Commit Graph

15926 Commits

Author SHA1 Message Date
Rossen Stoyanchev aee8a9c97b Upgrade to Bismuth-SR5
Issue: SPR-16385
2018-01-22 19:57:38 -05:00
Juergen Hoeller 91e6274b7d Upgrade to Apache HttpClient 4.5.5 and Tomcat 8.5.27 2018-01-22 22:11:00 +01:00
Juergen Hoeller 61579ffe7b Polishing 2018-01-22 21:34:44 +01:00
Arjen Poutsma b47491695e Support array of scalar values in Jackson2Tokenizer
Issue: SPR-16407
2018-01-22 17:08:08 +01:00
Juergen Hoeller 9d0e62ef68 Javadoc format and related polishing 2018-01-22 11:43:21 +01:00
Juergen Hoeller 3dd6c19c6d Upgrade to Netty 4.1.20 2018-01-22 11:43:06 +01:00
Arjen Poutsma 0befc60c8f Renamed DataBufferUtils/DataBufferFactory.compose to join
Issue: SPR-16365
2018-01-22 10:51:07 +01:00
sdeleuze 3f3141cdda Add additional test for SPR-16210
Issue: SPR-16210
2018-01-22 10:29:21 +01:00
Arjen Poutsma 646fcc5c2f Support Custom Headers for Multipart Async Data
This commit makes sure there is no custom Content-Disposition header
before setting one automatically.

This commit also adds a headers(Consumer<HttpHeaders>) method, so that
one can user the nicer methods of HttpHeaders, as opposed to basic
strings.

Issue: SPR-16376
2018-01-20 15:18:20 +01:00
Juergen Hoeller b2ce98e881 Upgrade to CGLIB 3.2.6
Issue: SPR-16267
2018-01-20 14:29:56 +01:00
Juergen Hoeller 395c9bd723 Consistently call HttpURLConnection.setFixedLengthStreamingMode(long)
Issue: SPR-16404
2018-01-20 14:29:40 +01:00
andrm 142f1ab42f Removed unnecessary cast to int
Since Java7 HttpURLConnection offers setFixedLengthStreamingMode method with long parameter which should be prefered over version with int argument, therefore casting ContentLength to int is no longer needed. Moreover it makes impossible to stream payload larger than Integer.MAX_VALUE
2018-01-20 14:11:15 +01:00
Stephane Nicoll 0ca4cd1cf8 Merge pull request #1647 from izeye:polish-20180120
* pr/1647:
  Polish InterceptorRegistration
2018-01-20 10:15:53 +01:00
Johnny Lim 36d42bcda4 Polish InterceptorRegistration
Closes gh-1647
2018-01-20 10:15:12 +01:00
Rossen Stoyanchev 283811b16b Ensure filename is written
Now that SPR-16307 makes it possible to write a part with a Publisher
we need to ensure we get the filename correctly when writing with
Publisher<Resource>.

Issue: SPR-16376
2018-01-19 22:12:53 -05:00
Rossen Stoyanchev aea6bb6357 Update STOMP docs on using dot as separator
Issue: SPR-16275
2018-01-19 21:51:31 -05:00
Rossen Stoyanchev 0fb31c5e36 Refine "." separator support for STOMP messaging
After this commit DefaultUserDestinationResolves no longer looks at
whether AntPathMatcher is configured with "." as separator and rather
expects to be explicitly told whether to keep the leading slash in
translated destinations which actually depends on what the message
broker supports (e.g. RabbitMQ "/", Artemis ".") or how it is
configured (simple broker could be either way).

There is also a minor improvement in SimpMessagingTemplate to ensure
user destinations are correctly formed based on what the
DefaultUserDestinationResolver expects. When using "." as separtor it
allows sending messages to "queue.q1" rather than "/queue.q1".

Issue: SPR-16275
2018-01-19 21:02:43 -05:00
Rossen Stoyanchev 17f9b61249 Polish MessageBrokerConfigurationTests 2018-01-19 21:02:43 -05:00
Rossen Stoyanchev afd248da8a MultipartHttpMessageWriter consumes source once only
The previous fix #09f1f7 did not actually address the issue but only
moved it further down, so instead of the subscribe(), it was consuming
it inside the MultipartHttpMessageWriter#write method which returned
this.body.then(), and then again for the actual request body writing.

In this commit MultipartHttpMessageWriter#write returns Mono.empty()
since we don't actually want to write the part content from there, but
only want to access it as soon as it is availabele, for writing to
the request body.

Issue: SPR-16402
2018-01-19 20:55:22 -05:00
Juergen Hoeller 572c668726 Polishing 2018-01-19 21:30:37 +01:00
Arjen Poutsma 5520e730f1 DataBufferUtils.read should not take input stream/channel as parameter
Fixed by creating `Callable`-based variants, as explained in the JIRA
issue.

Issue: SPR-16403
2018-01-19 17:38:14 +01:00
Rossen Stoyanchev 09f1f727a7 Remove explicit subscribe in MultipartHttpMessageWriter
Issue: SPR-16402
2018-01-19 10:41:08 -05:00
Juergen Hoeller 40127bd9ad Polishing 2018-01-19 12:30:10 +01:00
Juergen Hoeller 4c7414833b ResolvableType returns clone for cached state with original local source
Issue: SPR-16210
2018-01-19 12:30:03 +01:00
Rossen Stoyanchev 357fb48d13 Update docs on stomp.js library
Issue: SPR-15624
2018-01-18 13:09:40 -05:00
Rossen Stoyanchev 513461d4f1 Update STOMP overview
Issue: SPR-15624
2018-01-18 13:04:43 -05:00
Arjen Poutsma c53c8bfc5a Set 304 status on ServerResponse when ETag/LastModified match
This commit checks the Etag/LastModified headers on the incoming
request, and sets a 304 Not Modified status with no body when they
match, by delegating to ServerWebExchange.checkNotModified.

Issue: SPR-16348
2018-01-18 11:34:38 +01:00
Arjen Poutsma c211e3998b Polishing 2018-01-18 11:34:38 +01:00
Arjen Poutsma 61d330f805 Javadoc 2018-01-18 11:34:38 +01:00
Juergen Hoeller ea5f8f5a7c Consistently handle NoClassDefFoundError in BeanUtils and related places
Issue: SPR-16369
2018-01-17 20:24:57 +01:00
Rossen Stoyanchev 0c289283ff Separate from expectations from response creation
This commit separates the creation of a response so that it is executed
after the synchronized block inside which requests need to be matched
and validated (for order and count).

This allows a ResponseCreator to be slow or block if it has to.

Issue: SPR-16319
2018-01-17 13:01:40 -05:00
Rossen Stoyanchev 7ab4d0ca08 Polish MockRestServiceServer code 2018-01-17 13:01:40 -05:00
Juergen Hoeller 214576673a Proper NoClassDefFoundError check against BeanUtils.instantiateClass
Issue: SPR-16369
2018-01-17 18:29:05 +01:00
Juergen Hoeller 32b4279929 Upgrade to Kotlin 1.2.20 (and Undertow 1.4.22) 2018-01-17 18:28:26 +01:00
Stephane Nicoll 817254f2eb Merge pull request #1646 from JanStureNielsen:patch-2
* pr/1646:
  Fix doc typo
2018-01-17 18:07:04 +01:00
Jan Nielsen fed81c6f83 Fix doc typo
Closes gh-1646
2018-01-17 18:06:17 +01:00
Stephane Nicoll dfe4e01f51 Merge pull request #1645 from JanStureNielsen:patch-1
* pr/1645:
  Fix doc typo
2018-01-17 18:06:04 +01:00
Jan Nielsen df9761cb3f Fix doc typo
Closes gh-1645
2018-01-17 18:05:39 +01:00
Rossen Stoyanchev 12baafd4f0 WebFlux docs: security + consistent order for WebFlux config 2018-01-16 23:23:26 -05:00
Rossen Stoyanchev 1148b61dfa WebMVC Docs: remove old code-by-convention content
We cover this more succinctly from other places, e.g. in Model Methods
by referring to the Javadoc of Contentions, or pointing to the
RequestToViewNameTranslator.
2018-01-16 23:23:09 -05:00
Rossen Stoyanchev bb7152d6e7 WebFlux docs: matrix variables
Issue: SPR-16040
2018-01-16 22:53:40 -05:00
Rossen Stoyanchev 61fc321554 WebFlux docs: Request|ResponseBody + Http|ResponseEntity
Issue: SPR-16040
2018-01-16 22:10:04 -05:00
Rossen Stoyanchev 4f2e54fc2d Document multipart data support for WebFlux
Issue: SPR-16040
2018-01-16 16:53:19 -05:00
Rossen Stoyanchev b97fa4a5ee Update docs on multipart requests + table of parameters
Issue: SPR-16040
2018-01-16 11:33:27 -05:00
Juergen Hoeller 65a167f7fd JmsListenerAnnotationBeanPostProcessor properly deals with nested proxies
Issue: SPR-16196
2018-01-16 14:39:31 +01:00
Juergen Hoeller 0270808b3c Upgrade to EclipseLink 2.7.1 2018-01-16 14:39:14 +01:00
Timo Meinen 823a16ce17 Log SQL parameters in EclipseLink when using showSql
As of EclipseLink 2.4.0 - Juno this is not sufficient to log
SQL parameter binding. Additionally,
eclipselink.logging.parameters must be enabled.

Issue: SPR-16383
2018-01-16 14:20:49 +01:00
Sam Brannen 7b4ec72378 Upgrade to JUnit Jupiter 5.0.3
Issue: SPR-16384
2018-01-16 13:17:32 +01:00
Rossen Stoyanchev 437c33ba42 WebFlux @ModelAttribute coverage in reference
Issue: SPR-16040
2018-01-15 17:47:27 -05:00
sdeleuze 43d3abdfd5 Fix SockJs CorsConfiguration for forbidden origins
After this commit, AbstractSockJsService uses the configured allowed
origins when generating the CorsConfiguration instead of "*".

As a consequence, forbidden origin requests still result in a 403
response but now with no CORS headers in order to improve consistency
between the status code and the headers.

Issue: SPR-16304
2018-01-15 18:07:24 +01:00