minor refactoring in ensemble.py so the geofence can be loaded without loading the entire model

.github/workflows/python_style_checks.yml

@@ -12,7 +12,7 @@ jobs:
       fail-fast: false
       matrix:
         os: [ubuntu-latest]
-        python-version: ['3.9', '3.10', '3.11', '3.12']
+        python-version: ['3.10', '3.11', '3.12']
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4

pyproject.toml

@@ -51,6 +51,7 @@ dependencies = [
     "tqdm",
     "torch >= 2.0",
     "yolov5 >= 7.0.8, < 7.0.12",
+    "setuptools<82"
 ]
 
 [project.optional-dependencies]

speciesnet/ensemble.py

@@ -39,6 +39,32 @@
 PredictionType = tuple[PredictionLabelType, PredictionScoreType, PredictionSourceType]
 
 
+def _load_taxonomy_from_file(taxonomy_file):
+    """Loads the taxonomy .txt file into a dict."""
+
+    def _taxa_from_label(label: str) -> str:
+        return ";".join(label.split(";")[1:6])
+
+    # Create taxonomy map.
+    with open(taxonomy_file, mode="r", encoding="utf-8") as fp:
+        labels = [line.strip() for line in fp.readlines()]
+        taxonomy_map = {_taxa_from_label(label): label for label in labels}
+
+        for label in [
+            Classification.BLANK,
+            Classification.VEHICLE,
+            Classification.UNKNOWN,
+        ]:
+            taxa = _taxa_from_label(label)
+            if taxa in taxonomy_map:
+                del taxonomy_map[taxa]
+
+        for label in [Classification.HUMAN, Classification.ANIMAL]:
+            taxa = _taxa_from_label(label)
+            taxonomy_map[taxa] = label
+    return taxonomy_map
+
+
 class SpeciesNetEnsemble:
     """Ensemble component of SpeciesNet."""
 
@@ -76,27 +102,7 @@ def __init__(
     def load_taxonomy(self):
         """Loads the taxonomy from the model info."""
 
-        def _taxa_from_label(label: str) -> str:
-            return ";".join(label.split(";")[1:6])
-
-        # Create taxonomy map.
-        with open(self.model_info.taxonomy, mode="r", encoding="utf-8") as fp:
-            labels = [line.strip() for line in fp.readlines()]
-            taxonomy_map = {_taxa_from_label(label): label for label in labels}
-
-            for label in [
-                Classification.BLANK,
-                Classification.VEHICLE,
-                Classification.UNKNOWN,
-            ]:
-                taxa = _taxa_from_label(label)
-                if taxa in taxonomy_map:
-                    del taxonomy_map[taxa]
-
-            for label in [Classification.HUMAN, Classification.ANIMAL]:
-                taxa = _taxa_from_label(label)
-                taxonomy_map[taxa] = label
-        return taxonomy_map
+        return _load_taxonomy_from_file(self.model_info.taxonomy)
 
     def load_geofence(self):
         """Loads the geofence map from the model info."""

speciesnet/scripts/build_geofence_release.py

@@ -71,6 +71,7 @@
   may later get blocked
 
 """
+
 import copy
 import json
 import os

speciesnet/utils_test.py

@@ -481,8 +481,7 @@ def test_successful_loading(self, valid_predictions_json) -> None:
     def invalid_predictions_json(self, tmp_path) -> Generator[Path, None, None]:
         filepath = tmp_path / "invalid_predictions.json"
         with open(filepath, mode="w", encoding="utf-8") as fp:
-            fp.write(
-                """{
+            fp.write("""{
                     "predictions": [
                         {
                             "filepath": "a.jpg",
@@ -492,8 +491,7 @@ def invalid_predictions_json(self, tmp_path) -> Generator[Path, None, None]:
                             }
                         }
                     ]
-                }"""
-            )
+                }""")
         yield filepath
         filepath.unlink()