On start.spring.io, if you customize the artifactId it creates a zip file
with the same name. The `spring init` command did not have a similar
shortcut.
This commit updates the request to customize the artifactId if none is
set and a custom location was specified. Just as we check for the
presence of a dot to figure out if we have to extract the archive or not,
we check for it to generate an artifactId without an extension.
In practice, `spring init foo` creates a foo directory with a project
whose artifactId is `foo` and `spring init foo.zip` stores a foo.zip
file with the same project (i.e. the artifactId is `foo`).
Closes gh-3714
The `spring.datasource.name` property was hidden behind the 'name'
attribute of the Tomcat connection pool (since we are mapping all
datasource implementations on the `spring.datasource` namespace.
This commit replace the injected value by hand with the use of the
regular `DataSourceProperties`. That way, we generate proper meta-data
for it as well.
Closes gh-3755
A new `spring.datasource.type` property can now be used to configure the
connection pool implementation to use (rather than only relying on Boot's
preferences).
Closes gh-3705
Improve the documentation to explain the necessary steps to create a
custom Spring Boot starter. In particular, provide more details regarding
naming conventions.
Closes gh-2537
See gh-2927
Previously, to use a custom ConfigurableWebBindingInitializer, it was
necessary to extend WebMvcConfigurationSupport and override
getConfigurableWebBindingInitializer. This had the unwanted
side-effect of switching off the auto-configuration of Spring MVC.
This commit updates the auto-configuration to look for a
ConfigurableWebBindingInitializer bean and register it with Spring
MVC.
Closes gh-2526
Previously, PropertiesConfigurationFactory would only create a
DefaultPropertyNamePatternsMatcher that ignored case if it was
ignoring unknown fields. If the binding had a target name and
unknown fields were not being ignored the matcher would consider the
case when finding matches. This meant that SERVER_PORT would not being
to ServerProperties.port as SERVER did not match the target name,
server.
This commit updates PropertiesConfigurationFactory to use a
case-ignoring DefaultPropertyNamePatternsMatcher when binding with
a target name. The tests have also been updated to test binding both
with and without ignoring case and using either Properties or
PropertySources. The above-described matching is only performed
against PropertySources and the bug only occurred when using a target
name, making it important to test all combinations.
Closes gh-3745
Following the changes made to combines the /links and /hal endpoints
into a single /actuator endpoint, a web browser accessing /actuator
would receive a 406 response if HAL browser was not on the classpath.
This commit updates the /actuator main entry point so that it will
serve JSON to a web browser when HAL browser is not on the classpath.
The actuator's embedded documentation has also been updated to reflect
the recent changes.
Closes gh-3696
The annotation processor detects `@ConfigurationProperties` bean or
method definition and merges manual meta-data. The former step will fail
with a NPE if the annotation is not present on the classpath. This could
happen if the annotation processor is added to a module that is not
actually using Spring Boot.
We now have a defensive check that skips that steps but still attempts to
merge manual meta-data if present.
Closes gh-3720
This commit provides a single endpoint, /actuator, that serves HTML
(the HAL browser) or JSON depending on the request’s accept header
that enables discovery of all of the actuator’s other endpoints.
When the management context path is configured, the /actuator endpoint
moves to the configured path, e.g. if the management context path is
set to /management, the actuator endpoint will be available from
/management.
Closes gh-3696
