With this commit, WebFlux server uses warning instead of error log level
for request handling, and also just print the message instead of the
stacktrace which is mostly meaningless in reactive world.
Complementary to this change, Reactor Netty removed additional logging
as part of https://github.com/reactor/reactor-netty/issues/339.
Issue: SPR-16688
ServerSentEventHttpMessageReader had logic to split on new lines
and buffer until an empty new line (start of a new event). To account
for random data chunking, it later re-assembled the lines for each
event and split again on new lines. However bufferUntil was still
unreliable a chunk may contain nothing but a newline, which doesn't
necessarily mean an empty newline in the overall SSE stream.
This commit simplifies the above by delegating the splitting of the
stream along newlines to StringDecoder.
Issue: SPR-16744
Prior to this commit, the various `HttpMessageConverter` instances
configured for a given `RestTemplate` instance could all contribute
`MediaType` values to the "Accept:" request header.
This could lead to duplicate media types in that request header,
cluttering for the HTTP request for no reason.
This commit ensures that only distinct values are added to the request.
Issue: SPR-16690
Using the simple example shown in the ticket but switching from
Mono<String> to Flux<String> (and 5,000,000 onNext calls) shows that
constant pausing causes significant overhead and is not worth the
trouble vs ignoring the onWritePossible in REQUESTED state.
Issue: SPR-16702
Undertow does not provide a way to check if we can write so with the
current implementation of isWritePossible, deep recursion can occur
when writing slows down. We now use a flag to keep track of write
ChannelListener callbacks.
This commit also addresses a related issue in
AbstractListenerWriteProcessor that went undected since #3c2d186
where after a large (single) buffer that is not written fully, the
completion signal is processed before the all data is written.
Issue: SPR-16702
This commit documents the difference between configuring the socket
timeout on the `RequestConfig` and on the `SocketConfig`.
The first one does not affect timeouts when establishing an SSL
connection or sending a CONNECT request to a proxy. For these use cases,
it is required to configure `SocketConfig` on the `HttpClient` instance
directly.
Issue: SPR-16697
This breaks the package dependency cycle between web.server/web.method and makes ServerErrorException more generally applicable. Includes deprecation of the plain reason constructor variant, in favor of providing a Method or MethodParameter context (which MatrixVariableMethodArgumentResolver does now).
Consistently return "*/*" if no media types were requested rather than
an empty list. Existing code has to check for both in any case to see
if nothing was requested.
Issue: SPR-16624
The web.server package is quite low-level and should not depend on web.bind in order to avoid a dependency cycle. Extracting the introspection of the ResponseStatus annotation into a WebFlux-level subclass resolves the cycle.
Issue: SPR-16567
Undertow does not provide a way to check if data is available to read
but instead we have to try to read and see if any data is returned.
This makes it impossible to implement checkOnDataAvailable without
trying to read and that can lead to infinite recursion like this:
...
UndertowServerHttpRequest$RequestBodyPublisher.checkOnDataAvailable(UndertowServerHttpRequest.java:156)
AbstractListenerReadPublisher.changeToDemandState(AbstractListenerReadPublisher.java:177)
AbstractListenerReadPublisher.access$900(AbstractListenerReadPublisher.java:47)
AbstractListenerReadPublisher$State$4.onDataAvailable(AbstractListenerReadPublisher.java:319)
AbstractListenerReadPublisher.onDataAvailable(AbstractListenerReadPublisher.java:85)
UndertowServerHttpRequest$RequestBodyPublisher.checkOnDataAvailable(UndertowServerHttpRequest.java:156)
This commit prevent the call to checkOnDataAvailable() when switching
states from READING->DEMAND which implies we exited the readAndPublish
loop because there was no more data to read.
Issue: SPR-16545
Sebastien Deleuze
Juergen Hoeller
Rossen Stoyanchev
Johnny Lim
Brian Clozel
Stephane Maldini
igor-suhorukov