This commit defers the resolution of the default ddl auto mode only when
it is absolutely necessary. This prevents Spring Boot to attempt to get
a connection when it isn't necessary
Closes gh-12374
Previously, when performing lazy initialisation of the context,
ApplicationContextRequestMatcher assigned the context field before it
called initialized. The context being non-null is used as the signal
that it’s ok to call a subclass’s matches method. If one thread checks
for a non-null context in between the field being assigned and
initialized being called on another thread, matches will be called
before the subclass is ready.
This commit closes the window for the race condition by only assigning
the context field once the subclass’s initialized method has been
called.
There is a secondary problem in each of the subclasses. Due to the use
of double-checked locking in ApplicationContextRequestMatcher, it’s
possible for a subclass’s matches method to be called by a thread that
has not synchronised on the context lock that’s held when initialized
is called and the delegate field is assigned. This means that the
value assigned to the field may not be visible to that thread.
This commit declares the delegate field of each
ApplicationContextRequestMatcher subclass as volatile to ensure that,
following initialisation, its value is guaranteed to be visible to
all threads.
Closes gh-12380
This commit detects if a `ConnectionNameStrategy` bean exists in the
context and associates it with the auto-configured RabbitMQ's
`ConnectionFactory` when that is the case.
Closes gh-12367
With JDK 8 being the baseline and JDK 7 not being supported anymore we
can get rid of the workaround for a JDK 7 bug in
ProcessBuilder.inheritIO on Windows machines.
Closes gh-12337
Previously, the JDK 10 build would fail as we build the plugin using
Gradle 4.0.x (the lowest version of Gradle that we support) and
Gradle 4.0.x doesn't work with Java 10.
Upgrading to Gradle 4.1, which appears to work with Java 10, was
considered but rejected for now as it introduces the risk that we
inadvertently use an API that's new in 4.1 and break our 4.0 support.
This commit goes for the extreme option and disables building the
Gradle Plugin when building with JDK.
See gh-12028
Spring Integration no longer requires a `MicrometerMetricsFactory` to
support Micrometer metrics so there's nothing for Boot to
auto-configure.
See gh-12287
Prior to this commit, the `ApplicationContext` couldn't start with a
JDK-proxied `HikariDataSource` as the JMX auto-configuration was
attempting to inject a (too narrowed) `HikariDataSource`.
This commit rather injects a regular `DataSource` and attempt to unwrap
it as a `HikariDataSource`.
Closes gh-12271
Reactor Netty 0.7.5 now supports all Compression options.
Unlike previous Reactor Netty versions, the minimum compression
threshold is now based on the content-length instead of
accumulating for a while the data.
See gh-12268
Update `EndpointRequest` to that the `PathMappedEndpoints` bean is
optional. A missing bean is treated as if there are no path mapped
endpoints.
Fixes gh-12238
Update `ApplicationContextRequestMatcher` and
`ApplicationContextServerWebExchangeMatcher` to use a supplier for
the context, rather than the context itself.
This allow exceptions to be propagated to subclasses which may choose
to deal with them.
See gh-12238
Update `spring-boot-actuator-autoconfigure` so that it no longer has
a dependency on `spring-boot-test-autoconfigure`. This will allow us to
add test support for actuator concerns in the future if we need to.
Fixed gh-12270
This commit simplifies the HTTP compression configuration for Reactor
Netty servers.
Also, this commit removes a test for the
`server.compression.min-response-size` support, as this is only
supported when the HTTP response contains a `Content-Length` header.
Since most Spring WebFlux responses are using
`Transfer-Encoding: chunked`, we should not test for that case.
See gh-12268
Previously, if the project's group, name, or version changed the
BuildInfo task would still be considered up-to-date as the values of
the project's properties were not reflected in the fields of the
BuildInfo instance.
This commit updates BuildInfo to copy the value of the project's
property to the corresponding BuildInfo field when the property is
read using its getter method on BuildInfo.
Closes gh-12266
Previously, if a user declared a custom conversionService bean that
was not an ApplicationConversionService instance, the binder lost
the ability to convert a String to a Duration (along with any other
conversions that are specific to ApplicationConversionService).
This commit updates BindConverter so that, if the ConversionService
with which it is created is not an ApplicationConversionService, it
will use one as an additional service when performing conversion.
Closes gh-12237
Previously, working with a JarFile created a large amount of garbage
that was allocated on the thread local allocation buffer (TLAB).
The TLAB allocations made a significant contribution to GC pressure
and slowed down startup. This commit reduces the amount of garbage
by making a number of changes.
Reading from a RandomAccessDataFile has been reworked to avoid
creating new RandomAccessFile instances. A single RandomAccessFile
is now created for an entire jar file and it is used to read data from
anywhere in that jar file, including entries in nested jar files. To
ensure that reads remain thread-safe, a lock is taken on the
RandomAccessFile that is shared by all RandomAccessDataFile instances
that are provided access to (portions of) the same jar file.
Reading all of the bytes from a RandomAccessData has been reworked to
avoid the use of an InputStream that was created, used to read the
data, and then thrown away. In place of the InputStream-based
mechanism a method has been introduced that returns all of the
RandomAccessData as a byte[]. Building on this change, a method has
also been introduced to read a portion of a RandomAccessData as a
byte[]. This avoids the need to create a new RandomAccessData
subsection where the subsection was only used to read its entire
contents and then thrown away.
Decoding of an MS-DOS datetime has been reworked to use LocalDataTime
rather than GregorianCalendar. The former produces less garbage than
the latter.
Closes gh-12226
Update `PropertySourceLoader` so that it no longer needs to deal with
matching multi-document files using the `spring.profile` property. The
loader now simply returns one or more `PropertSource` instances for a
given `Resource`.
All property matching now occurs in the `ConfigFileApplicationListener`.
This allows document processing logic to be contained in a single place,
and allows us to rationalize the algorithm so that negative matching
profiles are processed last.
Fixes gh-12159
Update `ValidationBindHandler` so that validation only occurs when a
value is bound, or the bind depth is zero. This prevents validation from
triggering getters which may throw an exception.
Fixes gh-12227
This commit changes invocations to immediately return the expression
instead of assigning it to a temporary variable. The method name should
be sufficient for callers to know exactly what will be returned.
Closes gh-12211
Exclude `FileEditor` from the `BindConverter` since it uses slightly
unusual logic to create the file. Specifically, given a value of "."
the editor will locate a `ClassPathResource` then return `getFile()`
from that resource. For back-compatibility, binding should use the
simpler conversion service logic.
Fixes gh-12163
Update `BindConverter` so that delegation to `SimpleTypeConverter` also
works with Collections and Arrays.
Prior to this commit, conversion that relied on a `PropertyEditor` would
only work for simple types. For example, "String -> Class<?>" would use
the `ClassEditor` but "String -> List<Class<?>>" would fail.
The `BindConverter` now uses a minimal `ConversionService` as an adapter
to the `SimpleTypeConverter`. This allows us to use the same delimited
string conversion logic as the `ApplicationConverter`.
Fixes gh-12166
This commit ensures that Tomcat is the first reactive server configured
if the Tomcat dependency is on classpath.
Spring Boot chose Reactor Netty as the default for the reactive web
starter, but the Reactor Netty dependency can be used also for its HTTP
client. In case developers are adding Tomcat, Undertow or Jetty on their
classpath, we must configure those and consider Reactor Netty for the
client only.
Fixes gh-12176
Drop `SpringProfileDocumentMatcher` and replace it with two new
implementations that restrict when YAML documents are loaded. YAML
sections are now restricted both on the specific profile that is being
loaded, and the profiles that are currently accepted.
The `PropertySourceLoader` interface has been refined to include a
predicate that determines when a profile is accepted. The
`ConfigFileApplicationListener` simply delegates the predicate logic to
the `Environment`.
Fixes gh-8011
Update the configurable embedded web server factory interfaces to
extend `ConfigurableWebServerFactory` so that the can be used in a
`WebServerFactoryCustomizer`.
Extract server specific customization to their own auto-configuration
and align reactive/servlet server auto-configuration.
Closes gh-8573
This commits move the Hikari metrics registration to a BeanPostProcessor
as the Hikari datasource cannot be modified once its configuration has
been sealed (usually happens when `getConnection` is invoked on the
pool).
Closes gh-12129
HikariCP has metrics integration with Micrometer. This configures all
`HikariDataSource` beans with the `MicrometerMetricsTrackerFactory` if a
`MeterRegistry` is available.
See gh-12129
Create a new `ApplicationConversionService` similar in design to the
DefaultFormattingConversionService from Spring Framework. The new
conversion service provides a central place for custom conversion logic
supported by Spring Boot.
Also replace the `BindingConversionService` with an internal
`BindConverter` class that now invokes the `SimpleTypeConverter`
directly. Binding for `@ConfigurationProperties` has been updated so
that any custom property editors registered with the BeanFactory can
be used.
Fixes gh-12095
Move `XADataSourceWrapper` to `jdb` and `XAConnectionFactoryWrapper` to
`jms` so that they are packaged in a similar way to the underlying XA
types.
Fixes gh-12061
Move leading slash logic from `MetricsClientHttpRequestInterceptor` to
`RestTemplateExchangeTags` so that URI based calls are also managed.
Closes gh-12126
This commit adds micrometer's default values for each monitoring
system with tests that validate those defaults are consistent. This
makes sure that those defaults are part of the metadata and any tooling
support that uses it.
Closes gh-12089
Previously, if a controll endpoint included a mapping with no path,
it would be mapped to the root of the application rather than the
root of the endpoint's path.
This commit updates the handler mappings for both reactive and Servlet
web applications to treat a mapping with no paths as being a mapping
with a single empty path. This ensures that a path with the endpoint's
path prepended is used for the mapping.
Closes gh-12104
Refactor `ConfigurationPropertiesBindingPostProcessor` to allow JSR-303
validation on `@ConfigurationProperties` defined at the `@Bean` method
level.
JSR-303 validation is now applied when a JSR-303 implementation is
available and `@Validated` is present on either the configuration
properties class itself or the `@Bean` method that creates it.
Standard Spring validation is also supported using a validator bean
named `configurationPropertiesValidator`, or by having the configuration
properties implement `Validator`.
The commit also consolidates tests into a single location.
Fixes gh-10803
This commit fixes two issues in `TestRestTemplate`:
* it improves the detection of the underlying request factory, using
reflection to look inside the intercepting request factory if
interceptors were configured
* it avoids reusing the same request factory when creating a new
`TestRestTemplate` with `withBasicAuth`. Sharing the same instance would
result in sharing authentication state (HTTP cookies). Since the
original request factory can't be detected consistently, a new one is
selected automatically
See gh-8697
Previously `TestRestTemplate` would override the configured
`ClientHttpRequestFactory` if the Apache HTTP client library was on
classpath.
This commit fixes two issues:
1. The existing `ClientHttpRequestFactory` is overridden *only* if it is
using the Apache HTTP client variant, in order to wrap it with the
`TestRestTemplate` custom support
2. Calling `withBasicAuth` will no longer directly use the request
factory returned by the internal `RestTemplate`; if client interceptors
are configured, the request factory is wrapped with an
`InterceptingClientHttpRequestFactory`. If we don't unwrap it,
interceptors are copied/applied twice in the newly created
`TestRestTemplate` instance. For that, we need to use reflection as the
underlying request factory is not accessible directly.
Closes gh-8697
Andy Wilkinson
Stephane Nicoll
seongwoon.lee
Madhura Bhave
inabajunmr
Cristian Greco
Jon Schneider
dreis2211
Johnny Lim
igor-suhorukov
Phillip Webb
Gary Russell
Brian Clozel
Stephane Maldini
Kedar Joshi
Arnaud Heritier
Raja Kolli
Nikola Kološnjaji
Tommy Ludwig
Simon Braconnier