Add the ontoportal mapping model json converter (#281)

* add conversion to ontoportal json tests

* add the to_ontoportal_json converter

* add the write_sssom_ontoportal_json test

* add the write_ontoportal_json function

* add the ontoportal writer to the list of writers of the CLI

* change the  serialisation parameter to use "ontoportal_json"

* fix linters issues

* add "ontoportal_json" to SSSOM_EXPORT_FORMATS array

* remove the process section form the ontoportal format

* auto lint code

* update the test for  the new ontoportal mappings format

sssom/util.py

@@ -93,7 +93,7 @@
     "obographs-json",
     "json",
 ]
-SSSOM_EXPORT_FORMATS = ["tsv", "rdf", "owl", "json", "fhir"]
+SSSOM_EXPORT_FORMATS = ["tsv", "rdf", "owl", "json", "fhir", "ontoportal_json"]
 
 SSSOM_DEFAULT_RDF_SERIALISATION = "turtle"
 

sssom/writers.py

@@ -20,6 +20,7 @@
 
 from .constants import SCHEMA_YAML
 from .parsers import to_mapping_set_document
+from .typehints import PrefixMap
 from .util import (
     PREFIX_MAP_KEY,
     RDF_FORMATS,
@@ -147,6 +148,19 @@ def write_owl(
     print(t.decode(), file=file)
 
 
+def write_ontoportal_json(
+    msdf: MappingSetDataFrame, output: TextIO, serialisation="ontoportal_json"
+) -> None:
+    """Write a mapping set dataframe to the file as the ontoportal mapping JSON model."""
+    if serialisation == "ontoportal_json":
+        data = to_ontoportal_json(msdf)
+        json.dump(data, output, indent=2)
+    else:
+        raise ValueError(
+            f"Unknown json format: {serialisation}, currently only ontoportal_json supported"
+        )
+
+
 # Converters
 # Converters convert a mappingsetdataframe to an object of the supportes types (json, pandas dataframe)
 
@@ -480,6 +494,38 @@ def to_json(msdf: MappingSetDataFrame) -> JsonObj:
     return json_obj
 
 
+def to_ontoportal_json(msdf: MappingSetDataFrame) -> List[Dict]:
+    """Convert a mapping set dataframe to a list of ontoportal mapping JSON nbjects."""
+    prefix_map = msdf.prefix_map
+    metadata: Dict[str, Any] = msdf.metadata if msdf.metadata is not None else {}
+    m_list = []
+
+    def resolve(x):
+        return _resolve_url(x, prefix_map)
+
+    if msdf.df is not None:
+        for _, row in msdf.df.iterrows():
+            json_obj = {
+                "classes": [resolve(row["subject_id"]), resolve(row["object_id"])],
+                "subject_source_id": _resolve_prefix(
+                    row.get("subject_source", ""), prefix_map
+                ),
+                "object_source_id": _resolve_prefix(
+                    row.get("object_source", ""), prefix_map
+                ),
+                "source_name": metadata.get("mapping_set_id", ""),
+                "source_contact_info": ",".join(metadata.get("creator_id", "")),
+                "date": metadata.get("mapping_date", row.get("mapping_date", "")),
+                "name": metadata.get("mapping_set_description", ""),
+                "source": resolve(row.get("mapping_justification", "")),
+                "comment": row.get("comment", ""),
+                "relation": [resolve(row["predicate_id"])],
+            }
+            m_list.append(json_obj)
+
+    return m_list
+
+
 # Support methods
 
 
@@ -506,6 +552,8 @@ def get_writer_function(
         return write_json, output_format
     elif output_format == "fhir_json":
         return write_fhir_json, output_format
+    elif output_format == "ontoportal_json":
+        return write_ontoportal_json, output_format
     elif output_format == "owl":
         return write_owl, SSSOM_DEFAULT_RDF_SERIALISATION
     else:
@@ -559,3 +607,18 @@ def _get_separator(serialisation: Optional[str] = None) -> str:
             f"Unknown table format: {serialisation}, should be one of tsv or csv"
         )
     return sep
+
+
+def _resolve_url(prefixed_url_str: str, prefix_map: PrefixMap) -> str:
+    if not prefixed_url_str:
+        return prefixed_url_str
+
+    prefix_url = prefixed_url_str.split(":")
+    if len(prefix_url) != 2:
+        return prefixed_url_str
+    else:
+        return _resolve_prefix(prefix_url[0], prefix_map) + prefix_url[1]
+
+
+def _resolve_prefix(prefix_str, prefix_map: PrefixMap) -> str:
+    return prefix_map.get(prefix_str, prefix_str + ":")

tests/test_conversion.py

@@ -4,6 +4,7 @@
 import json
 import logging
 import unittest
+from typing import Dict
 
 from rdflib import Graph
 
@@ -13,6 +14,7 @@
 from sssom.writers import (
     to_dataframe,
     to_json,
+    to_ontoportal_json,
     to_owl_graph,
     to_rdf_graph,
     write_json,
@@ -51,6 +53,8 @@ def test_conversion(self):
                 logging.info("Testing JSON export")
                 self._test_to_json_dict(mdoc, test)
                 self._test_to_json(mdoc, test)
+                logging.info("Testing ontoportal JSON export")
+                self._test_to_ontoportal_json(mdoc, test)
 
     def _test_to_owl_graph(self, mdoc, test):
         msdf = to_mapping_set_dataframe(mdoc)
@@ -74,6 +78,17 @@ def _test_to_json(self, mdoc, test: SSSOMTestCase):
         with open(test.get_out_file("json"), "w") as file:
             write_json(msdf, file, serialisation="json")
 
+    def _test_to_ontoportal_json(self, mdoc, test: SSSOMTestCase):
+        msdf = to_mapping_set_dataframe(mdoc)
+        jsonob = to_ontoportal_json(msdf)
+        self.assertEqual(len(jsonob), test.ct_data_frame_rows)
+        first_ob: Dict = jsonob[0]
+        self.assertTrue("classes" in first_ob)
+        self.assertTrue(len(first_ob.get("classes")) == 2)
+        self.assertTrue("relation" in first_ob)
+        self.assertIsInstance(first_ob.get("relation"), list)
+        self.assertGreater(len(first_ob.get("relation")), 0)
+
     def _test_to_rdf_graph(self, mdoc, test):
         msdf = to_mapping_set_dataframe(mdoc)
         g = to_rdf_graph(msdf)

tests/test_writers.py

@@ -6,7 +6,14 @@
 from jsonasobj2 import JsonObj
 
 from sssom.parsers import parse_sssom_json, parse_sssom_rdf, parse_sssom_table
-from sssom.writers import write_fhir_json, write_json, write_owl, write_rdf, write_table
+from sssom.writers import (
+    write_fhir_json,
+    write_json,
+    write_ontoportal_json,
+    write_owl,
+    write_rdf,
+    write_table,
+)
 from tests.constants import data_dir as test_data_dir
 from tests.constants import test_out_dir
 
@@ -84,3 +91,18 @@ def test_write_sssom_owl(self):
         # FIXME this test doesn't test anything
         # TODO implement "read_owl" function
         self.assertEqual(1, 1)
+
+    def test_write_sssom_ontoportal_json(self):
+        """Test writing as ontoportal JSON."""
+        path = os.path.join(test_out_dir, "test_write_sssom_ontoportal_json.json")
+        with open(path, "w") as file:
+            write_ontoportal_json(self.msdf, file)
+
+        with open(path, "r") as file:
+            d: list = json.load(file)
+
+        self.assertEqual(
+            len(d),
+            self.mapping_count,
+            f"{path} has the wrong number of mappings.",
+        )