workflow-interpreter

Image Workflow Parser & Executor

This example implements a workflow that executes various AI image processing steps, like image generation and transformations.

The workflow is specified dynamically in a json-based workflow definition. The workflow can contain steps that call different image processing services:

• to generate images by taking a screenshot with Puppeteer
• to rotate images
• to blur images

+--------+ Request +------------------------+
| Client | ------> | ImageProcessingWorkflow|
+--------+         +------------------------+
    ^                          |
    |                          | Parse JSON workflow definition
    |                          v
    |              +------------------------+
    |              |   Workflow Execution   |----+
    |              +------------------------+    |
    |                          |                 |
    |                          |    +------------------------+
    |                          |    | Optional Services      |
    |                          |    | (any order)            |
    |                          |    |  +-----------------+   |
    |                          |    |  | PuppeteerService|   |
    |                          |    |  +-----------------+   |
    |                          |    |  | TransformerSvc  |   |
    |                          |    |  | (/blur)         |   |
    |                          |    |  +-----------------+   |
    |                          |    |  | TransformerSvc  |   |
    |                          |    |  | (/rotate)       |   |
    |                          |    |  +-----------------+   |
    |                          |    +------------------------+
    |                          |                 |
    |                          <-----------------+
    |                          |
    |                          | Concatenate
    |                          v
    |              +------------------------+
    |              |    Result Assembly     |
    |              +------------------------+
    |                          |
    +<-------------------------|
           Result Report

Note: For simplicity, this example stores images locally in the folder generated-images.

The example deploys each of the possible steps (generation, transformation) as a separate service, and then has a workflow service call them as specified in the workflow definition.

Download the example

You can clone the example repository (git clone https://github.com/restatedev/examples) or just download this example via

• CLI: restate example java-workflow-interpreter

Running the example

1. Start Restate Server in a separate shell: restate-server

2. Start the workflow and image transformation services: mvn compile spring-boot:run

3. Register the example at Restate server by calling restate -y deployment register "localhost:9080".

Demo scenario

Here is list of example workflow execution requests that you can send to the workflow executor:

Example workflow

Puppeteer screenshot -> rotate -> blur:

curl localhost:8080/ImageProcessingWorkflow/user123-wf1/run -H 'content-type: application/json' \
  -d '[
          {"action":"puppeteer","parameters":{"url":"https://restate.dev"}},
          {"action":"rotate","parameters":{"angle":90}},
          {"action":"blur","parameters":{"blur":5}}
       ]'

Have a look at the generated-images folder to see the end result.

Retrieving state

You can retrieve the workflow state via via the CLI. For example for id user123-wf1, do:

restate kv get ImageProcessingWorkflow user123-wf1

Result:

🤖 State:
―――――――――
                                  
 Service  ImageProcessingWorkflow 
 Key      user123-wf1             

 KEY     VALUE                                                             
 status  {                                                                 
           "imgName": "c7879b2b-c213-f723-60c1-593422c98bc2",              
           "output": [                                                     
             "[Took screenshot of website with url: https://restate.dev]", 
             "[Rotated image with angle: 90]",                             
             "[Blurred image with strength param 5]"                       
           ],                                                              
           "status": "Finished"                                            
         }

More complex workflows

Try more complex workflows, and have a look at the end result.

• Puppeteer screenshot -> blur -> rotate -> rotate -> rotate -> rotate:

curl localhost:8080/ImageProcessingWorkflow/user123-wf2/run -H 'content-type: application/json' \
  -d '[
        {"action":"puppeteer","parameters":{"url":"https://restate.dev"}},
        {"action":"blur","parameters":{"blur":5}}, 
        {"action":"rotate","parameters":{"angle":90}}, 
        {"action":"rotate","parameters":{"angle":90}}, 
        {"action":"rotate","parameters":{"angle":90}}, 
        {"action":"rotate","parameters":{"angle":90}}
      ]' 

• Invalid workflow definition (no image source is defined):

curl localhost:8080/ImageProcessingWorkflow/user123-wf3/run -H 'content-type: application/json' \
  -d '[
        {"action":"invalid","parameters":{"angle":90}},
        {"action":"blur","parameters":{"blur":5}}]}'