Merge branch 'develop'

Author: rhnfzl

sct/sct.py

@@ -80,7 +80,11 @@ def __init__(self, cfg: Optional[TextCleanerConfig] = None):
                 torch_model_names=torch_model_names,
                 ner_batch_size=self.cfg.ner_batch_size,
                 ensemble_models=dict(self.cfg.ner_ensemble) if self.cfg.ner_ensemble is not None else None,
-                ensemble_default_keys=tuple(self.cfg.ner_ensemble_default_keys) if self.cfg.ner_ensemble_default_keys is not None else None,
+                ensemble_default_keys=(
+                    tuple(self.cfg.ner_ensemble_default_keys)
+                    if self.cfg.ner_ensemble_default_keys is not None
+                    else None
+                ),
             )
         else:
             self.GeneralNER = None

sct/utils/ner.py

@@ -70,7 +70,11 @@ def __init__(self, cache_dir: Optional[Path] = None, device: str = None,
         self._ner_batch_size = ner_batch_size
         self._gliner_pipe = None
         self._ensemble_models: Dict[str, tuple] = ensemble_models if ensemble_models is not None else DEFAULT_NER_ENSEMBLE
-        self._ensemble_default_keys: tuple = ensemble_default_keys if ensemble_default_keys is not None else NER_ENSEMBLE_DEFAULT_KEYS
+        self._ensemble_default_keys: tuple = (
+            ensemble_default_keys
+            if ensemble_default_keys is not None
+            else NER_ENSEMBLE_DEFAULT_KEYS
+        )
 
         # Device detection
         if device:

tests/test_sct.py

@@ -928,12 +928,12 @@ def test_aprocess_batch_uses_running_loop(self):
     def test_sentence_boundary_no_false_split_on_abbreviation(self):
         """SENTENCE_BOUNDARY_PATTERN must not split abbreviation-dot + space."""
         from sct.utils.constants import SENTENCE_BOUNDARY_PATTERN
-        for text in ["Dr. Smith visited Berlin.",
-                     "Mr. Jones called Mrs. Smith.",
-                     "The U.S. Army Base is here."]:
-            pieces = SENTENCE_BOUNDARY_PATTERN.split(text)
+        for abbrev_text in ["Dr. Smith visited Berlin.",
+                            "Mr. Jones called Mrs. Smith.",
+                            "The U.S. Army Base is here."]:
+            pieces = SENTENCE_BOUNDARY_PATTERN.split(abbrev_text)
             self.assertEqual(len(pieces), 1,
-                             f"Should not split abbreviation in: {text!r}")
+                             f"Should not split abbreviation in: {abbrev_text!r}")
 
     def test_sentence_boundary_splits_real_sentence(self):
         """SENTENCE_BOUNDARY_PATTERN should split genuine sentence boundaries."""