Page History

When developing an application with Saga Library embedded, and connecting to the ElasticSearch manage by the Saga Server

Note

This tutorial assumes:

The reader ability to create a project with Maven Framework support
The data Saga will use is manage through the Saga's user interface

Step-by-step guide

Table of Contents

maxLevel	3
minLevel	3
outline	true

Configure pom.xml

In the pom.xml of the project, a basic example of the minimum configuration can be found below. But elementary section of this configuration is the dependencies section, where we need two main libraries.

saga-library	saga-elastic-provider
The core library of Saga, this dependency includes the Engine, Stages, Tag Manager, Pipeline Manager and Resource Manager which are all the parts necessary to use Saga in any application.

saga-elastic-provider

This dependency will grant us access to ElasticSearch as a provider for Saga, which means our Stages and Managers will be able to fetch the data directly from this provider

Info
More providers will be available in the future, but to use Saga full functionality we recommend the use of the saga-elastic-provider.

Other important configuration to notice is the use of Java 11, for the compilation of the code and the encoding UTF-8, as you can see in the lines 36-38

Code Block

language	xml
theme	RDark
firstline	0
title	pom.xml
linenumbers	true

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>

    <groupId>com.accenture.saga</groupId>
    <artifactId>saga-howto</artifactId>
    <version>1.0.0-SNAPSHOT</version>

    <properties>
        <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
        <saga.version>1.0.0</saga.version>
    </properties>

    <build>
        <plugins>
            <plugin>
                <groupId>org.apache.maven.plugins</groupId>
                <artifactId>maven-compiler-plugin</artifactId>
                <version>3.8.0</version>
                <executions>
                    <execution>
                        <id>compile</id>
                        <phase>compile</phase>
                        <goals>
                            <goal>compile</goal>
                        </goals>
                    </execution>
                    <execution>
                        <id>testCompile</id>
                        <phase>test-compile</phase>
                        <goals>
                            <goal>testCompile</goal>
                        </goals>
                    </execution>
                </executions>
                <configuration>
                    <source>11</source>
                    <target>11</target>
                    <encoding>${project.build.sourceEncoding}</encoding>
                </configuration>
            </plugin>
            <plugin>
                <artifactId>maven-assembly-plugin</artifactId>
                <version>3.1.0</version>
                <configuration>
                    <archive>
                        <manifest>
                            <mainClass>com.accenture.saga.server.SagaServer</mainClass>
                        </manifest>
                    </archive>
                    <descriptorRefs>
                        <descriptorRef>jar-with-dependencies</descriptorRef>
                    </descriptorRefs>
                    <appendAssemblyId>false</appendAssemblyId>
                </configuration>
                <executions>
                    <execution>
                        <id>make-assembly</id> <!-- this is used for inheritance merges -->
                        <phase>package</phase> <!-- bind to the packaging phase -->
                        <goals>
                            <goal>single</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>
        </plugins>
    </build>

    <dependencies>
        <dependency>
            <groupId>com.accenture.saga</groupId>
            <artifactId>saga-library</artifactId>
            <version>${saga.version}</version>
            <scope>compile</scope>
        </dependency>

        <dependency>
            <groupId>com.accenture.saga</groupId>
            <artifactId>saga-elastic-provider</artifactId>
            <version>${saga.version}</version>
            <scope>compile</scope>
        </dependency>
    </dependencies>
</project>

Setting Up The Configuration

For starters, we will create a main class, which will hold a SagaEngine, ResourceManager, TagManager and PipelineManager.

Code Block

language	java
theme	RDark
linenumbers	true

package com.accenture.saga;

import com.accenture.saga.engine.PipelineManager;
import com.accenture.saga.engine.SagaEngine;
import com.accenture.saga.resourcemgr.ResourceManager;
import com.accenture.saga.tags.TagManager;

public class Main {

    SagaEngine engine;
    ResourceManager resourceManager;
    TagManager tagManager;
    PipelineManager pipelineManager;

    /**
     * Constructor
     */
    public Main() {
        
        
    }
    
    public static void main(String[] args) {
        Main _instance = new Main();
    }
}

First we start by creating an configuring the ResourceManager, and adding a provider to it, but this configuration will be hard-coded, so we need to add SagaJsonFactory class, which allow us to create SagaJson objects (the standard document of Saga) from text, files or readers.

The configuration we are going to use for the provider is the following

Saga_json
"name": "saga-provider", "type": "Elastic", "scheme": "http", "hostname": "localhost", "port": 9200, "timestamp": "updatedAt", "exclude": [ "updatedAt", "createdAt" ]

Each field from the top, starting with the common

Parameter
summary The name we are going to use for the provider. It doesn't which name you use, but our take is "saga-provider"
name name
required true
Parameter
summary Indicates the type of provider we are using, in this case since we are using saga-elastic-provider, it's type would be "Elastic"
name type
required true

from here on, all the properties are specific to saga-elastic-indexer

Parameter
summary Schema for the url to Elasticsearch
default http
name shema
Parameter
summary Name of the hosting server
default localhost
name hostname
Parameter
summary Port of ElasticSearch
default 9200
name 9200
type integer
Parameter
summary Name of the field reflecting any change done to the data
default updatedAt
name timestamp
Parameter
summary Name of the fields omitted (when possible) from the response of ElasticSearch
name exclude
type string array

Our code should look this this

Code Block

language	java
theme	RDark
linenumbers	true

public Main() {

    resourceManager = new ResourceManager();

    resourceManager.registerProvider(
        SagaJsonFactory.getInstance(
            "{"
            + "\"name\":\"saga-provider\","
            + "\"type\":\"FileSystem\","
            + "\"baseDir\":\"testdata\""
            + "}"
        )
    );
}

Next we proceed with the configuration of the TagManager below the ResourceManager, once again we will hard-code the configuration for this one

Saga_json
"resource": "saga-provider:saga_tags"

In the configuration above, saga-provider is representing the provider we add to the ResourceManager in the previous configuration, then the colon (:) indicates the division between the provider and the actual resource; since we are using a saga-elastic-provider, the resources will be indexes names, and since were are connecting to a Saga index, created by the Saga server, all the indexes will be a combination between the solution's name (usually will be saga), and underscore (_) and the type of data the index holds, in this case tags, forming the name saga_tags

The code should look now like this

Code Block

language	java
theme	RDark
linenumbers	true

public Main() {

    resourceManager = new ResourceManager();

    resourceManager.registerProvider(
            SagaJsonFactory.getInstance(
                    "{"
                            + "\"name\":\"saga-provider\","
                            + "\"type\":\"FileSystem\","
                            + "\"baseDir\":\"testdata\""
                            + "}"
            )
    );

    tagManager = new TagManager(resourceManager, SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_tags\"}"));
}

As you can see the TagManager, receives as a parameter the ResourceManager, which grant us access to the resource saga-provider:saga_tags.

In the same way, we proceed to configure the PipelinesManager using the following configuration

Saga_json
"resource": "saga-provider:saga_pipelines"

Once again, saga-provider, does reference to the provider added in the ResourceManager, and saga_pipelines is a combination between the name of the solution's name and the type of data, in this case pipelines.

Code Block

language	java
theme	RDark
linenumbers	true

public Main() {

    resourceManager = new ResourceManager();

    resourceManager.registerProvider(
            SagaJsonFactory.getInstance(
                    "{"
                            + "\"name\":\"saga-provider\","
                            + "\"type\":\"FileSystem\","
                            + "\"baseDir\":\"testdata\""
                            + "}"
            )
    );

    tagManager = new TagManager(resourceManager, SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_tags\"}"));

    pipelineManager = new PipelineManager(resourceManager,  SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_pipelines\"}"));
}

Configuring the Engine

So one we are at the fun part, once we got the Resource, Tag and Pipeline Manager all set up, we assign the ResourceManager and the TagManager to the Engine

Code Block

language	java
theme	RDark
linenumbers	true

public Main() {

    resourceManager = new ResourceManager();

    resourceManager.registerProvider(
            SagaJsonFactory.getInstance(
                    "{"
                            + "\"name\":\"saga-provider\","
                            + "\"type\":\"FileSystem\","
                            + "\"baseDir\":\"testdata\""
                            + "}"
            )
    );

    tagManager = new TagManager(resourceManager, SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_tags\"}"));

    pipelineManager = new PipelineManager(resourceManager,  SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_pipelines\"}"));

    engine = new SagaEngine();

    engine.setResourceManager(resourceManager);
    engine.setTagManager(tagManager);
}

Working With The Pipeline Manager

From here we got 2 options, first option is letting the PipelineManager build the pipeline for us, using a set of tags (that we will provide); the second option is to manually provide a complete pipeline configuration to the PipelineManager

Pros & Cons

Automatic Pipeline

Manual Pipeline

Pros

Uses configuration set up through Saga's UI
Loads only the necessary from the resource (tags, stages, ...)
Builds pipeline based on tag dependency
Can generate multiple and different pipelines
Each Recognizer can have a base pipeline as dependency

Pros

Complete control over the flow of the data

Cons

Pipelines not always the most efficient (...yet)
Each base pipeline must be configure manually (... for the moment)

Cons

Configuration of every stage must be done manually
Relies strongly in the knowledge of the user for each possible stage configuration
Lack of flexibly when changing to another pipeline

Tie

Needs a stage of type TextBlockReader configure manually

Tie

Needs a stage of type TextBlockReader configure manually

Since the first one is the most flexible and the one that makes use of the configuration in ElasticSearch, we will use that one.

Request A Pipeline

Before asking the PipelineManager for a pipeline, we need to provide a stage of type TextBlockReader, at the moment we only have one stage of that type, the SimpleReaderStage, which requires a splitRegex in the configuration, as a SagaJson object. So let's add that to the code shall we.

Code Block

language	java
theme	RDark
linenumbers	true

public Main() {

    resourceManager = new ResourceManager();

    resourceManager.registerProvider(
            SagaJsonFactory.getInstance(
                    "{"
                            + "\"name\":\"saga-provider\","
                            + "\"type\":\"FileSystem\","
                            + "\"baseDir\":\"testdata\""
                            + "}"
            )
    );

    tagManager = new TagManager(resourceManager, SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_tags\"}"));

    pipelineManager = new PipelineManager(resourceManager,  SagaJsonFactory.getInstance("{ \"resource\": \"saga-provider:saga_pipelines\"}"));

    //The Fun Part

    engine = new SagaEngine();

    engine.setResourceManager(resourceManager);
    engine.setTagManager(tagManager);


    SimpleReaderStage simpleReaderStage = new SimpleReaderStage(engine, SagaJsonFactory.getInstance("{ \"splitRegex\": \"[\\r|\\n]+\"}"));
}

The splitRegex is the way the SimpleReader know when to split the text for processing, with the regex [\r\n]+ we are indicating the character signaling a break line; also note the SimpleReaderStage receives the engine as the first parameter.

Tip
The regex [\r\n]+, is the standard for mostly all the text you will be processing

Page tree

Versions Compared

Old Version 3

New Version 4

Key

Step-by-step guide

Configure pom.xml

Setting Up The Configuration

Configuring the Engine

Working With The Pipeline Manager

Pros & Cons

Request A Pipeline

Related articles