Update `AutoConfigureAnnotationProcessor` to ensure that the generated
properties file is fully repeatable. Properties are now sorted and
written out directly to ensure that the timestamp comment is not
present.
Closes gh-19370
- Previously, we would create a JarFileArchive for all nested jars.
This was an additional overhead. We only need to create a JarFileArchive
for jars that can have nested jars in them. For all other jars we only need
the URL to build the classpath.
- While iterating over nested entries in the exploded jar, we only need to
look at BOOT-INF and we can skip any entry that does not match that.
Closes gh-16655
Co-authored-by: Phillip Webb <pwebb@pivotal.io>
Previously, reflective access to the archiveBaseName property
incorrectly treated the property as a String. It should have been
treated as a Property<String>. This caused an exception to be thrown
and the deprecated baseName property to be used as a fallback.
This commit corrects the reflective access to the archiveBaseName
property. It also updates the tests to fail if a build outputs a
deprecation warning. Tests that use Gradle's Maven plugin have been
updated to expect deprecation warnings when run with Gradle 6.0 where
the plugin is deprecated. Tests that configure an archive's base name
have been updated to use archiveBaseName when running with Gradle 6.0
and later.
Closes gh-18663
Previously, our Gradle plugin was not tested against Gradle 6.0,
a number of deprecation warnings were output when using the plugin
with Gradle 6, and some functionality related to the application
plugin did not work as expected.
This commit tests the plugin against Gradle 6. It also avoids calling
deprecated APIs. The plugin is compatibile against Gradle 4.10 where
the deprecated APIs' replacements are not available so reflection is
used to call the replcaements. Lastly, the way in which the base name
of the boot distribution that is created when the application plugin
is applied has been modified to ensure that it is effective when using
Gradle 6.
Closes gh-18663
Previously, when RunProcess handled a SIGINT it would immediately
attempt to destroy the process that it had run. This created a race
condition between the SIGINT being handled by the child process
and RunProcess destroying the child. The exact behavior of destroy
is implementation dependent and it may result in forcible termination
of the process where shutdown hooks are not called. This is what
happens on Windows. The exit code in such a case is 1 which prevents
anything from waiting for the process to complete from detecting
that it ended as a result of a SIGINT, leaving it with no choice but
to report an error. This is what happens with mvn spring-boot:run
with a forked process on Windows and results in the build failing.
This commit updates RunProcess to allow the child process to handle
the SIGINT itself, waiting for up to five seconds for that to happen
before the process is then destroyed. Given this time, the child
process exits with 130 which RunMojo already handles correctly as
indicating that the process died due to SIGINT and the build completes
with success as a result.
Fixes gh-18936
Previously, when the configuration properties annotation processor
encountered a property that was the same as an outer type that had
already been processed, it would fail with a stack overflow error.
This commit introduces the use of a stack to track the types that
have been processed. Types that have been seen before are skipped,
thereby preventing a failure from occurring. We do not fail upon
encountering a recursive type to allow metadata generation to
complete. At runtime, the recursive property will not cause a problem
if it is not bound.
Fixes gh-18365
Update `AutoConfigureAnnotationProcessor` to no longer store
`@Configuration.value` in the meta-data JSON since we never actually
read it.
Closes gh-16608
Update the configuration processor to use the newly introduced
`@ConstructorBinding` annotation to determine when meta data
should be generated from constructor parameters.
Prior to this commit, the processor had no good way to tell when
constructor parameters should be used instead of getters/setters.
Closes gh-17035
Previously, Maven's default behaviour was relied up which resulted
in the artifact ID being appended to each URL as it was inherited.
This behaviour can only be disabled in Maven 3.6 and later, a version
that we cannot use due to an incompatibility with the Flatten Plugin.
This commit works around Maven's default behaviour by defining
properties for the SCM URL, connection, and developer connection and
then explicitly defining the settings in each pom using these
properties. The explicit definition of the properties in each pom
prevents them being inherited from the parent, thereby disabling the
unwanted appending of the artifact ID to the URL.
Fixes gh-18328
This commit makes sure that a ConfigurationProperties type contributed
by a `@Bean` factory method uses properties binding regardless of the
presence of a matching constructor.
`@Bean` method makes sure the user is in control and will be responsible
of creating the instance. As a result, binding of properties will not
happen there and therefore can only happen with regular JavaBean
accessors.
Closes gh-18184
Previously, calling getComment() on a nested jar file would result
in the outer jar file's comment being returned.
This commit updates the loader's JarFile to read the file's comment
from the central directory end record and return it from getComment().
Fixes gh-18128
Previously, the configuration processor would ignore any
@ConfigurationProperties-annotated methods that were not public. This
prevented metadata generation for package-private @Bean methods such
as those in DataSourceConfiguration's inner-classes for DBCP2, Hikari,
and Tomcat JDBC.
This commit updates the annotation processor so that it will process
any non-private method annotated with @ConfigurationProperties.
Fixes gh-18124
This commit prevents a potential NPE if the startTime of the
MavenSession is not available and fallbacks to the current time. This
can happen when invoking the plugin with Maven embedded in an IDE.
Closes gh-17810
Update `BuildInfoMojo` so that the time property now defaults to
`${session.request.startTime}` rather than the time the Mojo was
created. Also update javadoc to make it clear that any supplied
value will be passed to `Instant.parse`.
See gh-17390
Meta-annotate `ClassPathExclusions` and `ClassPathOverrides` with
so that the `ModifiedClassPathExtension` no longer needs to be
used directly.
See gh-17491
Replace any direct `junit-platform-launcher` dependencies and instead
rely on the test runner providing it. Launcher related class are not
handled via reflection.
This update allows us to workaround SUREFIRE-1679.
Closes gh-17517
Add a JUnit 5 extension that allows tests to be run with a
modified classpath. Since JUnit 5 does not currently offer a way
to run tests with a different classpath, we instead fake the
original invocation and launch an entirely new run for each
method.
See gh-17491
Apply checkstyle rule to ensure that private and package private
classes do not have unnecessary public methods. Test classes have
also been unified as much as possible to use default scoped
inner-classes.
Closes gh-7316
Previously, a regular expression of /./ was used to replace /./ with
/. The '.'' in the expression matches any single character so the
replacement was more broadly applicable than it should have been. For
example, /a/ would be replaced with /.
This commit uses Pattern.LITERAL to compile the regular expression
from the CURRENT_DIR (/./) contant. This allows the constant to be
used to check for occurances of /./ in the string before attempting
replacement, while also ensuring that the '.' is treated literally.
Closes gh-17341
Update Gradle archive tasks to ensure that `META-INF/` and
`META-INF/MANIFEST.MF` remain as the first entries of the archive.
Prior to this commit, rewritten archives would violate the implicit
specification of `JarInputStream` that these entries should be first.
Fixes gh-16698
Since the move to JUnit 5, a number of tests were failing on Windows.
The majority were failing due to open file handles preventing the
clean up of the tests' temporary directory. This commit addresses
these failures by updating the tests to close JarFiles, InputStreams,
OutputStreams etc.
A change has also been made to CachingOperationInvokerTests to make
a flakey test more robust. Due to System.currentTimeMillis() being
less precise on Windows than it is on *nix platforms, the test could
fail as it would not sleep for long enough for the TTL period to have
expired.
Stephane Nicoll
Dmytro Nosan
如梦技术
Phillip Webb
Andy Wilkinson
dreis2211
Madhura Bhave
Johnny Lim
Alessandro Falappa
Leo Li
Kant Leung
Yuyan
Brian Clozel
Wagner Macedo
Camille Vienot
thelproad
SaberXu
Dave Syer
vivganes
Marc Rohlfs
Andreas Gebhardt
gaurav bhatnagar