Workflow API

workflow class

Module: hermes.workflow.workflow

The central class that loads JSON workflow definitions and builds executable task networks.

Constructor

workflow(workflowJSON, WD_path=None, Resources_path=None, name=None)

Parameter	Type	Description
workflowJSON	dict	The workflow JSON definition
WD_path	str	Working directory path
Resources_path	str	Path to the Resources directory
name	str	Optional workflow name

Properties

Property	Type	Description
json	dict	Full workflow JSON with GUI and final nodes
workflowJSON	dict	The workflow section only
nodeList	list	Ordered list of node names
nodes	dict	Node definitions dictionary
parametersJSON	dict	Extracted parameters only
taskRepresentations	dict	Map of node names to TaskWrapper lists
solver	str	Solver configuration

Methods

build(buildername)

Build execution code for the specified engine.

wf.build("luigi")

Parameter	Type	Description
buildername	str	Engine name (e.g., "luigi")

write(workflowName, directory, fullJSON=False)

Persist the workflow to a JSON file.

wf.write("my_workflow", "/output", fullJSON=True)

updateNodes(parameters)

Update node parameters from a dictionary.

wf.updateNodes({"Parameters": {"OFversion": "10"}})

getRootTaskName()

Returns the name of the root (final) task in the workflow.

Dictionary Access

The workflow supports dictionary-like access to nodes:

node = wf["NodeName"]          # Get node
wf["NewNode"] = node_def       # Set node
del wf["OldNode"]              # Delete node
for name in wf.keys():         # Iterate names
    ...

---

hermesNode class

Module: hermes.workflow.workflow

Provides a convenient interface to individual node JSON definitions.

Constructor

hermesNode(name, nodeJSON)

Properties

Property	Type	Description
name	str	Node name
parameters	dict	Node input parameters
executionJSON	dict	Execution configuration only
parametersTable	DataFrame	Pandas-formatted parameters table

Methods

Dictionary Access

value = node["paramName"]      # Get parameter
node["paramName"] = value      # Set parameter
for key in node.keys():        # Iterate parameter names
    ...