from pathlib import Path
from typing import Hashable, Optional, Union
import networkx
from ewoksutils.import_utils import qualname
from .. import inittask
from .subgraph import extract_graph_nodes
from .subgraph import add_subgraph_links
from . import analysis
from . import serialize
from .validate import validate_graph
from .multigraph import flatten_multigraph
from .error_handlers import connect_default_error_handlers
from .execute.sequential import execute_graph
from .compare import graphs_are_equal
from .models import GraphSource
[docs]
class TaskGraph:
"""The API for graph analysis is provided by `networkx`.
Any directed graph is supported (cyclic or acyclic).
Loop over the dependencies of a task
.. code-block:: python
for source in taskgraph.predecessors(target):
link_attrs = taskgraph.graph[source][target]
Loop over the tasks dependent on a task
.. code-block:: python
for target in taskgraph.successors(source):
link_attrs = taskgraph.graph[source][target]
For acyclic graphs, sequential task execution can be done like this:
.. code-block:: python
taskgraph.execute()
"""
def __init__(
self,
source: Optional[GraphSource] = None,
representation: Optional[Union[serialize.GraphRepresentation, str]] = None,
**load_options,
):
self.load(source=source, representation=representation, **load_options)
def __repr__(self):
return self.graph_label
@property
def graph_id(self) -> Hashable:
return self.graph.graph.get("id", qualname(type(self)))
@property
def graph_label(self) -> str:
return str(self.graph.graph.get("label", self.graph_id))
def __eq__(self, other):
if not isinstance(other, type(self)):
raise TypeError(other, type(other))
return graphs_are_equal(self.graph, other.graph)
[docs]
def load(
self,
source: Optional[GraphSource] = None,
representation: Optional[Union[serialize.GraphRepresentation, str]] = None,
subgraph_representation: Optional[
Union[serialize.GraphRepresentation, str]
] = None,
**load_options,
) -> None:
graph = serialize.load(
source=source, representation=representation, **load_options
)
if subgraph_representation is not None:
representation = subgraph_representation
subgraphs = get_subgraphs(graph, representation=representation, **load_options)
if subgraphs:
# Extract
edges, update_attrs = extract_graph_nodes(graph, subgraphs)
graph = flatten_multigraph(graph)
# Merged
self.graph = graph
graphs = [self] + list(subgraphs.values())
rename_nodes = [False] + [True] * len(subgraphs)
graph = merge_graphs(
graphs,
graph_attrs=graph.graph,
rename_nodes=rename_nodes,
**load_options,
).graph
# Re-link
add_subgraph_links(graph, edges, update_attrs)
graph = flatten_multigraph(graph)
connect_default_error_handlers(graph)
validate_graph(graph)
self.graph = graph
[docs]
def dump(
self,
destination: Optional[Union[str, Path]] = None,
representation: Optional[Union[serialize.GraphRepresentation, str]] = None,
**kw,
) -> Optional[Union[str, Path, dict]]:
return serialize.dump(
self.graph, destination=destination, representation=representation, **kw
)
[docs]
def serialize(self) -> str:
return self.dump(representation=serialize.GraphRepresentation.json_string)
@property
def is_cyclic(self) -> bool:
return analysis.graph_is_cyclic(self.graph)
@property
def has_conditional_links(self) -> bool:
return analysis.graph_has_conditional_links(self.graph)
[docs]
def execute(self, *args, **kw):
return execute_graph(self.graph, *args, **kw)
[docs]
def load_graph(
source: Optional[Union[TaskGraph, GraphSource]] = None,
representation: Optional[Union[serialize.GraphRepresentation, str]] = None,
**load_options,
):
if isinstance(source, TaskGraph):
return source
else:
return TaskGraph(source=source, representation=representation, **load_options)
[docs]
def node_has_links(graph, node_id):
try:
next(graph.successors(node_id))
except StopIteration:
try:
next(graph.predecessors(node_id))
except StopIteration:
return False
return True
[docs]
def merge_graphs(graphs, graph_attrs=None, rename_nodes=None, **load_options):
lst = list()
if rename_nodes is None:
rename_nodes = [True] * len(graphs)
else:
assert len(graphs) == len(rename_nodes)
for g, rename in zip(graphs, rename_nodes):
g = load_graph(g, **load_options)
gname = repr(g)
g = g.graph
if rename:
mapping = {s: (gname, s) for s in g.nodes}
g = networkx.relabel_nodes(g, mapping, copy=True)
lst.append(g)
ret = load_graph(networkx.compose_all(lst), **load_options)
if graph_attrs:
ret.graph.graph.update(graph_attrs)
return ret
[docs]
def get_subgraphs(graph: networkx.DiGraph, **load_options):
subgraphs = dict()
for node_id, node_attrs in graph.nodes.items():
task_type, task_info = inittask.task_executable_info(
node_id, node_attrs, all=True
)
if task_type == "graph":
g = load_graph(task_info["task_identifier"], **load_options)
g.graph.graph["id"] = node_id
node_label = node_attrs.get("label")
if node_label:
g.graph.graph["label"] = node_label
subgraphs[node_id] = g
return subgraphs