Doi patch for citation field in uns (#74)

* Updated version to 0.1.13

* Added dataset object to rdf representation

* Disabled supercluster_of removal to fix missing nodes in graph generation

* Added add_metadata_nodes method

* Rounded up percentages in add_metadata_nodes

* Refactored add_metadata_nodes to utilize OWL reification for improved visualisation in neo4j.

* Added parsing mechanism for citation field in uns

pandasaurus_cxg/graph_generator/graph_generator.py

@@ -5,7 +5,6 @@
 
 import matplotlib.pyplot as plt
 import networkx as nx
-import pandas as pd
 from pandasaurus.graph.graph_generator import GraphGenerator as graphgen
 from rdflib import OWL, RDF, RDFS, BNode, Graph, Literal, Namespace, URIRef
 from rdflib.plugins.sparql import prepareQuery
@@ -19,9 +18,11 @@
     add_edge,
     add_node,
     add_outgoing_edges_to_subgraph,
+    citation_field_name,
     colour_mapping,
     find_and_rotate_center_layout,
     generate_subgraph,
+    parse_citation_field_into_dict,
     remove_special_characters,
     select_node_with_property,
 )
@@ -129,6 +130,17 @@ def generate_rdf_graph(self):
         for key, value in uns.items():
             if not isinstance(value, str):
                 continue
+            if key == citation_field_name:
+                citation_dict = parse_citation_field_into_dict(value)
+                for citation_key, citation_value in citation_dict.items():
+                    self.graph.add(
+                        (
+                            dataset_class,
+                            URIRef(self.ns[citation_key]),
+                            Literal(citation_value),
+                        )
+                    )
+
             self.graph.add((dataset_class, URIRef(self.ns[key]), Literal(value)))
         has_source = URIRef(HAS_SOURCE["iri"])
         self.graph.add((has_source, RDFS.label, Literal(HAS_SOURCE["label"])))

pandasaurus_cxg/graph_generator/graph_generator_utils.py

@@ -1,4 +1,5 @@
 import re
+from typing import Dict
 
 import networkx as nx
 from rdflib import OWL, RDF, RDFS, BNode, Graph, Literal, Namespace, URIRef
@@ -14,6 +15,8 @@
     "http://purl.obolibrary.org/obo/PCL_0010001": "cyan",
 }
 
+citation_field_name = "citation"
+
 
 def add_edge(nx_graph: nx.Graph, subject, predicate, obj):
     edge_data = {
@@ -126,3 +129,32 @@ def select_node_with_property(graph: Graph, _property: str, value: str):
 
 def remove_special_characters(input_string: str) -> str:
     return re.sub(r"[^a-zA-Z0-9_]", "", input_string.replace(" ", "_"))
+
+
+def parse_citation_field_into_dict(value: str) -> Dict[str, str]:
+    """
+    Parses a citation string into a dictionary by extracting key citation fields.
+
+    Args:
+        value: The string containing citation fields and values.
+
+    Returns:
+        A dictionary with keys such as 'Publication', 'Dataset Version', and 'Collection',
+          and corresponding values extracted from the input string.
+    """
+    # Split the input string on the key terms
+    parts = value.split(" ")
+    keys = ["Publication:", "Version:", "Collection:"]
+    key_indices = [parts.index(key) for key in keys if key in parts]
+    # Break down into key-value pairs
+    key_value_pairs = {}
+    for i, index in enumerate(key_indices):
+        current_value = " ".join(parts[index + 1 : index + 2])
+        key_value_pairs.update(
+            {
+                "download_link"
+                if parts[index][:-1].lower() == "version"
+                else parts[index][:-1].lower(): current_value
+            }
+        )
+    return key_value_pairs