Merge 8dc81d8 into 2f3cb47

kg_covid_19/load_utils/merge_kg.py

@@ -1,10 +1,10 @@
 import logging
 from typing import Dict, List
 import yaml
-
+import networkx as nx
 from kgx import Transformer, NeoTransformer
 from kgx.cli.utils import get_file_types, get_transformer
-
+from kgx.operations.graph_merge import GraphMerge
 
 def parse_load_config(yaml_file: str) -> Dict:
     """Parse load config YAML.
@@ -21,16 +21,17 @@ def parse_load_config(yaml_file: str) -> Dict:
     return config
 
 
-def load_and_merge(yaml_file: str) -> Transformer:
+def load_and_merge(yaml_file: str) -> nx.MultiDiGraph:
     """Load and merge sources defined in the config YAML.
 
     Args:
         yaml_file: A string pointing to a KGX compatible config YAML.
 
     Returns:
-        kgx.Transformer: The merged transformer that contains the merged graph.
+        networkx.MultiDiGraph: The merged graph.
 
     """
+    gm = GraphMerge()
     config = parse_load_config(yaml_file)
     transformers: List = []
 
@@ -52,19 +53,17 @@ def load_and_merge(yaml_file: str) -> Transformer:
             logging.error("type {} not yet supported".format(target['type']))
 
     # merge all subgraphs into a single graph
-    merged_transformer = Transformer()
-    merged_transformer.merge_graphs([x.graph for x in transformers])
-    merged_transformer.report()
+    merged_graph = gm.merge_all_graphs([x.graph for x in transformers])
 
     # write the merged graph
     if 'destination' in config:
         destination = config['destination']
         if destination['type'] in ['csv', 'tsv', 'ttl', 'json', 'tar']:
-            destination_transformer = get_transformer(destination['type'])(merged_transformer.graph)
+            destination_transformer = get_transformer(destination['type'])(merged_graph)
             destination_transformer.save(destination['filename'], extension=destination['type'])
         elif destination['type'] == 'neo4j':
             destination_transformer = NeoTransformer(
-                merged_transformer.graph,
+                merged_graph,
                 uri=destination['uri'],
                 username=destination['username'],
                 password=destination['password']
@@ -73,4 +72,4 @@ def load_and_merge(yaml_file: str) -> Transformer:
         else:
             logging.error("type {} not yet supported for KGX load-and-merge operation.".format(destination['type']))
 
-    return merged_transformer
+    return merged_graph

kg_covid_19/transform_utils/scibite_cord/scibite_cord.py

@@ -97,9 +97,10 @@ def parse_annotation_doc(self, node_handle, edge_handle, doc: Dict, subset: str
         """
         terms = set()
         paper_id = doc['paper_id']
-
+        title = None
         if 'metadata' in doc:
             metadata = doc['metadata']
+            title = metadata['title'].replace('\n', ' ')
             # extract hits from metadata
             terms.update(self.extract_termite_hits(metadata))
 
@@ -125,7 +126,7 @@ def parse_annotation_doc(self, node_handle, edge_handle, doc: Dict, subset: str
             header=self.node_header,
             data=[
                 f"CORD:{paper_id}",
-                f"{metadata['title']}",
+                f"{title}",
                 "biolink:Publication",
                 ""
             ]