Clone
 

petr široký <siroky.petr@gmail.com> in Drools

[DROOLS-2152] fixes after mockito upgrade (#1628)

[DROOLS-1776] add Automatic-Module-Name MANIFEST attribute

  1. … 31 more files in changeset.
Remove unsused FilteredPatternLayout (#1271)

When upgrading to logback 1.2.0, this class gets broken. No logback-*.xml

file (that I was able to find) is referencing it.

Remove unsused FilteredPatternLayout (#1271)

When upgrading to logback 1.2.0, this class gets broken. No logback-*.xml

file (that I was able to find) is referencing it.

(cherry picked from commit e40621441b4b6958e6030854a8f348095cef84f2)

Don't refer to non existing findbugs config file (#1094)

Remove bogus source file (#1079)

[RHBPMS-4435] make sure scanner state is always consistent (#1047)

Before this change it was possible to "break" the consistency

of the scanner, in case the polling interval was very small (20ms).

It was because of the unsynchronized method run() in the ScanTask.

It was possible to schedule new run() method (by the Timer)

during the call to stop(). The run() would stop at scanNow()

(as that is synchronized), but would be executed after the stop()

method finished. So the scanner would be running one last time

even after it was stopped. This then lead to inconsistent scanner

status as the run() method changed it to RUNNING after the scanNow()

finished.

With this fix, the scanNow() + status=RUNNING is synchronized via

the scanner lock making sure it can not be called in the middle of

the stop() method.

[DROOLS-1390] make sure scanner state is always consistent (#1046)

Before this change it was possible to "break" the consistency

of the scanner, in case the polling interval was very small (20ms).

It was because of the unsynchronized method run() in the ScanTask.

It was possible to schedule new run() method (by the Timer)

during the call to stop(). The run() would stop at scanNow()

(as that is synchronized), but would be executed after the stop()

method finished. So the scanner would be running one last time

even after it was stopped. This then lead to inconsistent scanner

status as the run() method changed it to RUNNING after the scanNow()

finished.

With this fix, the scanNow() + status=RUNNING is synchronized via

the scanner lock making sure it can not be called in the middle of

the stop() method.

Upgrade JPA (2.0 -> 2.1) + Hibernate (4.x -> 5.x)

    • -0
    • +2
    /drools-persistence-jpa/src/test/resources/bitronix-default-config.properties
Fixes after JPA 2.0 -> 2.1 upgrade (#1019)

Fix and clean-up POMs (#11)

[DROOLS-1320,RHBRMS-2636] fix org.kie.scanner import package name (#951)

[DROOLS-1320] fix org.kie.scanner import package name (#949)

[DROOLS-898] org.drools.osgi -> org.kie.osgi (#902)

[DROOLS-1255] remove some of the java.beans usage (#879) (#882)

* there are still few instances remaining which can

not be just removed, but we need proper replacement

for those

  1. … 15 more files in changeset.
[DROOLS-1255] remove some of the java.beans usage (#879)

* there are still few instances remaining which can

not be just removed, but we need proper replacement

for those

  1. … 14 more files in changeset.
[DROOLS-1212,RHBRMS-2501] properly close streams/resources (#876)

* also fixed small typo in a method name

Don't report turtle tests as skipped (#845)

* reporting the thousands of generated tests

as skipped skews the overall "skipped tests"

statistics which is very bad

* before the change 25k tests were reported as skipped

* turtle tests can be executed using specific property, e.g.

`mvn clean test -DrunTurtleTests`

Don't report turtle tests as skipped (#844)

* reporting the thousands of generated tests

as skipped skews the overall "skipped tests"

statistics which is very bad

* before the change 25k tests were reported as skipped

* turtle tests can be executed using specific property, e.g.

`mvn clean test -DrunTurtleTests`

Don't report turtle tests as skipped (#837)

* reporting the thousands of generated tests

as skipped skews the overall "skipped tests"

statistics which is very bad

* before the change 25k tests were reported as skipped

* turtle tests can be executed using specific property, e.g.

`mvn clean test -DrunTurtleTests`

[DROOLS-1220] configure NativeJavaCompiler using the provided settings (#824)

* only source+target levels are configured right now (via the

JavaDialectConfiguration). More options can be added later

if needed

[DROOLS-1213] do not assume getClass().getClassLoader() is an instance of URLClassLoader (#822)

[DROOLS-1216] exclude commons-logging and use jcl-over-slf4j instead (#815)

[DROOLS-1213] restore system properties after test finishes (#819)

* property "drools.warning.filters" is no longer being used

[DROOLS-1213] drools-pmml: explicitly depend on EE deps to make it work on JDK9 (#816)

* works both on JDK8 and JDK9

* the drools-pmml needs to transitively bring the jaxb

and javax.activation deps, otherwise projects depending

on drools-pmml (e.g. drools-scorecards) still won't work

[DROOLS-1212] use try-with-resources to properly close streams (#817)

* also fixed small typo in a method name

[DROOLS-1170] drools-pmml: use 'xjc' binary instead of jaxb2-maven-plugin (#814)

* jaxb2-maven-plugin is not yet supported on JDK9. The original

idea was to create new profile which would be jdk9-specific

and use the exec-maven-plugin only there. However, this does not

work because the classes generated directly by 'xjc' and the

plugin are not exactly the same. The one big difference is that

the plugin useis 'is' prefix for Boolean properties whereas

xjc will use 'get' prefix (e.g. public Boolean isAdult() vs

public Boolean getAdult()). It seems the later is correct as

'is' should only be used for primitive boolean, and not for object

type Boolean (http://download.oracle.com/otndocs/jcp/7224-javabeans-1.01-fr-spec-oth-JSpec/).

To make the configuration easier, the plugin was removed

and the jxc is called directly.

Fixes after IP BOM 7.0.0.CR4 upgrade (#813)

Remove no longer valid 'Java5Test' (#811)

kie-parent now brings all 3rd party deps (#793)