def merge_wrapper(merge_config: str, source: List, destination: List,
processes: int):
"""
Load nodes and edges from files and KGs, as defined in a config YAML, and merge them into a single graph.
The merged graph can then be written to a local/remote Neo4j instance OR be serialized into a file.
\f
.. note::
Everything here is driven by the ``merge-config`` YAML.
Parameters
----------
merge_config: str
Merge config YAML
source: List
A list of source to load from the YAML
destination: List
A list of destination to write to, as defined in the YAML
processes: int
Number of processes to use
"""
try:
merge(merge_config, source, destination, processes)
exit(0)
except Exception as me:
get_logger().error(f"kgx.merge error: {str(me)}")
exit(1)
def test_merge2():
"""
Transform selected source from test merge YAML and
write selected destinations.
"""
merge_config = os.path.join(RESOURCE_DIR, 'test-merge.yaml')
merge(merge_config=merge_config, destination=['merged-graph-json'])
assert os.path.join(TARGET_DIR, 'merged-graph.json')
def test_merge1():
"""
Transform from test merge YAML.
"""
merge_config = os.path.join(RESOURCE_DIR, 'test-merge.yaml')
merge(merge_config=merge_config)
assert os.path.join(TARGET_DIR, 'merged-graph_nodes.tsv')
assert os.path.join(TARGET_DIR, 'merged-graph_edges.tsv')
assert os.path.join(TARGET_DIR, 'merged-graph.json')
def load_and_merge(yaml_file: str, processes: int = 1) -> nx.MultiDiGraph:
"""Load and merge sources defined in the config YAML.
Args:
yaml_file: A string pointing to a KGX compatible config YAML.
processes: Number of processes to use.
Returns:
networkx.MultiDiGraph: The merged graph.
"""
merged_graph = merge(yaml_file, processes=processes)
return merged_graph
