Workflow Operation Handler

Introduction

Workflows are the central element to define how a media package is being processed by the Opencast services. Their definitions consist of a list of workflow operations, which basically map a piece of configuration to Opencast code:

<definition xmlns="http://workflow.opencastproject.org">
    ....
    <operation
      id="tag"
      <configurations>
        <configuration key="source-flavors">presentation/trimmed</configuration>
        <configuration key="target-flavor">presentation/tagged</configuration>
      </configurations>
   </operation>
   ...
</definition>

Default Workflow Operations

The following table contains the workflow operations that are available in an out-of-the-box Opencast installation:

Operation Handler Description Details
add-catalog Add a catalog to the media package Documentation
analyze-audio Analyze first audio stream Documentation
analyze-tracks Analyze tracks in media package Documentation
analyze-mediapackage Analyze media package Documentation
animate Create animated video sequence Documentation
amberscript-start-transcription Start AmberScript Transcription Documentation
amberscript-attach-transcription Attach AmberScript Transcription Documentation
assert Verify preconditions with assertion Documentation
asset-delete Deletes the current mediapackage from the Archive Documentation
attach-watson-transcription Attaches automated transcripts to mediapackage Documentation
changetype Change type of media package elements Documentation
cleanup Cleanup the working file repository Documentation
clone Clone media package elements to another flavor Documentation
comment Add, resolve or delete a comment Documentation
composite Compose two videos on one canvas. Documentation
concat Concatenate multiple video tracks into one video track Documentation
conditional-config Configure workflow configuration variable based on coditions Documentation
configure-by-dcterm Set workflow parameter if dublincore term matches value Documentation
copy Copy media package elements to target directory Documentation
cover-image Generate a cover-image containing metadata Documentation
crop-video Checks for black bars on the sides of the video Documentation
cut-marks-to-smil Parses timestamps into a SMIL for the editor workflow Documentation
defaults Applies default workflow configuration values Documentation
demux Demuxes streams to multiple output files Documentation
duplicate-event Create an event by cloning an existing one Documentation
editor Waiting for user to review, then cut video based on edit-list Documentation
encode Encode media files to differents formats in parallel Documentation
error-resolution Internal operation to pause a workflow in error Documentation
execute-many Execute a command for each matching element in a MediaPackage Documentation
execute-once Execute a command for a MediaPackage Documentation
export-wf-properties Export workflow properties Documentation
extract-text Extracting text from presentation segments Documentation
google-speech-attach-transcription Attaches automated transcripts to mediapackage Documentation
google-speech-start-transcription Starts automated transcription provided by Google Speech Documentation
http-notify Notifies an HTTP endpoint about the process of the workflow Documentation
image Extract images from a video using FFmpeg Documentation
image-convert Convert images using FFmpeg Documentation
image-to-video Create a video track from a source image Documentation
import-wf-properties Import workflow properties Documentation
incident Testing incidents on a dummy job Documentation
include Include workflow definition in current workflow Documentation
ingest-download Download files from external URL for ingest Documentation
inspect Inspect the media (check if it is valid) Documentation
log Log workflow status Documentation
mattermost-notify Send messages to Mattermost Documentation
metadata-to-acl Add read/write access based on metadata Documentation
microsoft-azure-attach-transcription Attaches automated transcripts to mediapackage Documentation
microsoft-azure-start-transcription Starts automated transcription provided by Microsoft Azure Documentation
move-storage Move files between asset manager storage systems Documentation
multiencode Encode to multiple profiles in one operation Documentation
normalize-audio Normalize first audio stream Documentation
partial-import Import partial tracks and process according to a SMIL document Documentation
post-mediapackage Send mediapackage to remote service Documentation
prepare-av Preparing audio and video work versions Documentation
probe-resolution Set workflow instance variables based on video resolution Documentation
process-smil Edit and Encode media defined by a SMIL file Documentation
publication-to-workspace Copy publication element to mediapackage in workspace Documentation
publish-configure-aws Distribute and publish media to the configured publication using Amazon S3 and Cloudfront Documentation
publish-configure Distribute and publish media to the configured publication Documentation
publish-engage-aws Distribute and publish media to the engage player using Amazon S3 and Cloudfront Documentation
publish-engage Distribute and publish media to the engage player Documentation
publish-oaipmh Distribute and publish media to a OAI-PMH repository Documentation
publish-youtube Distribute and publish media to YouTube Documentation
rename-files Rename media files using metadata Documentation
republish-oaipmh Update media in a OAI-PMH repository Documentation
retract-configure-aws Retracts media from configured publication in AWS S3 and Cloudfront Documentation
retract-configure Retracts media from configured publication Documentation
retract-engage-aws Retracts media from AWS S3 and Cloudfront publication Documentation
retract-engage Retracts media from Opencast Media Module publication Documentation
retract-partial-aws Retract a subset of the mediapackage from a publication in Amazon S3 and Cloudfront Documentation
retract-partial Retract a subset of the mediapackage from a publication Documentation
retract-oaipmh Retracts media from a OAI-PMH repository Documentation
retract-partial Retract a subset of the mediapackage from a publication Documentation
retract-youtube Retracts media from YouTube Documentation
sanitize-adaptive Fix references to media files in a playlist Documentation
segment-video Extracting segments from presentation Documentation
segmentpreviews Extract segment images from a video using FFmpeg Documentation
select-streams Select streams for further processing Documentation
select-version Select a version of the media package to run the current workflow with Documentation
send-email Sends email notifications at any part of a workflow Documentation
series Apply series to the mediapackage Documentation
silence Silence detection on audio of the mediapackage Documentation
snapshot Archive the current state of the mediapackage Documentation
speechtotext Create subtitles for video and audio sources Documentation
start-watson-transcription Starts automated transcription provided by IBM Watson Documentation
start-workflow Start a new workflow for given media package ID Documentation
statistics-writer Log statistical data about the video Documentation
tag Modify the tag sets of media package elements Documentation
tag-by-dcterm Modify the tags if dublincore term matches value Documentation
tag-engage Modify the tag sets of media package elements in the engage publication Documentation
theme Make settings of themes available to processing Documentation
timelinepreviews Create a preview image stream from a given video track Documentation
transfer-metadata Transfer metadata fields between catalogs Documentation
video-grid Put parallel video streams on a single video canvas Documentation
waveform Create a waveform image of the audio of the mediapackage Documentation
webvtt-to-cutmarks Create cutting suggestions from subtitles Documentation
zip Create zipped archive of the current state of the mediapackage Documentation

State Mappings

Technically, a workflow can be in one of the following states:

Technical State Description What the Admin UI displays in the events table
instantiated The workflow is queued and will be started as soon as possible "Pending"
running The workflow is running, no problems so far "Running"
stopped The workflow was aborted by the user "Processing canceled"
paused The workflow was paused and can be continued "Paused"
succeeded The workflow has completed successfully "Finished"
failed The workflow failed due to an error "Processing failure"
failing The workflow is still running, but there were errors. It will fail. "Running"

Using state mappings, it is possible to refine the labels displayed in the Admin UI events table for a particular workflow.

Here is an example which displays "Retracting" instead of "Running" for the retract workflow:

<?xml version="1.0" encoding="UTF-8"?>
<definition xmlns="http://workflow.opencastproject.org">
  <id>retract</id>
  ...
  <state-mappings>
    <state-mapping state="running">retracting</state-mapping>
    <state-mapping state="failing">retracting</state-mapping>
  </state-mappings>

When no state mappings are configured for a workflow, the generic default labels will be displayed.

When a workflow includes other workflows, the event table only shows the state of the including workflow.