test: verify suicide_attempt_modality and fix typos

Co-Authored-By: Ariel Cohen <[email protected]>
aphp · Jul 9, 2024 · da4321c · da4321c
1 parent 53c79f0
commit da4321c
Show file tree

Hide file tree

Showing 3 changed files with 32 additions and 24 deletions.
diff --git a/edsnlp/pipes/ner/suicide_attempt/suicide_attempt.py b/edsnlp/pipes/ner/suicide_attempt/suicide_attempt.py
@@ -17,25 +17,25 @@ class SuicideAttemptMatcher(BaseNERComponent):
     The `eds.suicide_attempt` pipeline component detects mentions of Suicide Attempt.
     It can be used with a span qualifier
     for contextualisation of the entity (history) and to detect false positives as
-    negation, hypothesis or family. We recomend to use a Machine Learning qualifier
+    negation, hypothesis or family. We recommend to use a machine learning qualifier
     to disambiguate polysemic words, as proposed in [@bey_natural_2024].
 
 
-    It sets the entity label `suicide_attempt`.
+    Every matched entity will be labelled `suicide_attempt`.
 
     Extensions
     ----------
     Each entity span will have the suicide attempt modality as an attribute.
     The available modalities are:
 
-        - suicide_attempt_unspecific
-        - autolysis
-        - intentional_drug_overdose
-        - jumping_from_height
-        - cuts
-        - strangling
-        - self_destructive_behavior
-        - burn_gas_caustic
+    - `suicide_attempt_unspecific`
+    - `autolysis`
+    - `intentional_drug_overdose`
+    - `jumping_from_height`
+    - `cuts`
+    - `strangling`
+    - `self_destructive_behavior`
+    - `burn_gas_caustic`
 
     Examples
     --------
@@ -55,7 +55,7 @@ class SuicideAttemptMatcher(BaseNERComponent):
     # Out: 'intentional_drug_overdose'
     ```
 
-    ??? info "Patterns used for the NER"
+    ??? info "Patterns used for the named entity recognition"
         ```{ .python .no-check }
         # fmt: off
         --8<-- "edsnlp/pipes/ner/suicide_attempt/patterns.py"

diff --git a/edsnlp/utils/examples.py b/edsnlp/utils/examples.py
@@ -17,7 +17,7 @@ class Modifier(BaseModel):
     key: str
     value: Union[int, float, bool, str, Dict[str, Any]]
 
-    @validator("value")
+    @validator("value", allow_reuse=True)
     def optional_dict_parsing(cls, v):
         try:
             return loads(v)
@@ -30,6 +30,10 @@ class Entity(BaseModel):
     end_char: int
     modifiers: List[Modifier]
 
+    @property
+    def modifiers_dict(self) -> Dict[str, Any]:
+        return {m.key: m.value for m in self.modifiers}
+
 
 entity_pattern = re.compile(r"(<ent[^<>]*>[^<>]+</ent>)", flags=re.DOTALL)
 text_pattern = re.compile(r"<ent.*>(.+)</ent>", flags=re.DOTALL)

diff --git a/tests/pipelines/ner/test_suicide_attempt.py b/tests/pipelines/ner/test_suicide_attempt.py
@@ -1,17 +1,17 @@
-from edsnlp.utils.examples import parse_example
+from edsnlp.utils.examples import Entity, parse_example
 
 examples = [
-    "J'ai vu le patient à cause d'une <ent>TS</ent> médicamenetuse."
-    "J'ai vu le patient à cause d'une ts médicamenetuse.",
-    "J'ai vu le patient à cause d'une <ent>IMV</ent>.",
-    "surface TS",
-    "Patiente hospitalisée à cause d'une <ent>Tentative d'autolyse</ent>.",
-    "Le patient exprime des idées de défenestration",
-    "vu aux urgences suite à une <ent>défenestration volontaire</ent>",
-    "amené par les pompiers à cause d'une <ent>phlebotomie</ent>",
-    "Antécédents :\n- <ent>pendaison</ent> (2010)",
-    "copain du patient : plusieurs événements d'<ent>autodestruction</ent>",
-    "suspicion d'<ent>ingestion de caustique</ent> avec des idées suicidaires",
+    "J'ai vu le patient à cause d'une <ent modality=suicide_attempt_unspecific>TS</ent> médicamenetuse."  # noqa: E501
+    "J'ai vu le patient à cause d'une ts médicamenetuse.",  # noqa: E501
+    "J'ai vu le patient à cause d'une <ent modality=intentional_drug_overdose>IMV</ent>.",  # noqa: E501
+    "surface TS",  # noqa: E501
+    "Patiente hospitalisée à cause d'une <ent modality=autolysis>Tentative d'autolyse</ent>.",  # noqa: E501
+    "Le patient exprime des idées de défenestration",  # noqa: E501
+    "vu aux urgences suite à une <ent modality=jumping_from_height>défenestration volontaire</ent>",  # noqa: E501
+    "amené par les pompiers à cause d'une <ent modality=cuts>phlebotomie</ent>",  # noqa: E501
+    "Antécédents :\n- <ent modality=strangling>pendaison</ent> (2010)",  # noqa: E501
+    "copain du patient : plusieurs événements d'<ent modality=self_destructive_behavior>autodestruction</ent>",  # noqa: E501
+    "suspicion d'<ent modality=burn_gas_caustic>ingestion de caustique</ent> avec des idées suicidaires",  # noqa: E501
 ]
 
 
@@ -24,4 +24,8 @@ def test_suicide_attempt(blank_nlp):
         assert len(doc.ents) == len(entities)
 
         for ent, entity in zip(doc.ents, entities):
+            entity: Entity
             assert ent.text == text[entity.start_char : entity.end_char]
+            assert ent._.suicide_attempt_modality == entity.modifiers_dict.get(
+                "modality"
+            )