Scraper

While the Apache PLC4X API allows simple access to PLC resources, if you want to continuously monitor some values and have them retrieved in a pre-defined interval, the core PLC4X API method is a little bit uncomfortable.

Especially when you have multiple batches of data you want to have refreshed in different intervals.

In this case you need to take care of the scheduling of queries, need to manage the connection state (Check if the connection is still available and to apply countermeasures, if there are problems)

As we have encountered exactly the same problem for about every integration module we created, the Apache PLC4X team has created a tool called the Scraper.

This tool automatically handles all of the tasks mentioned above.

Getting started with the Scraper

In general, you need 3 parts to work with the Scraper:

1) A Scraper Configuration 2) A Scraper Implementation 3) A Handler to handle the results of Scraper jobs

In the Scraper Configuration you define the so-called jobs.

Sources

Sources define connections to PLCs using PLC4X drivers.

Generally you can think of a Source as a PLC4X connection string, given an alias name.

Jobs

A Job defines which resources (PLC Addresses) should be collected from which Sources with a given Trigger.

All resources in a job will be collected as a batch.

Generally multiple types of triggers could theoretically be supported, but for now only a time triggered job (Aka SCHEDULED) is actually supported.

In the near future we’re hoping that we will be able to support: - External triggers - Triggering collection based upon PLC-values

But, as to now, this has not been implemented yet.

Configuration using the Java API

The core of the Scraper configuration is the ScraperConfigurationTriggeredImplBuilder class. Use this to build the configuration objects used to bootstrap the Scraper.

ScraperConfigurationTriggeredImplBuilder builder = new ScraperConfigurationTriggeredImplBuilder();

As soon as you have your builder instance, you should add at least one source to it.

builder.addSource({connectionName}, {plc4xConnectionString});

The connectionName will be what we use when configuring the job to reference which source it should use to collect.

In order to configure a job we have to get an instance of a JobConfigurationTriggeredImplBuilder.

JobConfigurationTriggeredImplBuilder jobBuilder = builder.job({jobName}, {triggerCommand});

This creates a new job with a given name which is executed based on the information in the triggerCommand.

As mentioned above, we currently only support a time-scheduled collection.

This genrally just requires one parameter: The number of milliseconds between each collection.

SCHEDULED:1000

Above would schedule a collection every 1000ms - so once every second.

Up to now this job would not be run anywhere, and it would also not collect anything. So in order to have the job actually do something, we should assign it a source to collect from.

jobBuilder.source({connectionName});

Here we could theoretically collect on multiple sources, by simply calling the source() method multiple times.

All sources would be collected at the same time, whenever the trigger tells it to.

So the last thing we need to configure our first Scraper job, is to add a few fields for it to collect.

jobBuilder.field({fieldName}, {fieldAddress});

The field method has to be called for every field we want to add to the current job configuration. It gives a PLC4X address string an easy to understand string name, just like when using the core PLC4X API.

As soon as we’re done adding fields, we configure the job by calling the build method.

jobBuilder.build();

This configures the finished job and attaches that to the overall Scraper configuration of the scraper configuration.

As soon as we’re done configuring jobs, we need to create the Scraper configuration by calling the buiold method on the builder:

ScraperConfigurationTriggeredImpl scraperConfig = builder.build();

Running the Scraper

In order to run the Scraper, the following boilerplate code is needed.

       try {
            PlcDriverManager plcDriverManager = new PooledPlcDriverManager();
            TriggerCollector triggerCollector = new TriggerCollectorImpl(plcDriverManager);
            TriggeredScraperImpl scraper = new TriggeredScraperImpl(scraperConfig, (jobName, sourceName, results) -> {

                ...

            }, triggerCollector);
            scraper.start();
            triggerCollector.start();
        } catch (ScraperException e) {
            log.error("Error starting the scraper", e);
        }

At first a new PooledPlcDriverManager is created (It actually doesn’t have to be the pooled version, but we strongly suggest you use it as for some protocols the connection process is stressfull for the connected PLC).

With this plcDriverManager we can then create a so-called TriggerCollector, which we pass in the driver manager as argument.

Next comes the probably most important part: We configure the scraper, by binding a Scraper Configuration, a ResultHandler and a TriggerCollector together.

After this, the scraper is ready to start, which is then done by calling start on the scraper as well as the triggerCollector.

For the sake of clarity, here comes the definition of the ResultHandler interface:

@FunctionalInterface
public interface ResultHandler {

    /**
     * Callback handler.
     * @param jobName name of the job (from config)
     * @param connectionName alias of the connection (<b>not</b> connection String)
     * @param results Results in the form alias to result value
     */
    void handle(String jobName, String connectionName, Map<String, Object> results);

}

Configuration using a JSON or YAML file

As an alternative to using the Java API, the Scraper Configuration can also be read from a JSON or YAML document.

Here come some examples:

JSON:

{
    "sources": {
        "connectionName": "connectionString"
    },
    "jobs": [
        {
            "name": "jobName",
            "triggerConfig": (SCHEDULED,10000)
            "sources": [
                "connectionName"
            ],
            "fields": {
                "a": "{address-a}",
                "b": "{address-b}"
            }
        }
    ]
}

YAML:

---
sources:
  connectionName: connectionString
jobs:
  - name: jobName
    triggerConfig: (SCHEDULED,10000)
    sources:
      - connectionName
    fields:
      a: {address-a}
      b: {address-b}

In both cases, you can create the ScraperConfiguration with the following code:

ScraperConfiguration conf = ScraperConfiguration.fromFile("{path to the JSON or YAML file}", ScraperConfigurationTriggeredImpl.class);