workflowlite

pip install workflowlite

A workflow engine that allows the execution of jobs composed of multiple steps. Each step is defined by an action and a set of inputs and outputs.

The job definition can include on_except and on_finish hooks to handle the job's completion or failure. The context is updated during execution with runtime information like job_run_id and job_status. All the system variables are prefixed with $.

Example Usage:

from workflowlite import WorkflowEngine, Context

engine = WorkflowEngine()

@engine.register_action('prepare')
def prepare(inputs: List[str], context: Context):
   context['prepared_output'] = "prepared_output.mp4"

@engine.register_action('ffmpeg')
def ffmpeg(inputs: List[str], context: Context):
   context['mp3_output'] = "output.mp3"

@engine.register_hook('on_finish')
def on_finish(context: Context):
   print("Job finished successfully!")

@engine.register_hook('on_except')
def on_except(context: Context):
   print("Job encountered an exception.")

job = {
   "name": "video_processing",
   "env": {"input_file": "video.mp4"},
   "steps": [
       {
           "action": "prepare",
           "input": ["input_file"],
           "output": ["prepared_output"]
       },
       {
           "action": "ffmpeg",
           "input": ["prepared_output"],
           "output": ["mp3_output"]
       }
   ],
   "output": ["mp3_output"],
   "on_finish": "on_finish",
   "on_except": "on_except"
}

with engine:
   job_future = engine.submit_job(job)
   try:
       result = job_future.wait()
       print(f"Job completed with output: {result}")
   except Exception as e:
       print(f"Job failed with exception: {e}")