Updating AEM content with Sling Pipes

Matija Kovacek

Cover Image for Updating AEM content with Sling Pipes

Matija Kovacek

May 22, 2020

Updating AEM content with Sling Pipes

You are still updating content manually? Try out Sling pipes.

Several weeks ago I saw some tweet about how Adobe guys are using Sling pipes for updating their AEM content, so I was curious and I try it out.

Let's see what I have found out...

Sling pipes are a simple tool for executing CRUD operations over resources in the AEM repository. But is it powerful enough to replace your groovy scripts? Let's discover it in the following examples.

Introduction

Just to get your attention, let's say in past you have made some bad decisions during the development of some of your components and now you need to update all content with that component. With Sling Pipes that is very easy.

Example

plumber.newPipe(resourceResolver)
  .echo("/content/yoursite")
  .$("nt:unstructured[sling:resourceType=projectName/components/content/imageSlider]")
  .write("autoplay", true)
  .run()

5 lines? Yes, you can set autoplay property to true in all your content with imageSlider component in 5 lines, and I'm pretty much sure you can't do that so simply with your groovy code, or you can do it manually in crx/de for the whole day...

So what the hell is happening in this code, what are these methods?!

I will explain it later in more details but in tl;dr;

plumber is OSGi service and to start all this magic you need to pass resourceResolver object to newPipe method. After that, you need to call pipes (methods) to achieve what you want.
echo("path") is used to receive root resources on a specific path.
$(expression) is the wrapper of Sling query expressions and I'm using it here to filter out specific resources with imageSlider component
write(conf) is a pipe to write/update resources/properties, in this case, I'm creating or updating autoplay property with value true
run() is used to execute code

Still interested? Let me explain to you Sling pipes in more details.

About Sling Pipes

Sling pipes is a toolset for doing extract - transform - load operations by chaining proven code blocks. A sling pipe is essentially a sling resource stream, encapsulating well-known sling operations.

There are 3 types of pipes:

reader pipes - used to get resources
writer pipes - for writing/updating resources in the repository, depending on configuration and input
logical pipes - they refer to other pipes, chaining them or using their results in a general way

Reader pipes

The reader pipes are used to get resources.

Base pipe: echo(path) Used to receive resource on the given path. Let's say you want to get "/content/we-retail" resource.

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail")
  .run()

XPathPipe: xpath(expr) Used to receive resource with given xpath query Let's say you want to get all nt:unstructured resources with heroimage component under "/content/we-retail".

plumber.newPipe(resourceResolver)
  .xpath("/jcr:root/content/we-retail//element(*, nt:unstructured)
        [(@sling:resourceType = 'weretail/components/content/heroimage')]")
  .run()

Sling Query Pipes Wrapper for executing Sling Query methods in sling pipe way.

Find Pipe: $(expr)
Children Pipe: children(expr)
Siblings Pipe: siblings(expr)
Parent Pipe: parent()
Closest Pipe: closest(expr)

Here is more details about Sling Query

The same example as the previous one, let's say you want to get all resources with heroimage component under "/content/we-retail".

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail")
  .$("nt:unstructured[sling:resourceType=weretail/components/content/heroimage]")
  .run()

Writer pipes

The writer pipes are used to modify resources.

Write Pipe: write(conf) Used to write given nodes and properties to current input resource.

Let's say you want to create/update property "foo" with value "bar" and property "foo2" with value "bar2".

Please let me know is it possible to write different values except for strings, like long, date, boolean?

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail")
  .write('foo','bar','foo2','bar2')
  .run()

PathPipe: mkdir(expr) Used to create resources on the given location.

Let's you want to create resource "master" on "/content/we-retail/master" location. This will create "sling:Folder" node.

Please let me know if is it possible to specify different jcr:primaryType?

plumber.newPipe(resourceResolver)
  .mkdir("/content/we-retail/master")
  .run()

MovePipe: mv(expr) Moves resource or property to the target path.

Following example will move master resource from "/content/we-retail/master" to "/content/we-retail/language-master".

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail/master")
  .mv("/content/we-retail/language-master")
  .run()

RemovePipe: rm() Removes resources or property on the specific location.

The following example will remove the master resource on the given path "/content/we-retail/master".

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail/master")
  .rm()
  .run()

PackagePipe: pkg(expr) Creates a package and adds current resources as a filter.

The following example will package all we retail content

plumber.newPipe(resourceResolver)
  .echo("/content/we-retail")
  .pkg("/etc/packages/we-retail-content.zip")
  .run()

How to use it in your AEM project

Okay, how to run all this in your AEM project? You can configure and execute pipes with java, groovy console, http, or jmx

Dependencies

You need to add "org.apache.sling.query" & "org.apache.sling.pipes" as dependencies and you need to embedded them as part of your bundle or install them as separate bundles (my recommended way).

<dependency>
    <groupId>org.apache.sling</groupId>
    <artifactId>org.apache.sling.query</artifactId>
    <version>4.0.2</version>
    <scope>provided</scope>
</dependency>
<dependency>
    <groupId>org.apache.sling</groupId>
    <artifactId>org.apache.sling.pipes</artifactId>
    <version>3.1.0</version>
    <scope>provided</scope>
</dependency>

<plugin>
  <groupId>org.apache.jackrabbit</groupId>
  <artifactId>filevault-package-maven-plugin</artifactId>
  <extensions>true</extensions>
  <configuration>
      <allowIndexDefinitions>true</allowIndexDefinitions>
      <packageType>mixed</packageType>
      <group>pipes</group>
      <embeddeds>
          <!--sling pipes related-->
          <embedded>
              <groupId>org.apache.sling</groupId>
              <artifactId>org.apache.sling.query</artifactId>
              <target>/apps/pipes-vendor-packages/application/install</target>
          </embedded>
          <embedded>
              <groupId>org.apache.sling</groupId>
              <artifactId>org.apache.sling.pipes</artifactId>
              <target>/apps/pipes-vendor-packages/application/install</target>
          </embedded>
      </embeddeds>
  </configuration>
</plugin>

Code execution

You can write in groovy console.

def plumber = getService("org.apache.sling.pipes.Plumber")

plumber.newPipe(resourceResolver)
  .echo("/content/yoursite")
  .$("nt:unstructured[sling:resourceType=projectName/components/content/imageSlider]")
  .write("autoplay", true)
  .run()

You can write in you Java class.

@Component(service = SlingPipesExamples.class, immediate = true)
public class SlingPipesExamples {

    private static final Logger logger = LoggerFactory.getLogger(SlingPipesExamples.class);

    private static final String SERVICE_USER = "sling-pipes-service-user";

    @Reference
    private ResourceResolverFactory resourceResolverFactory;

    @Reference
    private Plumber plumber;

    public void example() {
        try (final ResourceResolver resourceResolver = this.resourceResolverFactory
                    .getServiceResourceResolver(this.getAuthenticationInfo())) {
            PipeBuilder pipeBuilder = this.plumber.newPipe(resourceResolver);
            // your logic with pipes
        } catch (final LoginException e) {
            logger.error("Exception during creating service resource resolver ", e);
        }
    }

    private Map<String, Object> getAuthenticationInfo() {
        return Collections.singletonMap(ResourceResolverFactory.SUBSERVICE, SERVICE_USER);
    }

}

You can create a resource on a specific location e.g "/etc/sling-pipes/example" and execute it with the HTTP request.

<?xml version="1.0" encoding="UTF-8"?>
<jcr:root xmlns:sling="http://sling.apache.org/jcr/sling/1.0" 
  xmlns:jcr="http://www.jcp.org/jcr/1.0"
  jcr:description="Create directory pipe"
  jcr:primaryType="sling:OrderedFolder"
  sling:resourceType="slingPipes/mkdir"
  expr="/content/sling-pipes/foo"/>

Sum up

Basically that's what I discover so far. In general, I like Sling Pipes, especially how with a few lines of code you can do a lot of stuff. Didn't try with some more complex examples but I guess it would be almost the same.

For more details take a look at offical documentation.

Let me know if for mkdir pipe is possible to specify different jcr:primaryType? If it's only "sling:Folder" then 👎

Also for write pipe, is it possible to write in a property some value different than String?

I hope that you learn something and that you will give it a try Sling Pipes. Let me know what you think and what are your experiences with Sling Pipes :)

Home | Blog

Updating AEM content with Sling Pipes

Updating AEM content with Sling Pipes

Introduction

About Sling Pipes

Reader pipes

Writer pipes

How to use it in your AEM project

Dependencies

Code execution

Sum up