From 12247ab2a21623497013be323f4e80011cee4722 Mon Sep 17 00:00:00 2001
From: Claude <noreply@anthropic.com>
Date: Wed, 21 Oct 2026 17:02:00 +0000
Subject: [PATCH] lookup(): trim the prefix and remainder as MATLAB does

getOntologyNameFromPrefix (ontology.m:493) strtrims both halves of the
lookup string before mapping the prefix and calling lookupTermOrID.
Python matched the untrimmed prefix and passed the untrimmed remainder,
so 'CL: 0000540' and ' CL:0000540' failed in Python and resolved in
MATLAB. The cache stays keyed on the string as typed, as MATLAB's is.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UaMQ8AiBeK9Bj7tZ7U1CS2
---
 src/ndi_ontology/__init__.py                   |  7 +++++--
 src/ndi_ontology/ndi_matlab_python_bridge.yaml |  8 ++++++++
 tests/test_ontology.py                         | 18 ++++++++++++++++++
 3 files changed, 31 insertions(+), 2 deletions(-)

diff --git a/src/ndi_ontology/__init__.py b/src/ndi_ontology/__init__.py
index 9d09800..81cca1c 100644
--- a/src/ndi_ontology/__init__.py
+++ b/src/ndi_ontology/__init__.py
@@ -225,7 +225,10 @@ def lookup(lookup_string: str) -> OntologyResult:
             f"{lookup_string!r} has no ontology prefix; expected 'PREFIX:term'"
         )
 
-    prefix, remainder = lookup_string.split(":", 1)
+    # MATLAB trims both halves before mapping the prefix or calling the
+    # provider (getOntologyNameFromPrefix, ontology.m:493), so ' CL : neuron'
+    # resolves there. The cache stays keyed on the string as typed, as MATLAB's is.
+    prefix, remainder = (part.strip() for part in lookup_string.split(":", 1))
 
     prefix_map = _load_prefix_map()
 
@@ -274,7 +277,7 @@ def lookup(lookup_string: str) -> OntologyResult:
     # NDI's subject tables name treatment columns with it, so a provider
     # returning OLS's short_form or '' gave Python different columns from
     # MATLAB for the same dataset.
-    result.prefix = prefix.strip()
+    result.prefix = prefix
     result.synonyms = [
         s
         for item in result.synonyms
diff --git a/src/ndi_ontology/ndi_matlab_python_bridge.yaml b/src/ndi_ontology/ndi_matlab_python_bridge.yaml
index 4219709..4c695b9 100644
--- a/src/ndi_ontology/ndi_matlab_python_bridge.yaml
+++ b/src/ndi_ontology/ndi_matlab_python_bridge.yaml
@@ -179,6 +179,14 @@ classes:
           'taxonomy:9606'). NDI-python names treatment columns in its subject
           tables from shortName, so the same dataset got different columns
           in each language. The shared-case runner now asserts both fields.
+          FIXED 2026-10-21: lookup() now trims the prefix and the remainder
+          before mapping the prefix and calling the provider, as
+          getOntologyNameFromPrefix does (ontology.m:493, read at 44efcec).
+          Python used to match the untrimmed prefix and pass the untrimmed
+          remainder, so 'CL: 0000540' and ' CL:0000540' failed in Python and
+          resolved in MATLAB. The cache is still keyed on the string as typed,
+          as MATLAB's is. Python's str.strip() stands in for strtrim; the two
+          differ only on NUL characters, which strtrim also removes.
 
       # --- Instance: lookupTermOrID ---
       - name: lookupTermOrID
diff --git a/tests/test_ontology.py b/tests/test_ontology.py
index 00c9241..e17bc46 100644
--- a/tests/test_ontology.py
+++ b/tests/test_ontology.py
@@ -170,6 +170,24 @@ def test_case_insensitive_prefix(self):
             lookup("cl:1")  # lowercase
             assert mock_lt.called
 
+    def test_remainder_is_trimmed_before_the_provider_sees_it(self):
+        """ontology.m:493 -- remainder = strtrim(...) before lookupTermOrID."""
+        from ndi_ontology import OntologyResult, lookup
+
+        with patch("ndi_ontology.providers.OLSProvider.lookup_term") as mock_lt:
+            mock_lt.return_value = OntologyResult(id="CL:0000540", name="neuron")
+            lookup("CL:  0000540 \t")
+        mock_lt.assert_called_once_with("0000540", "CL")
+
+    @pytest.mark.parametrize("typed", ["NDIC: 8", " NDIC:8", "NDIC :8", "  ndic : 8  "])
+    def test_whitespace_around_prefix_or_term_resolves(self, typed):
+        """ontology.m:493 trims both halves; the prefix before it is mapped."""
+        from ndi_ontology import lookup
+
+        result = lookup(typed)
+        assert result.name == "Experienced cricket hunting"
+        assert result.prefix == typed.split(":", 1)[0].strip()
+
     def test_provider_exception_is_chained(self):
         """A provider error surfaces as NDIOntologyLookupError, cause intact.