - which is set, if not null, on org.apache.xml.resolver.CatalogManager, apparently CatalogManager can also check Catalog.properties file Perhaps OASISCatalogManager can also be enhanced... Re the episodes files - have a look at CustomizationParser in wadlto/jaxrs, I copied some code there from the wsdl processor, you can see jaxb bindings are modified with the target namespaces. Perhaps the similar approach can help with episodes ? Cheers, Sergey On 21/01/16 05:00, Vjacheslav V. Borisov wrote: I am talking about my progress is slow :) Where can I find xml catalog resolution code in cxf wadl2java ? May be i can patch it to enable debugging info 2016-01-20 18:24 GMT+04:00 Sergey Beryozkin <sberyoz...@gmail.com>: Well, this is one of those issues which I can not prioritize upon - I've never worked with episodes before, and given that workarounds exists it is hard to prioritize on this arguably minor issue. In many case we can temp stop working on our current commitments (which all of us have) and react fast enough when the issues are easy to tackle or when the issues are indeed critical or patches exist. I'd like to encourage you to look at providing a patch whenever you feel like the progress is slow, you obviously understand what is needed as far as episodes are concerned, so how about looking at JAXRSContainerTest in wadlto/jaxrs, look for the tests which end with "Binding" and see how CXF manages the custom jaxb binding perhaps that can give you an idea how episodes can also be handled nicely Cheers, Sergey
Re: How to generate episode file using wadl2java ?
Well, this is one of those issues which I can not prioritize upon - I've
never worked with episodes before, and given that workarounds exists it
is hard to prioritize on this arguably minor issue.
In many case we can temp stop working on our current commitments (which
all of us have) and react fast enough when the issues are easy to tackle
or when the issues are indeed critical or patches exist.
I'd like to encourage you to look at providing a patch whenever you feel
like the progress is slow, you obviously understand what is needed as
far as episodes are concerned, so how about looking at
JAXRSContainerTest in wadlto/jaxrs, look for the tests which end with
"Binding" and see how CXF manages the custom jaxb binding perhaps that
can give you an idea how episodes can also be handled nicely
Cheers, Sergey
On 20/01/16 13:20, Vjacheslav V. Borisov wrote:
2015-11-24 19:48 GMT+04:00 Sergey Beryozkin <sberyoz...@gmail.com>:
I'll look in detail at one of the existing tests where an external binding
is applied (I recall copying some JAXWS tooling code which modifies the
binding files), and will let you know, a bit later though
Found temporal workaround to recoginze geenerated episode files (binding
files for multiple schemas), i need to supply them to xjc directly, like
-xjc-b
-xjc${basedir}/src/main/resources/schemas/dictionary/dictionary.episode
Altough i cannot refer to classpath for episode file here, still waitng for
CXF realization with these bindings.
P.S.
And playing with oasis catalog file (eg to resolve schemas located in
jars), I cannot enable any debug info about applying catalog.xml
transformations, this really slows work
Still not sure what to do about WADLToJava actually recognizing those
episodes
I know that maven-jaxb2-plugin works with this generated binding (with
many
schemas attached to generation)
How can I help to investigate the problem?
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Accessing init-param in spring with servlet transport
Hi Paul
Sorry for a delay.
I'm not sure I understand, so you'd like to have some context-params and
CXFServlet specific init-params set in a given web.xml and make them
available to the application context ?
Can you please show some configuration fragments which will help to
understand what exactly are you trying to achieve ?
Cheers, Sergey
I have a small JAX-RS application that I'm running through the servlet
transport.
I'd like to make the application configurable via init-param settings in
the corresponding web.xml file. I would like to keep the code
transport-neutral, so wanted to inject properties using the
applicationContext.xml spring file.
Prima facie, this should be possible. The CXF servlet transport
populates the "contextParameters" bean (using context-param, and
init-params setting) and this bean is accessible in spring via SpEL;
e.g., #{contextParameters['key']}.
However, this is not working: while context-param values are available,
the init-params are not.
A potential complication is that this is currently done with
jetty-maven-plugin, which I'm binding to the integration testing phase
to allow integration/functional testing as part of the build process.
Ultimately, my goal is to expose configuration options. If init-param
is not supported, what is the best way of achieving this?
I think the most important dependencies are:
spring-context 4.2.4.RELEASE
spring-web 4.2.4.RELEASE
cxf-rt-frontend-jaxrs 3.1.4
Cheers,
Paul.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: How to limit concurrent requests served by resource?
Hi, jaxrs:server can have features or native CXF interceptors registered within jaxrs:features and jaxrs:inInterceptors (or jaxrs:outInterceptors), so that demo should work the same way with JAXRS servers it works with JAXWS. Unfortunately I haven't experimented much myself yet, though as I said one thing we noticed that supporting optionally returning 503 instead of suspending the overflowing client requests will be needed - that should be not be difficult to do. Cheers, Sergey On 18/01/16 10:54, Vjacheslav V. Borisov wrote: Sergey, how should we adopt this approach to jax-rs services? Which is there jax-rs analogs to AbstractPhaseInterceptor ? 2016-01-15 16:19 GMT+04:00 Sergey Beryozkin <sberyoz...@gmail.com>: Hi CXF ships its own throttling feature, Dan did it, see https://fisheye6.atlassian.com/browse/cxf/distribution/src/main/release/samples/throttling The threads will be temporarily suspended If the threshold is reached. One thing that still needs to be done is to support actually returning 503 (if preferred) as opposed to suspending the clients. Sergey On 15/01/16 11:12, Vjacheslav V. Borisov wrote: Hi, How to limit concurrent requests served by specific resource, or by application iteself? Found this thread http://stackoverflow.com/questions/5563442/how-to-set-limit-to-the-number-of-concurrent-request-in-servlet Should i do something like showed in accepted answer, or cxf has his limit configuration ? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: How to limit concurrent requests served by resource?
Hi CXF ships its own throttling feature, Dan did it, see https://fisheye6.atlassian.com/browse/cxf/distribution/src/main/release/samples/throttling The threads will be temporarily suspended If the threshold is reached. One thing that still needs to be done is to support actually returning 503 (if preferred) as opposed to suspending the clients. Sergey On 15/01/16 11:12, Vjacheslav V. Borisov wrote: Hi, How to limit concurrent requests served by specific resource, or by application iteself? Found this thread http://stackoverflow.com/questions/5563442/how-to-set-limit-to-the-number-of-concurrent-request-in-servlet Should i do something like showed in accepted answer, or cxf has his limit configuration ? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Where is my JAX-RS Invoke going?
Hi
The log shows that a method accepting a single parameter is invoked,
while the one you typed accepts 3 parameters.
Looks like you may have 2 methods which prove to be equal JAX-RS method
candidates.
Cheers, Sergey
On 12/01/16 20:09, Jim Talbut wrote:
Hi,
In a JAX-RS application running in tomcat 8 using CXF 3.1.4 with the
logs turned up to DEBUG I'm seeing this:
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor org.apache.cxf.interceptor.OneWayProcessorInterceptor@3b258a06
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor org.apache.cxf.interceptor.ServiceInvokerInterceptor@22d2b4b
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.s.invoker.AbstractInvoker - Invoking method public
javax.ws.rs.core.Response
com.groupgti.scheduler.api.resource.TaskRunsResource.get(int) throws
java.io.IOException,org.apache.cxf.jaxrs.ext.search.SearchParseException
on object com.groupgti.scheduler.api.resource.TaskRunsResource@61f150ac
with params [426347].
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor org.apache.cxf.interceptor.OutgoingChainInterceptor@6bdd002e
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by bus:
[com.groupgti.esb.cxf.interceptors.MessageLoggerOutbound@716e7107,
org.apache.cxf.ws.policy.PolicyOutInterceptor@71c4ebf3]
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by service: []
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by endpoint:
[org.apache.cxf.interceptor.MessageSenderInterceptor@2061f639]
Note that it invokes my method and then the next log record is for an
interceptor on the outgoing chain.
An excerpt of my resource class looks like this:
@Path("/taskruns")
@LinkRelation( name = "taskruns", relation = "taskruns", title = "Task
Runs" )
public class TaskRunsResource extends AbstractResource {
private static final Logger LOG =
LoggerFactory.getLogger(TaskRunsResource.class);
@GET
@Path("{id: \\d+}")
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML, MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML, MediaType.TEXT_XML})
public Response get( @PathParam("id") int taskRunId
, @QueryParam("field") final String[] fields
, @QueryParam("embed") final String[] embeds
) throws IOException, SearchParseException {
LOG.warn("Get taskrun {}", taskRunId);
So somehow CXF thinks it is invoking that method, but it never gets as
far as that LOG expression.
Other calls are working, this is the only one I've found that isn't, but
it fails consistently.
It does work when run in a junit that isn't using tomcat.
Can anyone suggest where I should start to look for the cause of this
problem?
Thanks.
Jim
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Where is my JAX-RS Invoke going?
Hi
On 13/01/16 13:05, Jim Talbut wrote:
Gaah
You're right.
Thank you.
I haven't touched this code for a long while and didn't realise I had
overloaded that method (I've now renamed the methods to help me out next
time).
I'm surprised that a request with an Accept of this:
application/json, text/plain, */*
Do you have any 'q' parameters in this Accept ?
Thanks, Sergey
Preferred this method:
@Produces({MediaType.TEXT_HTML})
over this:
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML,
MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML,
MediaType.TEXT_XML
})
I guess you just iterate through the methods (with unpredictable
ordering) and take the first that matches, rather than trying to
establish the best?
Or do you order based on the Produces order, where the two matching
types are both third in the list?
I've changed the request to have:
Accept: application/hal+json, application/json
(which is what it should have been all along, the client couldn't cope
with anything else) and now it hits the right method.
Thank you again.
Jim
On 13/01/2016 12:32, Sergey Beryozkin wrote:
Hi
The log shows that a method accepting a single parameter is invoked,
while the one you typed accepts 3 parameters.
Looks like you may have 2 methods which prove to be equal JAX-RS method
candidates.
Cheers, Sergey
On 12/01/16 20:09, Jim Talbut wrote:
Hi,
In a JAX-RS application running in tomcat 8 using CXF 3.1.4 with the
logs turned up to DEBUG I'm seeing this:
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.OneWayProcessorInterceptor@3b258a06
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor org.apache.cxf.interceptor.ServiceInvokerInterceptor@22d2b4b
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.s.invoker.AbstractInvoker - Invoking method public
javax.ws.rs.core.Response
com.groupgti.scheduler.api.resource.TaskRunsResource.get(int) throws
java.io.IOException,org.apache.cxf.jaxrs.ext.search.SearchParseException
on object com.groupgti.scheduler.api.resource.TaskRunsResource@61f150ac
with params [426347].
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor org.apache.cxf.interceptor.OutgoingChainInterceptor@6bdd002e
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by bus:
[com.groupgti.esb.cxf.interceptors.MessageLoggerOutbound@716e7107,
org.apache.cxf.ws.policy.PolicyOutInterceptor@71c4ebf3]
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by
service: []
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by endpoint:
[org.apache.cxf.interceptor.MessageSenderInterceptor@2061f639]
Note that it invokes my method and then the next log record is for an
interceptor on the outgoing chain.
An excerpt of my resource class looks like this:
@Path("/taskruns")
@LinkRelation( name = "taskruns", relation = "taskruns", title = "Task
Runs" )
public class TaskRunsResource extends AbstractResource {
private static final Logger LOG =
LoggerFactory.getLogger(TaskRunsResource.class);
@GET
@Path("{id: \\d+}")
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML, MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML, MediaType.TEXT_XML})
public Response get( @PathParam("id") int taskRunId
, @QueryParam("field") final String[] fields
, @QueryParam("embed") final String[] embeds
) throws IOException, SearchParseException {
LOG.warn("Get taskrun {}", taskRunId);
So somehow CXF thinks it is invoking that method, but it never gets as
far as that LOG expression.
Other calls are working, this is the only one I've found that isn't, but
it fails consistently.
It does work when run in a junit that isn't using tomcat.
Can anyone suggest where I should start to look for the cause of this
problem?
Thanks.
Jim
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Where is my JAX-RS Invoke going?
On 13/01/16 13:53, Sergey Beryozkin wrote:
Hi
On 13/01/16 13:05, Jim Talbut wrote:
Gaah
You're right.
Thank you.
I haven't touched this code for a long while and didn't realise I had
overloaded that method (I've now renamed the methods to help me out next
time).
I'm surprised that a request with an Accept of this:
application/json, text/plain, */*
Do you have any 'q' parameters in this Accept ?
And no text/html ? Usually a browser inserts it
Cheers, Sergey
Thanks, Sergey
Preferred this method:
@Produces({MediaType.TEXT_HTML})
over this:
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML,
MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML,
MediaType.TEXT_XML
})
I guess you just iterate through the methods (with unpredictable
ordering) and take the first that matches, rather than trying to
establish the best?
Or do you order based on the Produces order, where the two matching
types are both third in the list?
I've changed the request to have:
Accept: application/hal+json, application/json
(which is what it should have been all along, the client couldn't cope
with anything else) and now it hits the right method.
Thank you again.
Jim
On 13/01/2016 12:32, Sergey Beryozkin wrote:
Hi
The log shows that a method accepting a single parameter is invoked,
while the one you typed accepts 3 parameters.
Looks like you may have 2 methods which prove to be equal JAX-RS method
candidates.
Cheers, Sergey
On 12/01/16 20:09, Jim Talbut wrote:
Hi,
In a JAX-RS application running in tomcat 8 using CXF 3.1.4 with the
logs turned up to DEBUG I'm seeing this:
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.OneWayProcessorInterceptor@3b258a06
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.ServiceInvokerInterceptor@22d2b4b
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.s.invoker.AbstractInvoker - Invoking method public
javax.ws.rs.core.Response
com.groupgti.scheduler.api.resource.TaskRunsResource.get(int) throws
java.io.IOException,org.apache.cxf.jaxrs.ext.search.SearchParseException
on object com.groupgti.scheduler.api.resource.TaskRunsResource@61f150ac
with params [426347].
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.OutgoingChainInterceptor@6bdd002e
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by bus:
[com.groupgti.esb.cxf.interceptors.MessageLoggerOutbound@716e7107,
org.apache.cxf.ws.policy.PolicyOutInterceptor@71c4ebf3]
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by
service: []
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by
endpoint:
[org.apache.cxf.interceptor.MessageSenderInterceptor@2061f639]
Note that it invokes my method and then the next log record is for an
interceptor on the outgoing chain.
An excerpt of my resource class looks like this:
@Path("/taskruns")
@LinkRelation( name = "taskruns", relation = "taskruns", title = "Task
Runs" )
public class TaskRunsResource extends AbstractResource {
private static final Logger LOG =
LoggerFactory.getLogger(TaskRunsResource.class);
@GET
@Path("{id: \\d+}")
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML, MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML, MediaType.TEXT_XML})
public Response get( @PathParam("id") int taskRunId
, @QueryParam("field") final String[] fields
, @QueryParam("embed") final String[] embeds
) throws IOException, SearchParseException {
LOG.warn("Get taskrun {}", taskRunId);
So somehow CXF thinks it is invoking that method, but it never gets as
far as that LOG expression.
Other calls are working, this is the only one I've found that isn't,
but
it fails consistently.
It does work when run in a junit that isn't using tomcat.
Can anyone suggest where I should start to look for the cause of this
problem?
Thanks.
Jim
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: CXF remove some namespaces
Hi
It looks like you'd like to optimize (remove possibly redundant
prefixes) as opposed to change the namespaces, the transformation
feature does not do it, perhaps you can use XSLT feature to drop
namespaces prefixes
Cheers, Sergey
On 12/01/16 16:34, r.bott...@afterbit.com wrote:
Hi,
i use CXF 3.14
i would like to remove some namespaces for some elements and preserve
other..
the original XML is like that :
http://schemas.xmlsoap.org/soap/envelope/;>
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-wssecurity-secext-1.0.xsd;
xmlns:wsu="http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-wssecurity-utility-1.0.xsd;
soap:mustUnderstand="1">
MYUSER
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-username-token-profile-1.0#PasswordDigest;>QTh0CuaMLVIUpyLCM8vvobZZg28=
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-soap-message-security-1.0#Base64Binary;>B2BolkOOPUgz4uXRDQJlJA==
2016-01-12T15:58:31.142Z
http://www.w3.org/2005/08/addressing;>http://webservices.afterbit.com/HSFREQ_07_3_1A
http://www.w3.org/2005/08/addressing;>urn:uuid:59eeb370-c04c-4457-bd77-b67784c37eab
http://www.w3.org/2005/08/addressing;>https://test.webservices.afterbit.com/MYUSER
http://www.w3.org/2005/08/addressing;>
http://www.w3.org/2005/08/addressing/anonymous
http://xml.afterbit.com/2010/06/Security_v1;
xmlns:ns70="http://xml.afterbit.com/HSFREQ_07_3_1A;
xmlns:ns71="http://xml.afterbit.com/HSFRES_07_3_1A; >
http://xml.afterbit.com/2010/06/Security_v1;
xmlns:ns70="http://xml.afterbit.com/HSFREQ_07_3_1A;
xmlns:ns71="http://xml.afterbit.com/HSFRES_07_3_1A; >
Hi
and this is what i would want :
http://schemas.xmlsoap.org/soap/envelope/;>
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-wssecurity-secext-1.0.xsd;
xmlns:wsu="http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-wssecurity-utility-1.0.xsd;
soap:mustUnderstand="1">
MYUSER
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-username-token-profile-1.0#PasswordDigest;>QTh0CuaMLVIUpyLCM8vvobZZg28=
http://docs.oasis-open.org/wss/2004/01/oasis-200401-wss-soap-message-security-1.0#Base64Binary;>B2BolkOOPUgz4uXRDQJlJA==
2016-01-12T15:58:31.142Z
http://www.w3.org/2005/08/addressing;>http://webservices.afterbit.com/HSFREQ_07_3_1A
http://www.w3.org/2005/08/addressing;>urn:uuid:59eeb370-c04c-4457-bd77-b67784c37eab
http://www.w3.org/2005/08/addressing;>https://test.webservices.afterbit.com/MYUSER
http://www.w3.org/2005/08/addressing;>
http://www.w3.org/2005/08/addressing/anonymous
http://xml.afterbit.com/2010/06/Security_v1;>
Hi
that is :
1) http://xml.afterbit.com/2010/06/Security_v1;
xmlns:ns70="http://xml.afterbit.com/HSFREQ_07_3_1A;
xmlns:ns71="http://xml.afterbit.com/HSFRES_07_3_1A; >
becomes :
http://xml.afterbit.com/2010/06/Security_v1; >
2) http://xml.afterbit.com/2010/06/Security_v1;
xmlns:ns70="http://xml.afterbit.com/HSFREQ_07_3_1A;
xmlns:ns71="http://xml.afterbit.com/HSFRES_07_3_1A; >
becomes :
this is my java unsuccessful code :
Map<String, String> outTransformMap = new HashMap<String, String>();
outTransformMap.put("{http://xml.afterbit.com/HSFREQ_07_3_1A}*;, "*");
outTransformMap.put("{http://xml.afterbit.com/HSFRES_07_3_1A}*;, "*");
outTransformMap.put("{http://xml.afterbit.com/2010/06/Security_v1}*;, "*");
outTransformMap.put("AMA_SecurityHostedUser",
"{http://xml.afterbit.com/2010/06/Security_v1}AMA_SecurityHostedUser;);
org.apache.cxf.interceptor.transform.TransformOutInterceptor
transformOutInterceptor = new
org.apache.cxf.interceptor.transform.TransformOutInterceptor();
transformOutInterceptor.setOutTransformElements(outTransformMap);
client.getOutInterceptors().add(transformOutInterceptor);
Thanks!
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Where is my JAX-RS Invoke going?
Indeed, I reproduced it, but it looks like to be correct
S(*/*, text/html) = text/html;q=1;qs=1;d=0
S(application/json, application/json) = application/json;q=1;qs=1;d=0
(Section 3.7.2, 3.b)
both methods are equal candidates so the 1st method (or whichever method
from these two gets on top) is selected
add 'qs=0.9' to text/html in Produces and you will get a method with
application/json selected
Cheers, Sergey
On 13/01/16 14:22, Jim Talbut wrote:
No q, no text/html, this is coming from AngularJS using its default
Accept header.
Jim
On 13/01/2016 13:57, Sergey Beryozkin wrote:
On 13/01/16 13:53, Sergey Beryozkin wrote:
Hi
On 13/01/16 13:05, Jim Talbut wrote:
Gaah
You're right.
Thank you.
I haven't touched this code for a long while and didn't realise I had
overloaded that method (I've now renamed the methods to help me out
next
time).
I'm surprised that a request with an Accept of this:
application/json, text/plain, */*
Do you have any 'q' parameters in this Accept ?
And no text/html ? Usually a browser inserts it
Cheers, Sergey
Thanks, Sergey
Preferred this method:
@Produces({MediaType.TEXT_HTML})
over this:
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML,
MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML,
MediaType.TEXT_XML
})
I guess you just iterate through the methods (with unpredictable
ordering) and take the first that matches, rather than trying to
establish the best?
Or do you order based on the Produces order, where the two matching
types are both third in the list?
I've changed the request to have:
Accept: application/hal+json, application/json
(which is what it should have been all along, the client couldn't cope
with anything else) and now it hits the right method.
Thank you again.
Jim
On 13/01/2016 12:32, Sergey Beryozkin wrote:
Hi
The log shows that a method accepting a single parameter is invoked,
while the one you typed accepts 3 parameters.
Looks like you may have 2 methods which prove to be equal JAX-RS
method
candidates.
Cheers, Sergey
On 12/01/16 20:09, Jim Talbut wrote:
Hi,
In a JAX-RS application running in tomcat 8 using CXF 3.1.4 with the
logs turned up to DEBUG I'm seeing this:
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.OneWayProcessorInterceptor@3b258a06
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.ServiceInvokerInterceptor@22d2b4b
15:59:52.116 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.s.invoker.AbstractInvoker - Invoking method public
javax.ws.rs.core.Response
com.groupgti.scheduler.api.resource.TaskRunsResource.get(int) throws
java.io.IOException,org.apache.cxf.jaxrs.ext.search.SearchParseException
on object
com.groupgti.scheduler.api.resource.TaskRunsResource@61f150ac
with params [426347].
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.cxf.phase.PhaseInterceptorChain - Invoking handleMessage on
interceptor
org.apache.cxf.interceptor.OutgoingChainInterceptor@6bdd002e
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by bus:
[com.groupgti.esb.cxf.interceptors.MessageLoggerOutbound@716e7107,
org.apache.cxf.ws.policy.PolicyOutInterceptor@71c4ebf3]
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by
service: []
15:59:52.117 [http-apr-172.28.9.212-8080-exec-31] DEBUG
o.a.c.i.OutgoingChainInterceptor - Interceptors contributed by
endpoint:
[org.apache.cxf.interceptor.MessageSenderInterceptor@2061f639]
Note that it invokes my method and then the next log record is for an
interceptor on the outgoing chain.
An excerpt of my resource class looks like this:
@Path("/taskruns")
@LinkRelation( name = "taskruns", relation = "taskruns", title =
"Task
Runs" )
public class TaskRunsResource extends AbstractResource {
private static final Logger LOG =
LoggerFactory.getLogger(TaskRunsResource.class);
@GET
@Path("{id: \\d+}")
@Produces({RepresentationFactory.HAL_JSON,
RepresentationFactory.HAL_XML, MediaType.APPLICATION_JSON,
MediaType.APPLICATION_XML, MediaType.TEXT_XML})
public Response get( @PathParam("id") int taskRunId
, @QueryParam("field") final String[] fields
, @QueryParam("embed") final String[] embeds
) throws IOException, SearchParseException {
LOG.warn("Get taskrun {}", taskRunId);
So somehow CXF thinks it is invoking that method, but it never
gets as
far as that LOG expression.
Other calls are working, this is the only one I've found that isn't,
but
it
Re: Attaching context to an AsyncResponse
Hi Benson The idea of having a utility 'put' method was not accepted awhile back, I agree it would be useful, but you can have a map of AsyncResponse to some context bean values, you might need to keep a reference to a given AsyncResponse so that you can accumulate the response context from various code parts Cheers, Sergey On 01/01/16 16:19, Benson Margulies wrote: I have some code that calls MessageContext#put to avoid passing a handful of parameters through many functions. It occurs to me that this won't work with async; if some mechanism of mine calls back into the resource class to complete processing, the injected message context could be long gone. I'm wishing that AsyncResponse had a 'put', but no such luck. Am I missing anything? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Exceptions and Async
Hi Benson The very first call into this method does not suspend the continuation, so there's no need to resume it if the exception is returned. Do you see a mapped exception response not being returned ? Please clarify Cheers, Sergey On 02/01/16 02:01, Benson Margulies wrote: If I am in a method with an @Suspended AsyncResponse, and I neglect to catch an exception, CXF runs mappers but they don't do any good, of course, because nothing resumes the response. How about if CXF instead logged at ERROR about any exception thrown through a suspended resource method? I'm assuming that there's no way to arrange for any uncaught exception to end up as asyncResponse.resume(e).
Re: Exceptions and Async
Sounds good, the unmapped/escaped exception must not cause blocking the response, so yeah, please create a test I see a test testUnmappedAfterTimeout which asserts 500 is returned. This is a case where the current async response is suspended for some time and when it returns the unmapped exception is thrown. Perhaps when the exception escapes immediately then something needs to be tweaked Cheers, Sergey On 11/01/16 11:56, Benson Margulies wrote: On Mon, Jan 11, 2016 at 6:49 AM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi Benson On 11/01/16 11:42, Benson Margulies wrote: Sergey, I annotate @Suspended. So it is suspended on the first call. It is not, believe me :-) because this is how I implemented it. The runtime does not have to physically suspend a call, given that the method can immediately call AsyncResponse.resume(). JAXRSInvoker will only suspend the thread if no resume/cancel was called by the time the method returns. The runtimes are allowed to optimize like this. I think that perhaps we're using the word 'suspended' in two different ways? I'm only meaning that the response object itself thinks it's suspended. However, it's not important. If it's supposed to handle _thrown_ exceptions in the first call, I'll make a branch with a test case, because what I saw what is that it hung. The quote I read from the spec says that returned values from @Suspended calls are ignored. It did not say what the spec says about exceptions from those calls. If they are also supposed to be ignored, then a big fat WARN log message would be what I have in mind. The exceptions escaped from a method are supposed to be caught, 'the returned values are ignored' is probably a reference to a 'void' return type, but of course a method can immediately do asyncResponse.resume(myResponse). Yes, it says that you must use 'void' and if you return something else it will be ignored. Cheers, Sergey On Mon, Jan 11, 2016 at 5:17 AM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi Benson The very first call into this method does not suspend the continuation, so there's no need to resume it if the exception is returned. Do you see a mapped exception response not being returned ? Please clarify Cheers, Sergey On 02/01/16 02:01, Benson Margulies wrote: If I am in a method with an @Suspended AsyncResponse, and I neglect to catch an exception, CXF runs mappers but they don't do any good, of course, because nothing resumes the response. How about if CXF instead logged at ERROR about any exception thrown through a suspended resource method? I'm assuming that there's no way to arrange for any uncaught exception to end up as asyncResponse.resume(e). -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: CXF Web Client Proxy Exception
Hi Can you clarify please, how does it work for you if say a plain HttpUrlConnection is used directly ? Sounds like CXF HttpConduit proxy configuration might be applied, but I'd like to get more information first Thanks, Sergey On 04/01/16 20:14, williamo...@ups.com.INVALID wrote: Hi all, I am currently using CXF 3.1.2 with JAX-RS 2.0 and I have developed a web client to send RESTful messages to an internal url within my company. The issue I am having is with our proxy. When the message is generated and sent, it attempts to communicate with the proxy and fails. Since I am just trying to connect to a url on our intranet I do not want to communicate with the proxy. So my question is, is there a way I can add an exception to my client to force it to not attempt to connect to the internet and just search for the url on our local intranet? Just to note, I have attempted to add the proxy information to my client with proper credentials but that will not work since the url I want to connect to because that dns/ip does not exist on the internet. I have also tried using both the dns ip and hostname for the url and sill have the same issue for both. I have looked through all documentation pertaining to the client and cannot find anything addressing this issue. Thanks in advance, Chris
Re: MultipartBody Content-Type attributes dropped on upgrade from CXF 2.4.0 to 3.4.1
Hi
This is related to
https://issues.apache.org/jira/browse/CXF-4348?focusedCommentId=13488120=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13488120
The only multipart/* media type that may contain 'start' and
'start-info' attributes is, AFAIK, multipart/related.
Are you aware of any texts that support these attributes for
multipart/mixed ?
Even if it not but you do need these attributes when processing
multipart/mixed payloads then may be we can update the code to set them
but only if a property requesting it is set, let me know please
Sergey
On 04/01/16 18:40, Jacob Zwiers wrote:
After upgrading Apache CXF from 2.4.0 to 3.4.1, the `Content-Type` header on
responses from JAX-RS methods have dropped several attributes.
Under CXF 2.4.0, the header is:
Content-Type: multipart/mixed; type="application/octet-stream";
boundary="uuid:61b631f1-0aa9-4cc8-ad85-3c09129ec442"; start="";
start-info="application/octet-stream"
Under CXF 3.4.1, it is:
Content-Type: multipart/mixed;
boundary="uuid:804168d7-70ed-44e7-a471-9647372b9224"
Note: attributes `type`, `start`, `start-info` missing.
Here's the code we're using:
@GET
@Path( "{order_id}/document/{document_id}/file" )
@Produces("multipart/mixed")
public MultipartBody getDocument( @PathParam( "order_id") int _orderId, @PathParam(
"document_id") int _documentId) throws Exception {
FileInfo fileInfo = findFileInfo( _orderId, _documentId );
List atts = new ArrayList();
File internalFile = fileInfo.getActualFile();
String fileName = fileInfo.getOriginalDocumentName();
String fileSize = String.valueOf( internalFile.length() );
ContentDisposition cd = new ContentDisposition("attachment; filename=\"" + fileName +
"\"; size=" + fileSize );
InputStream inputStreamToUse = new FileInputStream( internalFile );
Attachment att = new Attachment(fileName, inputStreamToUse, cd);
atts.add( att );
return new MultipartBody(atts, true);
}
I can't find any references in the [Migration Guides][1] to changes in this
area and the style of the above method seems to match the one from the
[getBooks2() method in the JAX-RS Multipart documentation][2].
Any guidance on what might be causing the different behaviour?
[1]: http://cxf.apache.org/docs/migration-guides.html
[2]: http://cxf.apache.org/docs/jax-rs-multiparts.html#highlighter_731760
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: org.apache.cxf.interceptor.Fault: DISPATCHED,initial
] at org.apache.cxf.transport.http.AbstractHTTPDestination.flushHeaders(AbstractH TTPDestination.java:604)[158:org.apache.cxf.cxf-rt-transports-http:3.0.7] at org.apache.cxf.transport.http.AbstractHTTPDestination.flushHeaders(AbstractH TTPDestination.java:597)[158:org.apache.cxf.cxf-rt-transports-http:3.0.7] at org.apache.cxf.transport.http.AbstractHTTPDestination$WrappedOutputStream.on FirstWrite(AbstractHTTPDestination.java:775)[158:org.apache.cxf.cxf-rt-trans ports-http:3.0.7] at org.apache.cxf.io.AbstractWrappedOutputStream.write(AbstractWrappedOutputStr eam.java:47)[148:org.apache.cxf.cxf-core:3.0.7] at org.codehaus.jackson.impl.Utf8Generator._flushBuffer(Utf8Generator.java:1754 )[209:jackson-core-asl:1.9.5] at org.codehaus.jackson.impl.Utf8Generator.flush(Utf8Generator.java:1088)[209:j ackson-core-asl:1.9.5] at org.codehaus.jackson.map.ObjectMapper.writeValue(ObjectMapper.java:1615)[210 :jackson-mapper-asl:1.9.5] at org.codehaus.jackson.jaxrs.JacksonJsonProvider.writeTo(JacksonJsonProvider.j ava:558)[211:jackson-jaxrs:1.9.5] at org.apache.cxf.jaxrs.utils.JAXRSUtils.writeMessageBody(JAXRSUtils.java:1381) [165:org.apache.cxf.cxf-rt-frontend-jaxrs:3.0.7] at org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSO utInterceptor.java:235)[165:org.apache.cxf.cxf-rt-frontend-jaxrs:3.0.7] ... 41 more The next REST invocation after this exception gets a random response, from an unrelated REST method. Also there are timeouts when invoking REST methods. I would really appreciate some help with this issue, as aim out of ideas on this one. Greetings Lukasz Bialy -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Exceptions and Async
Hi Benson On 11/01/16 11:42, Benson Margulies wrote: Sergey, I annotate @Suspended. So it is suspended on the first call. It is not, believe me :-) because this is how I implemented it. The runtime does not have to physically suspend a call, given that the method can immediately call AsyncResponse.resume(). JAXRSInvoker will only suspend the thread if no resume/cancel was called by the time the method returns. The runtimes are allowed to optimize like this. The quote I read from the spec says that returned values from @Suspended calls are ignored. It did not say what the spec says about exceptions from those calls. If they are also supposed to be ignored, then a big fat WARN log message would be what I have in mind. The exceptions escaped from a method are supposed to be caught, 'the returned values are ignored' is probably a reference to a 'void' return type, but of course a method can immediately do asyncResponse.resume(myResponse). Cheers, Sergey On Mon, Jan 11, 2016 at 5:17 AM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi Benson The very first call into this method does not suspend the continuation, so there's no need to resume it if the exception is returned. Do you see a mapped exception response not being returned ? Please clarify Cheers, Sergey On 02/01/16 02:01, Benson Margulies wrote: If I am in a method with an @Suspended AsyncResponse, and I neglect to catch an exception, CXF runs mappers but they don't do any good, of course, because nothing resumes the response. How about if CXF instead logged at ERROR about any exception thrown through a suspended resource method? I'm assuming that there's no way to arrange for any uncaught exception to end up as asyncResponse.resume(e). -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: async problem ; Illegal State Exception
Hi Benson Does it happen only when org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity deals with the wrong user credentials, i.e, does it call HttpServletResponseWrapper.sendError ? May be that can explain it because that would probably bypass the async response channel (Jetty Continuation instance) allocated to a current request/response, but I'm only guessing at the moment Sergey On 29/12/15 23:11, Benson Margulies wrote: Removing jolokia made this disappear, so I don't think it belongs here, however odd it is that it only appeared with the use of CXF async. On Tue, Dec 29, 2015 at 2:38 PM, Benson Margulies <ben...@basistech.com> wrote: I switched some code of mine to use AsyncRequest, and I get some scary backtraces even though everything seems to be working just fine. Note the presence of org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity in the backtrace. Now, I am running in Karaf with the jolokia feature enabled, but until now I had no idea that jolokia was inserting itself into the process. Can anyone suggest a way to learn anything about what's happening here? 2015-12-29 14:32:14,328 | WARN | tp1821513140-131 | HttpChannel | 150 - org.eclipse.jetty.util - 9.2.10.v20150310 | /rest/worker/process java.lang.IllegalStateException: Committed at org.eclipse.jetty.server.Response.resetBuffer(Response.java:1253)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Response.sendError(Response.java:567)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Response.sendError(Response.java:544)[147:org.eclipse.jetty.server:9.2.10.v20150310] at javax.servlet.http.HttpServletResponseWrapper.sendError(HttpServletResponseWrapper.java:167)[51:javax.servlet-api:3.1.0] at org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity(BasicAuthenticationHttpContext.java:49)[164:org.jolokia.osgi:1.3.0] at org.ops4j.pax.web.service.jetty.internal.HttpServiceServletHandler.doHandle(HttpServiceServletHandler.java:68)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.server.handler.ScopedHandler.handle(ScopedHandler.java:143)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.security.SecurityHandler.handle(SecurityHandler.java:577)[146:org.eclipse.jetty.security:9.2.10.v20150310] at org.eclipse.jetty.server.session.SessionHandler.doHandle(SessionHandler.java:223)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ContextHandler.doHandle(ContextHandler.java:1127)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.ops4j.pax.web.service.jetty.internal.HttpServiceContext.doHandle(HttpServiceContext.java:271)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.servlet.ServletHandler.doScope(ServletHandler.java:515)[148:org.eclipse.jetty.servlet:9.2.10.v20150310] at org.eclipse.jetty.server.session.SessionHandler.doScope(SessionHandler.java:185)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ContextHandler.doScope(ContextHandler.java:1061)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ScopedHandler.handle(ScopedHandler.java:141)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.ops4j.pax.web.service.jetty.internal.JettyServerHandlerCollection.handle(JettyServerHandlerCollection.java:99)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.server.handler.HandlerWrapper.handle(HandlerWrapper.java:97)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Server.handleAsync(Server.java:553)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.HttpChannel.handle(HttpChannel.java:317)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.HttpChannel.run(HttpChannel.java:261)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.util.thread.QueuedThreadPool.runJob(QueuedThreadPool.java:635)[150:org.eclipse.jetty.util:9.2.10.v20150310] at org.eclipse.jetty.util.thread.QueuedThreadPool$3.run(QueuedThreadPool.java:555)[150:org.eclipse.jetty.util:9.2.10.v20150310] -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: async problem ; Illegal State Exception
Hi On 30/12/15 13:13, Benson Margulies wrote: On Wed, Dec 30, 2015 at 6:28 AM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: Hi Benson Does it happen only when org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity Sergey, there are no credentials anywhere that I know of. I'm invoking my endpoint, not jolokia, and my endpoint has no credentials configured of any kind. Jolokita context is still invoked though and it affects it somehow, so I guess going via its code may shed some light. Cheers, Sergey deals with the wrong user credentials, i.e, does it call HttpServletResponseWrapper.sendError ? May be that can explain it because that would probably bypass the async response channel (Jetty Continuation instance) allocated to a current request/response, but I'm only guessing at the moment Sergey On 29/12/15 23:11, Benson Margulies wrote: Removing jolokia made this disappear, so I don't think it belongs here, however odd it is that it only appeared with the use of CXF async. On Tue, Dec 29, 2015 at 2:38 PM, Benson Margulies <ben...@basistech.com> wrote: I switched some code of mine to use AsyncRequest, and I get some scary backtraces even though everything seems to be working just fine. Note the presence of org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity in the backtrace. Now, I am running in Karaf with the jolokia feature enabled, but until now I had no idea that jolokia was inserting itself into the process. Can anyone suggest a way to learn anything about what's happening here? 2015-12-29 14:32:14,328 | WARN | tp1821513140-131 | HttpChannel | 150 - org.eclipse.jetty.util - 9.2.10.v20150310 | /rest/worker/process java.lang.IllegalStateException: Committed at org.eclipse.jetty.server.Response.resetBuffer(Response.java:1253)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Response.sendError(Response.java:567)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Response.sendError(Response.java:544)[147:org.eclipse.jetty.server:9.2.10.v20150310] at javax.servlet.http.HttpServletResponseWrapper.sendError(HttpServletResponseWrapper.java:167)[51:javax.servlet-api:3.1.0] at org.jolokia.osgi.security.BasicAuthenticationHttpContext.handleSecurity(BasicAuthenticationHttpContext.java:49)[164:org.jolokia.osgi:1.3.0] at org.ops4j.pax.web.service.jetty.internal.HttpServiceServletHandler.doHandle(HttpServiceServletHandler.java:68)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.server.handler.ScopedHandler.handle(ScopedHandler.java:143)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.security.SecurityHandler.handle(SecurityHandler.java:577)[146:org.eclipse.jetty.security:9.2.10.v20150310] at org.eclipse.jetty.server.session.SessionHandler.doHandle(SessionHandler.java:223)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ContextHandler.doHandle(ContextHandler.java:1127)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.ops4j.pax.web.service.jetty.internal.HttpServiceContext.doHandle(HttpServiceContext.java:271)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.servlet.ServletHandler.doScope(ServletHandler.java:515)[148:org.eclipse.jetty.servlet:9.2.10.v20150310] at org.eclipse.jetty.server.session.SessionHandler.doScope(SessionHandler.java:185)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ContextHandler.doScope(ContextHandler.java:1061)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.handler.ScopedHandler.handle(ScopedHandler.java:141)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.ops4j.pax.web.service.jetty.internal.JettyServerHandlerCollection.handle(JettyServerHandlerCollection.java:99)[191:org.ops4j.pax.web.pax-web-jetty:4.2.2] at org.eclipse.jetty.server.handler.HandlerWrapper.handle(HandlerWrapper.java:97)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.Server.handleAsync(Server.java:553)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.HttpChannel.handle(HttpChannel.java:317)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.server.HttpChannel.run(HttpChannel.java:261)[147:org.eclipse.jetty.server:9.2.10.v20150310] at org.eclipse.jetty.util.thread.QueuedThreadPool.runJob(QueuedThreadPool.java:635)[150:org.eclipse.jetty.util:9.2.10.v20150310] at org.eclipse.jetty.util.thread.QueuedThreadPool$3.run(QueuedThreadPool.java:555)[150:org.eclipse.jetty.util:9.2.10.v20150310] -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Playing with readEntity(String.class)
Hi
This is to do with a default CXF JSONProvider which is Jettison based.
Jettison, historically, escapes forward slashes, I don't know why, it
was there when I started maintaining it.
What you can do is to configure CXF JSONProvider not to do it, set its
'escapeForwardSlashesAlways' to false.
Or use a Jackson provider instead (if you do - Make sure Jettison is on
on the classpath)
HTH, Sergey
On 29/12/15 14:40, Vincenzo D'Amore wrote:
Hi All,
I don't understand why when I receive a json encoded string this is not
decoded automatically.
I wrote this code:
Client client = ClientBuilder.newClient().register(JSONProvider.class);
WebTarget target = client.target("http://example.org/rest/service1;);
target = target.queryParam("method", "method1");
Entity entity = Entity.entity(new
EndpointRequest(""),
MediaType.APPLICATION_JSON);
Response response = builder.post(entity);
System.out.println( response.getStatus() );
if (response.getStatus() == 200) {
// The problem comes here
String basePath = response.readEntity(String.class);
System.out.println( basePath );
}
The request is successfully executed but basePath contains
"\/opt\/local\/application\/rest\/" (backslash and double quotes included)
basePath should instead contain this: /opt/local/application/rest/
It seems to me, the json deserialization hasn't be triggered when it should.
Thanks in advance for your help,
Vincenzo
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Log empty message (with HTTP status code only)
I think it might be because 202 is handled by a oneway processor and the outbound chain is not invoked though I might be wrong... Sergey On 29/12/15 11:17, Sergey Maslov wrote: The issue is reproduced for with 202 - Accepted http status. On Tue, Dec 29, 2015 at 1:28 PM, Sergey Maslov <therealma...@gmail.com> wrote: Hi! I need to log imcoming requests with corresponsing responses. According I add custom LoggingFeature and implement LogEventSender interface. The messages log fine except messages with emtpty body and containing http status code only (these are the responses to the service). According the contract, these messages are needed to be sent. How could I log them? -- *Sergey Maslov* -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Log empty message (with HTTP status code only)
I'm not sure right now, perhaps Aki or Dan may have an idea. May be you can register a custom Log interceptor, after LoggingInInterceptor, that will check CXF message if it is one way (there has to be some property set in the message) and log only if it is oneway... Cheers, Sergey On 29/12/15 18:00, Sergey Maslov wrote: Sergey, thank you. But how could I log the fact of 202 response? Segery Maslov On Tue, Dec 29, 2015 at 11:55 PM, Sergey Beryozkin <sberyoz...@gmail.com> wrote: I think it might be because 202 is handled by a oneway processor and the outbound chain is not invoked though I might be wrong... Sergey On 29/12/15 11:17, Sergey Maslov wrote: The issue is reproduced for with 202 - Accepted http status. On Tue, Dec 29, 2015 at 1:28 PM, Sergey Maslov <therealma...@gmail.com> wrote: Hi! I need to log imcoming requests with corresponsing responses. According I add custom LoggingFeature and implement LogEventSender interface. The messages log fine except messages with emtpty body and containing http status code only (these are the responses to the service). According the contract, these messages are needed to be sent. How could I log them? -- *Sergey Maslov* -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/ -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: Unit-testing jaxrs:server created by spring
Hi This is a log info from a CXF Server that is not controlled directly by the JAX-RS runtime. If you have say jaxrs:server/@address="/" then the value of @address is what you'll see in the logs, So you can assume the JAX-RS server address is something like "http://localhost:8080; + jaxrs:server/@address Sergey On 28/12/15 11:51, Maxim Solodovnik wrote: Hello, I'm trying to add automatic tests to our build. our applicationContext.xml contains our tests are subclasses of AbstractJUnit4SpringContextTests according to the logs server is starting: 17:16:26.901 [main] INFO org.apache.cxf.endpoint.ServerImpl - Setting the server's publish address to be / how can I connect to it using WebClient? what URL should I use? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
New OAuthDataProvider methods on the trunkk/3.1.x
Hi I had to update OAuthDataProvider interface and add two methods for listing the access and refresh tokens associated with a given OAuth2 client. Also, AuthorizationCodeDataProvider which extends OAuthDataProvider has a new method for listing the pending code grants associated with a given Client, so we have List getAccessTokens(Client) List getRefreshTokens(Client) and List getCodeGrants(Client) The reason I've done it is that it has become obvious OAuthDataProvider interface supports managing the existing tokens/grants in a very limited way. It is really important from a management point of view to show all the access and/or tokens associated with a given client or user. I could've introduced new interfaces and made only the utility (AbstractOAuthDataProvider) provider shipped with CXF implement them but I did not because: - I felt like that it would add an extra 'noise' into an already complex OAuth2 runtime/API - after all, listing the existing tokens is quite a fundamental thing which would not be a new 'challenge' for the existing provider implementations - they store the tokens, can extract them by their ids, so obviously they can get them listed too I had to do it for 3.1.5-SNAPSHOT because the OIDC module we do in Fediz does depend on 3.1.5-SNAPSHOT and we can not wait till 3.2.0 gets released (which is quite far away at this point of time). Hope those who are working with CXF 3.1.4 and doing some custom OAuth2 servers will not find it difficult to migrate. However if someone thinks it can pose some problems for their productions (and this is why I send this mail to users) then I can revisit the idea of getting the above methods added to dedicated interfaces. Let me know in this thread or contact me privately if you prefer FYI, besides the above changes I've deprecated OAuthDataProvider.removeAccessToken but will remove it only in 3.2.0-SNAPSHOT (master). The reason being is that OAuthDataProvider also has a revokeToken accepting a token hint which is part of the standard oAuth2 token revocation API. Removing OAuthDataProvider.removeAccessToken will make this interface a bit more minimalistic Thanks Sergey
Re: How to handle ClientAbortException in CXF ?
lve.java:123)
at
org.apache.catalina.authenticator.AuthenticatorBase.invoke(AuthenticatorBase.java:502)
at
org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:171)
at
org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:100)
at org.apache.catalina.valves.AccessLogValve.invoke(AccessLogValve.java:953)
at
org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:118)
at
org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:408)
at
org.apache.coyote.http11.AbstractHttp11Processor.process(AbstractHttp11Processor.java:1041)
at
org.apache.coyote.AbstractProtocol$AbstractConnectionHandler.process(AbstractProtocol.java:603)
at
org.apache.tomcat.util.net.JIoEndpoint$SocketProcessor.run(JIoEndpoint.java:312)
at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:895)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:918)
at java.lang.Thread.run(Thread.java:695)
Caused by: ClientAbortException: java.net.SocketException: Broken pipe
at
org.apache.catalina.connector.OutputBuffer.realWriteBytes(OutputBuffer.java:413)
at org.apache.tomcat.util.buf.ByteChunk.flushBuffer(ByteChunk.java:480)
at org.apache.tomcat.util.buf.ByteChunk.append(ByteChunk.java:366)
at
org.apache.catalina.connector.OutputBuffer.writeBytes(OutputBuffer.java:438)
at org.apache.catalina.connector.OutputBuffer.write(OutputBuffer.java:426)
at
org.apache.catalina.connector.CoyoteOutputStream.write(CoyoteOutputStream.java:91)
at
org.apache.cxf.io.AbstractWrappedOutputStream.write(AbstractWrappedOutputStream.java:51)
at
org.apache.cxf.io.CacheAndWriteOutputStream.write(CacheAndWriteOutputStream.java:80)
at
com.fasterxml.jackson.core.json.UTF8JsonGenerator._flushBuffer(UTF8JsonGenerator.java:1731)
at
com.fasterxml.jackson.core.json.UTF8JsonGenerator.flush(UTF8JsonGenerator.java:1070)
at
com.fasterxml.jackson.databind.ObjectWriter.writeValue(ObjectWriter.java:426)
at
com.fasterxml.jackson.jaxrs.json.JacksonJsonProvider.writeTo(JacksonJsonProvider.java:578)
at
org.apache.cxf.jaxrs.utils.JAXRSUtils.writeMessageBody(JAXRSUtils.java:1363)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSOutInterceptor.java:244)
... 36 more
Caused by: java.net.SocketException: Broken pipe
at java.net.SocketOutputStream.socketWrite0(Native Method)
at java.net.SocketOutputStream.socketWrite(SocketOutputStream.java:92)
at java.net.SocketOutputStream.write(SocketOutputStream.java:136)
at
org.apache.coyote.http11.InternalOutputBuffer.realWriteBytes(InternalOutputBuffer.java:215)
at org.apache.tomcat.util.buf.ByteChunk.flushBuffer(ByteChunk.java:480)
at org.apache.tomcat.util.buf.ByteChunk.append(ByteChunk.java:366)
at
org.apache.coyote.http11.InternalOutputBuffer$OutputStreamOutputBuffer.doWrite(InternalOutputBuffer.java:240)
at
org.apache.coyote.http11.filters.ChunkedOutputFilter.doWrite(ChunkedOutputFilter.java:117)
at
org.apache.coyote.http11.AbstractOutputBuffer.doWrite(AbstractOutputBuffer.java:192)
at org.apache.coyote.Response.doWrite(Response.java:517)
at
org.apache.catalina.connector.OutputBuffer.realWriteBytes(OutputBuffer.java:408)
... 49 more
Process finished with exit code 137
{http://X.X.X.X } - It was package name string ,which I hve replaced with
X.X.X.X
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: How to handle ClientAbortException in CXF ?
Well, first thing to check is whether this mapper is invoked, note it is
not auto-discovered by default, unless you explicitly enable the
auto-discovery, so it needs to be registered in jaxrs:providers.
But even if it is invoked, the problem is you still try to return some
error message - the runtime will again attempt to write it to the output
stream which is already terminated.
So make sure it is invoked, then make Throwable more specific and
finally, do not set an entity on Response
Sergey
On 23/12/15 16:09, Sumit Arora wrote:
I have tried that as well, that didn't work :
@Provider
public class GTMGenericExceptionMapper implements
ExceptionMapper {
public Response toResponse(Throwable th) {
ErrorMessage errorMessage = new
ErrorMessage(th.getMessage(),500,"Example Message");
return Response.status(Response.Status.INTERNAL_SERVER_ERROR)
.header("Content-Type", "application/json")
.entity(errorMessage)
.build();
}
}
On Wed, Dec 23, 2015 at 9:35 PM, Sergey Beryozkin <sberyoz...@gmail.com>
wrote:
Try registering ExceptionMapper and see what happens
Cheers. Sergey
On 23/12/15 16:03, Sumit Arora wrote:
Sergey,
I tried using to write Exception Mapper :
1.Here MapperException - > IOException, ClientErrorException
public class GTMClientExceptionMapper implements
ExceptionMapper {
public Response toResponse(IOException ce) {
ErrorMessage errorMessage = new
ErrorMessage(ce.getMessage(),400,"Example : XYZ");
return Response.status(Response.Status.NOT_FOUND)
.header("Content-Type", "application/json")
.entity(errorMessage)
.build();
}
}
But it doesn't work, other exception mapper stuff is working ?
May you please tell me If there is something else I supposed to do ? :
JAXRSOutInterceptor.handleWriteException
2. Suggest me one example of this : Another option to try is to wrap
Jackson with a custom JAX-RS MessageBoryWriter and handle the write
exception there.
3.May you suggest any example of this :
Finally, when working with AsyncResponse is used, one might want to
register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
'Sumit
On Wed, Dec 23, 2015 at 7:13 PM, Sergey Beryozkin <sberyoz...@gmail.com>
wrote:
Hi
JAXRSOutInterceptor.handleWriteException checks JAX-RS ExceptionMapper so
perhaps you can intercept that Exception and return Response with the
code
only from this exception mapper.
Another option to try is to wrap Jackson with a custom JAX-RS
MessageBoryWriter and handle the write exception there.
Finally, when working with AsyncResponse is used, one might want to
register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
HTH, Sergey
On 23/12/15 11:47, Sumit Arora wrote:
Hi All,
I have made a web application, which uses Apache CXF on backend. In
many
situation - Client(user) uses the web application to fetch some data to
view ( say click on button) which CXF web service provide from the
backend, But Client closes the browser, however the cxf web service was
busy to bring the data. So on logs it shows : ClientAbortException. This
behaviour is correct. Its fine.
But I need the help on following :
Is there any way to catch such exception in the CXF webservice code and
logs the customized error message, rather showing complete logs( as
below)
? If Client has aborted the connection then is there any way to handle
such
situation and act accordingly?
[http-bio-8080-exec-2:ERROR] org.apache.cxf.jaxrs.utils.JAXRSUtils:
Problem
with writing the data, class java.util.ArrayList, ContentType:
application/json
[http-bio-8080-exec-2:WARN] org.apache.cxf.phase.PhaseInterceptorChain:
Interceptor for {http://X.X.X.X }GTMDataServiceEndPointImpl
<http://X.X.X.X/%7DGTMDataServiceEndPointImpl> has thrown exception,
unwinding now
org.apache.cxf.interceptor.Fault
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleWriteException(JAXRSOutInterceptor.java:363)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSOutInterceptor.java:266)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleWriteException(JAXRSOutInterceptor.java:365)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSOutInterceptor.java:266)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.processResponse(JAXRSOutInterceptor.java:117)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleMessage(JAXRSOutInterceptor.java:80)
at
org.apache.cxf.phase.PhaseInterceptorChain.doIntercept(PhaseInterceptorChain.java:307)
at
org.apache.cxf.interceptor.OutgoingChainInterceptor.handleMessage(OutgoingChainInterceptor.java:81)
at
org.
Re: How to handle ClientAbortException in CXF ?
Try registering ExceptionMapper and see what happens
Cheers. Sergey
On 23/12/15 16:03, Sumit Arora wrote:
Sergey,
I tried using to write Exception Mapper :
1.Here MapperException - > IOException, ClientErrorException
public class GTMClientExceptionMapper implements
ExceptionMapper {
public Response toResponse(IOException ce) {
ErrorMessage errorMessage = new
ErrorMessage(ce.getMessage(),400,"Example : XYZ");
return Response.status(Response.Status.NOT_FOUND)
.header("Content-Type", "application/json")
.entity(errorMessage)
.build();
}
}
But it doesn't work, other exception mapper stuff is working ?
May you please tell me If there is something else I supposed to do ? :
JAXRSOutInterceptor.handleWriteException
2. Suggest me one example of this : Another option to try is to wrap
Jackson with a custom JAX-RS MessageBoryWriter and handle the write
exception there.
3.May you suggest any example of this :
Finally, when working with AsyncResponse is used, one might want to register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
'Sumit
On Wed, Dec 23, 2015 at 7:13 PM, Sergey Beryozkin <sberyoz...@gmail.com>
wrote:
Hi
JAXRSOutInterceptor.handleWriteException checks JAX-RS ExceptionMapper so
perhaps you can intercept that Exception and return Response with the code
only from this exception mapper.
Another option to try is to wrap Jackson with a custom JAX-RS
MessageBoryWriter and handle the write exception there.
Finally, when working with AsyncResponse is used, one might want to
register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
HTH, Sergey
On 23/12/15 11:47, Sumit Arora wrote:
Hi All,
I have made a web application, which uses Apache CXF on backend. In many
situation - Client(user) uses the web application to fetch some data to
view ( say click on button) which CXF web service provide from the
backend, But Client closes the browser, however the cxf web service was
busy to bring the data. So on logs it shows : ClientAbortException. This
behaviour is correct. Its fine.
But I need the help on following :
Is there any way to catch such exception in the CXF webservice code and
logs the customized error message, rather showing complete logs( as below)
? If Client has aborted the connection then is there any way to handle
such
situation and act accordingly?
[http-bio-8080-exec-2:ERROR] org.apache.cxf.jaxrs.utils.JAXRSUtils:
Problem
with writing the data, class java.util.ArrayList, ContentType:
application/json
[http-bio-8080-exec-2:WARN] org.apache.cxf.phase.PhaseInterceptorChain:
Interceptor for {http://X.X.X.X }GTMDataServiceEndPointImpl
<http://X.X.X.X/%7DGTMDataServiceEndPointImpl> has thrown exception,
unwinding now
org.apache.cxf.interceptor.Fault
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleWriteException(JAXRSOutInterceptor.java:363)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSOutInterceptor.java:266)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleWriteException(JAXRSOutInterceptor.java:365)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.serializeMessage(JAXRSOutInterceptor.java:266)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.processResponse(JAXRSOutInterceptor.java:117)
at
org.apache.cxf.jaxrs.interceptor.JAXRSOutInterceptor.handleMessage(JAXRSOutInterceptor.java:80)
at
org.apache.cxf.phase.PhaseInterceptorChain.doIntercept(PhaseInterceptorChain.java:307)
at
org.apache.cxf.interceptor.OutgoingChainInterceptor.handleMessage(OutgoingChainInterceptor.java:81)
at
org.apache.cxf.phase.PhaseInterceptorChain.doIntercept(PhaseInterceptorChain.java:307)
at
org.apache.cxf.transport.ChainInitiationObserver.onMessage(ChainInitiationObserver.java:121)
at
org.apache.cxf.transport.http.AbstractHTTPDestination.invoke(AbstractHTTPDestination.java:243)
at
org.apache.cxf.transport.servlet.ServletController.invokeDestination(ServletController.java:223)
at
org.apache.cxf.transport.servlet.ServletController.invoke(ServletController.java:197)
at
org.apache.cxf.transport.servlet.ServletController.invoke(ServletController.java:149)
at
org.apache.cxf.transport.servlet.CXFNonSpringServlet.invoke(CXFNonSpringServlet.java:171)
at
org.apache.cxf.transport.servlet.AbstractHTTPServlet.handleRequest(AbstractHTTPServlet.java:286)
at
org.apache.cxf.transport.servlet.AbstractHTTPServlet.doGet(AbstractHTTPServlet.java:211)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:621)
at
org.apache.cxf.transport.servlet.AbstractHTTPServlet.service(AbstractHTTPServlet.java:262)
at
org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:305)
at
org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:210)
at
org.apa
Re: How to handle ClientAbortException in CXF ?
As I have already said, return Response with the error code only.
Returning null indicates to the runtime that the mapping has not been
done in which case it has to rethrow the original exception.
If returning a response with the code only does not help then move to
the alternative 2 (write your own MessageBodyWriter which will
internally delegate to Jackson MessageBodyWriter - it will be loaded
inside your custom MBW, and register it instead of Jackson with the
endpoint).
If that does not work - move to the alternative 3. Introduce
AsyncResponse first into your code (see for ex AsyncResponse package
docs), and then once you have it, you will be able to register the
callback with this AsyncResponse
Sergey
On 23/12/15 16:28, Sumit Arora wrote:
It comes here on exception mapper :
public class GTMClientExceptionMapper implements
ExceptionMapper {
public Response toResponse(IOException ce) {
// ErrorMessage errorMessage = new
ErrorMessage(ce.getMessage(),400,"Exmaple Message ");
return null;
// return Response.status(Response.Status.NOT_FOUND)
//.header("Content-Type", "application/json")
// .build();
}
}
I tried to various stuff, but still throw the error, any guess here ?
On Wed, Dec 23, 2015 at 9:45 PM, Sergey Beryozkin <sberyoz...@gmail.com>
wrote:
Well, first thing to check is whether this mapper is invoked, note it is
not auto-discovered by default, unless you explicitly enable the
auto-discovery, so it needs to be registered in jaxrs:providers.
But even if it is invoked, the problem is you still try to return some
error message - the runtime will again attempt to write it to the output
stream which is already terminated.
So make sure it is invoked, then make Throwable more specific and finally,
do not set an entity on Response
Sergey
On 23/12/15 16:09, Sumit Arora wrote:
I have tried that as well, that didn't work :
@Provider
public class GTMGenericExceptionMapper implements
ExceptionMapper {
public Response toResponse(Throwable th) {
ErrorMessage errorMessage = new
ErrorMessage(th.getMessage(),500,"Example Message");
return Response.status(Response.Status.INTERNAL_SERVER_ERROR)
.header("Content-Type", "application/json")
.entity(errorMessage)
.build();
}
}
On Wed, Dec 23, 2015 at 9:35 PM, Sergey Beryozkin <sberyoz...@gmail.com>
wrote:
Try registering ExceptionMapper and see what happens
Cheers. Sergey
On 23/12/15 16:03, Sumit Arora wrote:
Sergey,
I tried using to write Exception Mapper :
1.Here MapperException - > IOException, ClientErrorException
public class GTMClientExceptionMapper implements
ExceptionMapper {
public Response toResponse(IOException ce) {
ErrorMessage errorMessage = new
ErrorMessage(ce.getMessage(),400,"Example : XYZ");
return Response.status(Response.Status.NOT_FOUND)
.header("Content-Type", "application/json")
.entity(errorMessage)
.build();
}
}
But it doesn't work, other exception mapper stuff is working ?
May you please tell me If there is something else I supposed to do ? :
JAXRSOutInterceptor.handleWriteException
2. Suggest me one example of this : Another option to try is to wrap
Jackson with a custom JAX-RS MessageBoryWriter and handle the write
exception there.
3.May you suggest any example of this :
Finally, when working with AsyncResponse is used, one might want to
register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
'Sumit
On Wed, Dec 23, 2015 at 7:13 PM, Sergey Beryozkin <sberyoz...@gmail.com
wrote:
Hi
JAXRSOutInterceptor.handleWriteException checks JAX-RS ExceptionMapper
so
perhaps you can intercept that Exception and return Response with the
code
only from this exception mapper.
Another option to try is to wrap Jackson with a custom JAX-RS
MessageBoryWriter and handle the write exception there.
Finally, when working with AsyncResponse is used, one might want to
register
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/container/ConnectionCallback.html
HTH, Sergey
On 23/12/15 11:47, Sumit Arora wrote:
Hi All,
I have made a web application, which uses Apache CXF on backend. In
many
situation - Client(user) uses the web application to fetch some data
to
view ( say click on button) which CXF web service provide from the
backend, But Client closes the browser, however the cxf web service
was
busy to bring the data. So on logs it shows : ClientAbortException.
This
behaviour is correct. Its fine.
But I need the help on following :
Is there any way to catch such exception in the CXF webservice code
and
logs the customized error message, rather showing complete logs( as
belo
Re: FIQL query validation
I've prototyped PropertyNameConverter and did some basic documentation here: https://cwiki.apache.org/confluence/display/CXF20DOC/JAX-RS+Search#JAX-RSSearch-Dealingwithmistypedpropertynames Cheers, Sergey On 04/11/15 11:07, Sergey Beryozkin wrote: Hi Thanks for experimenting with it, I've been actually keeping this thread in the 'priority' line because may be we can have Beanspector producing some default case-insensitive line that would be optionally used without the user having to build the map manually (unless aliases are used). Besides, I'm still planning to introduce an interface for more matches to be possible in principle, should the application wish to do it, example, from "supplierCod" (a user forgot to type 'e') or "supplierCodee" (a user typed an extra 'e'), etc, to "supplierCode" Cheers, Sergey On 04/11/15 10:52, Diether wrote: That is actually a great idea. I just tried it and it works fine. Seems like I was seriously overcomplicating things :-) -- View this message in context: http://cxf.547215.n5.nabble.com/FIQL-query-validation-tp5762439p5762553.html Sent from the cxf-user mailing list archive at Nabble.com. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: CXF JAX-RS response chunk/buffer size
Hi It might be possible to control with a non-Servlet, standalone CXF HTTP Jetty transport, but perhaps you can configure it with in Tomcat/etc server.xml, there might be a way to do it. Or add a ContainerResponseFilter with HttpServletResponse injected - this filter can be shared between all the application endpoints Sergey On 18/12/15 06:41, Julien Charon wrote: Hi, Does CXF somehow support the configuration of the response chunk/buffer size for Transfer-Encoding: chunked? Until now, using non-CXF/non-JAX-RS implementation, I'm doing this by calling javax.servlet.ServletResponse.setBufferSize(int). However, I'd like to use response streaming as described in http://cxf.apache.org/docs/jax-rs-basics.html#JAX-RSBasics-ResponseStreaming. One way I can think of is adding the javax.servlet.ServletResponse to my methods using the @Context annotation and calling javax.servlet.ServletResponse.setBufferSize(int) like before, but I don't know if that works and I'd prefer to be able to configure that on a per application basis rather than adding that call to each method. Best regards, Julien Avitech GmbH Engineering AxL Tel.: +49 (0)7541/282-177 Fax: +49 (0)7541/282-199 e-mail: julien.cha...@avitech.aero<mailto:julien.cha...@avitech.aero> Avitech GmbH Principal Office: Bahnhofplatz 1 | 88045 Friedrichshafen | Germany Court Registration: Amtsgericht Ulm | HRB 728293 Geschäftsführer/Managing Director: Antonio Maria Gonzalez Gorostiza http://avitech.aero<http://avitech.aero/> This message may contain confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
Re: RS service method accepting 2 parameters
Hi Max
It is unfortunate Form accepts parameters with String values, I think we
missed it should be Object instead given that ParamConverterProvider is
available in JAX-RS 2.0.
form() can deal with MultipartMap - CXF ships MetadataMap which
implements it, and JAX-RS 2.0 ships a simple implementation too:
https://jax-rs-spec.java.net/nonav/2.0-rev-a/apidocs/javax/ws/rs/core/MultivaluedHashMap.html
You'll still need registering ParamConverterProvider if values are not
String and the beans do not override toString()
HTH, Sergey
It has to be MultivaluedMap, JAX-RS
On 16/12/15 10:04, Maxim Solodovnik wrote:
Hello All,
I'm currently trying to write RS service accepting 2 objects of different
types as parameters:
ex. code:
@WebService(serviceName="org.apache.openmeetings.webservice.UserWebService")
@Features(features = "org.apache.cxf.feature.LoggingFeature")
@Produces({MediaType.APPLICATION_JSON})
@Path("/user")
public class UserWebService {
@POST
@Path("/hash")
public ServiceResult getRoomHash(
@FormParam("user") BeanA user
, @FormParam("options") BeanB options
) throws ServiceException
{}
}
and I'm trying to invoke this method from JUnit test:
ExternalUserDTO user = new ExternalUserDTO();
RoomOptionsDTO options = new RoomOptionsDTO();
Map<String, List> map = new LinkedHashMap<>();
map.put("user", Arrays.asList((Object)user));
map.put("options", Arrays.asList((Object)options));
Response resp = getClient(USER_SERVICE_URL)
.path("/hash")
.form(map);
//throws no body wrapper for LinkedHashMap class
I cannot use .form(new Form().) due to it can only accepts string
parameters
can someone please point me to an example or provide with the hint how this
can be implemented
Thanks in advance
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Async in OSGi / pax-web
Hi Benson I suspect you may need to avoid depending on the default CXF HTTP Transport and deploy your bundle as a web bundle (with web.xml). Do you work with Aries ? If yes then use CXFBlueprintServlet: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/resources/jaxrs_soap_blueprint/WEB-INF/web.xml and have a blueprint-web-osgi dependency Cheers, Sergey On 13/12/15 18:11, Benson Margulies wrote: How does the config in http://cxf.apache.org/docs/servlet-transport.html for async map out in pax-web? is there a ConfigAdmin param that gets set?
Re: CXF 3.0.7 JAXRS POST read body
Hi
Perhaps it is supposed to work with JAX-RS, so I've fixed it as part of
https://issues.apache.org/jira/browse/CXF-6709
I'd still recommend you to avoid dealing with Servlet request/response
directly - one of the main ideas of JAX-RS was to provide a
simpler/alternative mechanism.
Thanks, Sergey
On 14/12/15 08:23, Julien Charon wrote:
Hi everybody,
I am currently refactoring a REST application whose implementation is based on
a subclass of javax.servlet.http.HttpServlet and thus is using
javax.servlet.http.HttpServletRequest and
javax.servlet.http.HttpServletResponse quite extensively to use CXF and JAXRS.
The refactored version of the application extends
org.apache.cxf.transport.servlet.CXFNonSpringServlet and is using a service
class which defines methods using the javax.ws.rs.core.Context annotation to
obtain HttpServletRequest objets, e.g.
@POST
public void doPost(@Context HttpServletRequest request,
@Context HttpServletResponse response) {
...
}
Now, for GET operations this works as expected. When I try to read the body of
a POST request using the javax.servlet.ServletRequest.getInputStream(), the
returned InputStream ist empty.
According to JAX-RS spec, I guess I could extend my method to something like
@POST
public void doPost(@Context HttpServletRequest request,
@Context HttpServletResponse response, String body) {
...
}
in order to get the body. However, I'd like to avoid that because I don't want
to change the existing application implementation too much.
So I have 2 questions:
1. Why does javax.servlet.ServletRequest.getInputStream() return an
InputStream that contains no data whe I using CXF? This work fine in the old
style of the application.
2. What can I do to be able to read the body that way using CXF?
Viele Grüße,
Julien
Avitech GmbH
Engineering AxL
Tel.: +49 (0)7541/282-177
Fax: +49 (0)7541/282-199
e-mail: julien.cha...@avitech.aero<mailto:julien.cha...@avitech.aero>
Avitech GmbH
Principal Office: Bahnhofplatz 1 | 88045 Friedrichshafen | Germany
Court Registration: Amtsgericht Ulm | HRB 728293
Geschäftsführer/Managing Director: Antonio Maria Gonzalez Gorostiza
http://avitech.aero<http://avitech.aero/>
This message may contain confidential information and is intended only for the
individual named. If you are not the named addressee you should not
disseminate, distribute or copy this e-mail. Please notify the sender
immediately by e-mail if you have received this e-mail by mistake and delete
this e-mail from your system.
--
Sergey Beryozkin
Talend Community Coders
http://coders.talend.com/
Re: Async in OSGi / pax-web
Hi Benson It likely can be controlled from jetty.xml in pax-web (enabling the async mode) but as we have learnt recently, pax web only checks jetty.xml if a full blown war or web bundle is deployed. Web Bundle is your regular bundle, with only web.xml added to it, plus an extra WebAppContext instruction in Manifest Sergey On 14/12/15 15:09, Benson Margulies wrote: On consideration, I really don't want to get involved in a web.xml if I can avoid it. I'm not using the stock http transport as things are, somehow the code from the service factory plugs itself into the pax-web universe. I think I'll do some sleuthing. On Mon, Dec 14, 2015 at 9:22 AM, Benson Margulies <ben...@basistech.com> wrote: I have avoided blueprint in favor of DS but both are in karaf so I can probably go there. On Dec 14, 2015 12:00 PM, "Sergey Beryozkin" <sberyoz...@gmail.com> wrote: Hi Benson I suspect you may need to avoid depending on the default CXF HTTP Transport and deploy your bundle as a web bundle (with web.xml). Do you work with Aries ? If yes then use CXFBlueprintServlet: https://github.com/apache/cxf/blob/master/systests/jaxrs/src/test/resources/jaxrs_soap_blueprint/WEB-INF/web.xml and have a blueprint-web-osgi dependency Cheers, Sergey On 13/12/15 18:11, Benson Margulies wrote: How does the config in http://cxf.apache.org/docs/servlet-transport.html for async map out in pax-web? is there a ConfigAdmin param that gets set? -- Sergey Beryozkin Talend Community Coders http://coders.talend.com/
