Add ignore_separators option to by_name()

pipeline/src/additional_methods/by_name.py.txt

@@ -10,6 +10,7 @@
         all: bool = False,
         case_sensitive: bool = True,
         ignore_accents: bool = False,
+        ignore_separators: bool = False,
     ):
         """
         Search for instances in the openMINDS instance library based on their name.
@@ -29,6 +30,9 @@
                 other diacritical marks (cedilla, tilde, ring, etc.) when matching. Also treat
                 special letters (ß, œ, æ, ø, ł, etc.) as their closest plain-letter equivalents
                 (e.g. "ß" as "ss"). Defaults to False.
+            ignore_separators (bool, optional): Whether to ignore hyphens ("-"), underscores
+                ("_"), slashes ("/"), and repeated whitespace when matching, by collapsing
+                them all to a single space. Defaults to False.
         """
         namelike_properties = ("name", "lookup_label", "family_name", "full_name", "short_name", "abbreviation")
         if cls._instance_lookup is None:
@@ -72,25 +76,31 @@
                 s = s.casefold()
             if ignore_accents:
                 s = remove_accents(s)
+            if ignore_separators:
+                s = s.replace("-", " ").replace("_", " ").replace("/", " ")
+                s = " ".join(s.split())
             return s
 
         if match == "equals":
-            if case_sensitive and not ignore_accents:
+            if case_sensitive and not ignore_accents and not ignore_separators:
                 matches = cls._instance_lookup.get(name, [])
             else:
+                normalized_name = normalize(name)
                 matches = []
                 for key, instances in cls._instance_lookup.items():
-                    if normalize(key) == normalize(name):
+                    if normalize(key) == normalized_name:
                         matches.extend(instances)
         elif match == "contains":
+            normalized_name = normalize(name)
             matches = []
             for key, instances in cls._instance_lookup.items():
-                if normalize(name) in normalize(key):
+                if normalized_name in normalize(key):
                     matches.extend(instances)
         elif match == "within":
+            normalized_name = normalize(name)
             matches = []
             for key, instances in cls._instance_lookup.items():
-                if normalize(key) in normalize(name):
+                if normalize(key) in normalized_name:
                     matches.extend(instances)
         else:
             raise ValueError("'match' must be either 'equals', 'contains', or 'within'")

pipeline/tests/test_regressions.py

@@ -763,3 +763,32 @@ def test_pr0103_by_name_ignore_accents(om):
     for query, ignore_accents, expected_name in special_letter_cases:
         match = SovereignState.by_name(query, ignore_accents=ignore_accents)
         assert (match.name if match else None) == expected_name
+
+
+@pytest.mark.parametrize("om", [openminds.latest])
+def test_pr0105_by_name_ignore_separators(om):
+    # https://github.com/openMetadataInitiative/openMINDS_Python/pull/105
+    # by_name(..., ignore_separators=True) treats hyphens and underscores as spaces,
+    # and collapses repeated/mixed whitespace, before matching
+    Technique = om.controlled_terms.Technique
+
+    # (query, ignore_separators, expected match name or None)
+    cases = [
+        ("two-photon fluorescence microscopy", False, "two-photon fluorescence microscopy"),   # exact
+        ("two photon fluorescence microscopy", True, "two-photon fluorescence microscopy"),
+        ("two_photon_fluorescence_microscopy", True, "two-photon fluorescence microscopy"),
+        ("two   photon  fluorescence-microscopy", True, "two-photon fluorescence microscopy"),
+        ("two photon fluorescence microscopy", False, None),       # defaults: hyphen still matters
+        ("CLARITY/TDE", False, "CLARITY/TDE"),  # exact
+        ("CLARITY TDE", True, "CLARITY/TDE"),
+        ("CLARITY-TDE", True, "CLARITY/TDE"),
+        ("CLARITY TDE", False, None),           # defaults: slash still matters
+    ]
+    for query, ignore_separators, expected_name in cases:
+        match = Technique.by_name(query, ignore_separators=ignore_separators)
+        assert (match.name if match else None) == expected_name
+
+    # ignore_separators also applies to match="within"
+    assert Technique.by_name("CLARITY-TDE method", match="within") is None
+    match = Technique.by_name("CLARITY-TDE method", match="within", ignore_separators=True)
+    assert match.name == "CLARITY/TDE"