Red Hat Training
A Red Hat training course is available for Red Hat Decision Manager
6.4. Executable rule models
Executable rule models are embeddable models that provide a Java-based representation of a rule set for execution at build time. The executable model is a more efficient alternative to the standard asset packaging in Red Hat Decision Manager and enables KIE containers and KIE bases to be created more quickly, especially when you have large lists of DRL (Drools Rule Language) files and other Red Hat Decision Manager assets. The model is low level and enables you to provide all necessary execution information, such as the lambda expressions for the index evaluation.
Executable rule models provide the following specific advantages for your projects:
-
Compile time: Traditionally, a packaged Red Hat Decision Manager project (KJAR) contains a list of DRL files and other Red Hat Decision Manager artifacts that define the rule base together with some pre-generated classes implementing the constraints and the consequences. Those DRL files must be parsed and compiled when the KJAR is downloaded from the Maven repository and installed in a KIE container. This process can be slow, especially for large rule sets. With an executable model, you can package within the project KJAR the Java classes that implement the executable model of the project rule base and re-create the KIE container and its KIE bases out of it in a much faster way. In Maven projects, you use the
kie-maven-plugin
to automatically generate the executable model sources from the DRL files during the compilation process. -
Run time: In an executable model, all constraints are defined as Java lambda expressions. The same lambda expressions are also used for constraints evaluation, so you no longer need to use
mvel
expressions for interpreted evaluation nor the just-in-time (JIT) process to transform themvel
-based constraints into bytecode. This creates a quicker and more efficient run time. - Development time: An executable model enables you to develop and experiment with new features of the decision engine without needing to encode elements directly in the DRL format or modify the DRL parser to support them.
For query definitions in executable rule models, you can use up to 10 arguments only.
For variables within rule consequences in executable rule models, you can use up to 12 bound variables only (including the built-in drools
variable). For example, the following rule consequence uses more than 12 bound variables and creates a compilation error:
... then $input.setNo13Count(functions.sumOf(new Object[]{$no1Count_1, $no2Count_1, $no3Count_1, ..., $no13Count_1}).intValue()); $input.getFirings().add("fired"); update($input);
6.4.1. Embedding an executable rule model in a Maven project
You can embed an executable rule model in your Maven project to compile your rule assets more efficiently at build time.
Prerequisite
You have a Mavenized project that contains Red Hat Decision Manager business assets.
Procedure
In the
pom.xml
file of your Maven project, ensure that the packaging type is set tokjar
and add thekie-maven-plugin
build component:<packaging>kjar</packaging> ... <build> <plugins> <plugin> <groupId>org.kie</groupId> <artifactId>kie-maven-plugin</artifactId> <version>${rhdm.version}</version> <extensions>true</extensions> </plugin> </plugins> </build>
The
kjar
packaging type activates thekie-maven-plugin
component to validate and pre-compile artifact resources. The<version>
is the Maven artifact version for Red Hat Decision Manager currently used in your project (for example, 7.18.0.Final-redhat-00002). These settings are required to properly package the Maven project.注記Instead of specifying a Red Hat Decision Manager
<version>
for individual dependencies, consider adding the Red Hat Business Automation bill of materials (BOM) dependency to your projectpom.xml
file. The Red Hat Business Automation BOM applies to both Red Hat Decision Manager and Red Hat Process Automation Manager. When you add the BOM files, the correct versions of transitive dependencies from the provided Maven repositories are included in the project.Example BOM dependency:
<dependency> <groupId>com.redhat.ba</groupId> <artifactId>ba-platform-bom</artifactId> <version>7.3.0.GA-redhat-00002</version> <scope>import</scope> <type>pom</type> </dependency>
For more information about the Red Hat Business Automation BOM, see What is the mapping between RHDM product and maven library version?.
Add the following dependencies to the
pom.xml
file to enable rule assets to be built from an executable model:-
drools-canonical-model
: Enables an executable canonical representation of a rule set model that is independent from Red Hat Decision Manager -
drools-model-compiler
: Compiles the executable model into Red Hat Decision Manager internal data structures so that it can be executed by the decision engine
<dependency> <groupId>org.drools</groupId> <artifactId>drools-canonical-model</artifactId> <version>${rhdm.version}</version> </dependency> <dependency> <groupId>org.drools</groupId> <artifactId>drools-model-compiler</artifactId> <version>${rhdm.version}</version> </dependency>
-
In a command terminal, navigate to your Maven project directory and run the following command to build the project from an executable model:
mvn clean install -DgenerateModel=<VALUE>
The
-DgenerateModel=<VALUE>
property enables the project to be built as a model-based KJAR instead of a DRL-based KJAR.Replace
<VALUE>
with one of three values:-
YES
: Generates the executable model corresponding to the DRL files in the original project and excludes the DRL files from the generated KJAR. -
WITHDRL
: Generates the executable model corresponding to the DRL files in the original project and also adds the DRL files to the generated KJAR for documentation purposes (the KIE base is built from the executable model regardless). -
NO
: Does not generate the executable model.
Example build command:
mvn clean install -DgenerateModel=YES
-
For more information about packaging Maven projects, see Packaging and deploying a Red Hat Decision Manager project.