From 86557b0e24c3de6089bdea64d90831bf0b4acf21 Mon Sep 17 00:00:00 2001
From: Charles Tapley Hoyt <cthoyt@gmail.com>
Date: Sun, 24 Nov 2024 13:34:25 +0100
Subject: [PATCH] Reader bug fixes - parent parsing and empty quotes (#245)

---
 src/pyobo/reader.py  | 24 ++++++++++++++++++------
 tests/test_get.py    |  6 +++++-
 tests/test_reader.py | 25 +++++++++++++++++++++++++
 3 files changed, 48 insertions(+), 7 deletions(-)

diff --git a/src/pyobo/reader.py b/src/pyobo/reader.py
index afe52d35..c15dfb44 100644
--- a/src/pyobo/reader.py
+++ b/src/pyobo/reader.py
@@ -165,7 +165,11 @@ def from_obonet(graph: nx.MultiDiGraph, *, strict: bool = True) -> Obo:
         alt_ids = list(iterate_node_alt_ids(data, strict=strict))
         n_alt_ids += len(alt_ids)
 
-        parents = list(iterate_node_parents(data, node=reference, strict=strict))
+        parents = list(
+            iterate_node_parents(
+                data, node=reference, strict=strict, ontology_prefix=ontology_prefix
+            )
+        )
         n_parents += len(parents)
 
         synonyms = list(
@@ -377,6 +381,8 @@ def _extract_definition(
 
 def get_first_nonescaped_quote(s: str) -> int | None:
     """Get the first non-escaped quote."""
+    if not s:
+        return None
     if s[0] == '"':
         # special case first position
         return 0
@@ -523,7 +529,7 @@ def _handle_prop(
         )
         if obj_reference is None:
             logger.warning(
-                "[%s:%s] could not parse object: %s", node.curie, prop_reference.curie, value_type
+                "[%s - %s] could not parse object: %s", node.curie, prop_reference.curie, value_type
             )
             return None
         # TODO can we drop datatype from this?
@@ -579,10 +585,13 @@ def iterate_node_parents(
     *,
     node: Reference,
     strict: bool = True,
+    ontology_prefix: str,
 ) -> Iterable[Reference]:
     """Extract parents from a :mod:`obonet` node's data."""
     for parent_curie in data.get("is_a", []):
-        reference = Reference.from_curie(parent_curie, strict=strict)
+        reference = Reference.from_curie(
+            parent_curie, strict=strict, ontology_prefix=ontology_prefix, node=node
+        )
         if reference is None:
             logger.warning("[%s] could not parse parent curie: %s", node.curie, parent_curie)
             continue
@@ -612,7 +621,9 @@ def iterate_node_relationships(
         if relation_curie in RELATION_REMAPPINGS:
             relation_prefix, relation_identifier = RELATION_REMAPPINGS[relation_curie]
         else:
-            relation_prefix, relation_identifier = normalize_curie(relation_curie, strict=strict)
+            relation_prefix, relation_identifier = normalize_curie(
+                relation_curie, strict=strict, ontology_prefix=ontology_prefix, node=node
+            )
         if relation_prefix is not None and relation_identifier is not None:
             relation = Reference(prefix=relation_prefix, identifier=relation_identifier)
         else:
@@ -623,8 +634,9 @@ def iterate_node_relationships(
                 relation.curie,
             )
 
-        # TODO replace with omni-parser from :mod:`curies`
-        target = Reference.from_curie(target_curie, strict=strict)
+        target = Reference.from_curie(
+            target_curie, strict=strict, ontology_prefix=ontology_prefix, node=node
+        )
         if target is None:
             logger.warning("[%s] %s could not parse target %s", node.curie, relation, target_curie)
             continue
diff --git a/tests/test_get.py b/tests/test_get.py
index 86587025..2bcb615a 100644
--- a/tests/test_get.py
+++ b/tests/test_get.py
@@ -203,7 +203,11 @@ def test_get_node_properties(self):
     def test_get_node_parents(self):
         """Test getting parents from a node in a :mod:`obonet` graph."""
         data = self.graph.nodes["CHEBI:51990"]
-        parents = list(iterate_node_parents(data, node=Reference(prefix="chebi", identifier="XXX")))
+        parents = list(
+            iterate_node_parents(
+                data, node=Reference(prefix="chebi", identifier="XXX"), ontology_prefix="chebi"
+            )
+        )
         self.assertEqual(2, len(parents))
         self.assertEqual({"24060", "51992"}, {parent.identifier for parent in parents})
         self.assertEqual({"chebi"}, {parent.prefix for parent in parents})
diff --git a/tests/test_reader.py b/tests/test_reader.py
index eb3476fc..6b8c792a 100644
--- a/tests/test_reader.py
+++ b/tests/test_reader.py
@@ -30,6 +30,7 @@ class TestUtils(unittest.TestCase):
 
     def test_first_nonescaped_quote(self):
         """Test finding the first non-escaped double quote."""
+        self.assertIsNone(get_first_nonescaped_quote(""))
         self.assertEqual(0, get_first_nonescaped_quote('"'))
         self.assertEqual(0, get_first_nonescaped_quote('"abc'))
         self.assertEqual(0, get_first_nonescaped_quote('"abc"'))
@@ -726,3 +727,27 @@ def test_synonym_url(self) -> None:
             ],
             synonym.provenance,
         )
+
+    def test_parent(self) -> None:
+        """Test parsing out a parent."""
+        ontology = _read("""\
+            ontology: chebi
+            date: 20:11:2024 18:44
+
+            [Term]
+            id: CHEBI:1234
+            is_a: CHEBI:5678
+        """)
+        term = self.get_only_term(ontology)
+        self.assertEqual([Reference(prefix="CHEBI", identifier="5678")], term.parents)
+
+        ontology = _read("""\
+            ontology: chebi
+            date: 20:11:2024 18:44
+
+            [Term]
+            id: CHEBI:1234
+            is_a: http://purl.obolibrary.org/obo/CHEBI_5678
+        """)
+        term = self.get_only_term(ontology)
+        self.assertEqual([Reference(prefix="CHEBI", identifier="5678")], term.parents)