feat(medcat): CU-869b9h7y6 Add faster linker

medcat-v2/medcat/components/linking/only_primary_name_linker.py

@@ -0,0 +1,88 @@
+from typing import Iterator, Optional, Union
+import logging
+
+from medcat.tokenizing.tokens import MutableDocument, MutableEntity
+from medcat.components.linking.context_based_linker import Linker
+from medcat.components.linking.vector_context_model import (
+    PerDocumentTokenCache)
+from medcat.utils.defaults import StatusTypes
+from medcat.cdb import CDB
+from medcat.vocab import Vocab
+from medcat.config import Config
+
+
+logger = logging.getLogger(__name__)
+
+
+class OnlyPrimaryNamesLinker(Linker):
+    name = 'primary_name_only_linker'
+
+    def __init__(self, cdb: CDB, vocab: Vocab, config: Config) -> None:
+        super().__init__(cdb, vocab, config)
+        # don't need / use the context model
+        del self.context_model
+
+    def _process_entity_inference(
+            self, doc: MutableDocument,
+            entity: MutableEntity,
+            per_doc_valid_token_cache: PerDocumentTokenCache
+            ) -> Iterator[MutableEntity]:
+        cuis = entity.link_candidates
+        if not cuis:
+            return
+        # Check does it have a detected name
+        name = entity.detected_name
+        if name is None:
+            logger.info("No name detected for entity %s", entity)
+            return
+        cnf_l = self.config.components.linking
+        if cnf_l.filter_before_disamb:
+            cuis = [cui for cui in cuis if cnf_l.filters.check_filters(cui)]
+        if not cuis:
+            logger.debug("No CUIs that fit filter for %s", entity)
+            return
+        if len(cuis) == 1:
+            if cnf_l.filters.check_filters(cuis[0]):
+                logger.info("Choosing only possible CUI %s for %s",
+                            cuis[0], entity)
+                entity.cui = cuis[0]
+                entity.context_similarity = 1.0
+                yield entity
+            else:
+                logger.info(
+                    "A single CUI (%s) was mapped to for %s but not in filter",
+                    cuis[0], entity)
+            return
+        primary_cuis = [cui for cui in cuis
+                        if (self.cdb.name2info[name]['per_cui_status'][cui]
+                            in StatusTypes.PRIMARY_STATUS and
+                            cnf_l.filters.check_filters(cui))]
+        if not primary_cuis:
+            logger.info("No pimary CUIs for name %s", name)
+            return
+        if len(primary_cuis) > 1:
+            logger.info(
+                "Ambiguous pimary CUIs for name %s: %s", name, primary_cuis)
+            return
+        cui = primary_cuis[0]
+        entity.cui = cui
+        entity.context_similarity = 1.0
+        yield entity
+
+    def train(self, cui: str,
+              entity: MutableEntity,
+              doc: MutableDocument,
+              negative: bool = False,
+              names: Union[list[str], dict] = [],
+              per_doc_valid_token_cache: Optional[PerDocumentTokenCache] = None
+              ) -> None:
+        raise NoTrainingException("Training is not supported for this linker")
+
+    def _train_on_doc(self, doc: MutableDocument,
+                      ner_ents: list[MutableEntity]
+                      ) -> Iterator[MutableEntity]:
+        raise NoTrainingException("Training is not supported for this linker")
+
+
+class NoTrainingException(ValueError):
+    pass

medcat-v2/medcat/components/types.py

@@ -213,6 +213,10 @@ def train(self, cui: str,
     "medcat2_embedding_linker": (
         "medcat.components.linking.embedding_linker",
         "Linker.create_new_component"),
+    # primary name only
+    "primary_name_only_linker": (
+        "medcat.components.linking.only_primary_name_linker",
+        "OnlyPrimaryNamesLinker.create_new_component"),
 }
 
 

medcat-v2/tests/components/linking/test_primary_name_only_linker.py

@@ -0,0 +1,33 @@
+import os
+
+from medcat.cdb import CDB
+from medcat.cat import CAT
+from medcat.vocab import Vocab
+from medcat.components.linking.only_primary_name_linker import (
+    OnlyPrimaryNamesLinker)
+
+import unittest
+
+from ... import UNPACKED_EXAMPLE_MODEL_PACK_PATH
+
+
+EXAMPLE_CDB_PATH = os.path.join(UNPACKED_EXAMPLE_MODEL_PACK_PATH, "cdb")
+EXAMPLE_VOCAB_PATH = os.path.join(UNPACKED_EXAMPLE_MODEL_PACK_PATH, "vocab")
+
+
+class PrimaryNamesLinkerTests(unittest.TestCase):
+    TEXT = (
+        "Man was diagnosed with severe kidney failure and acute diabetes "
+        "and presented with a light fever")
+
+    @classmethod
+    def setUpClass(cls):
+        vocab = Vocab.load(EXAMPLE_VOCAB_PATH)
+        cdb = CDB.load(EXAMPLE_CDB_PATH)
+        cdb.config.components.linking.comp_name = OnlyPrimaryNamesLinker.name
+        cls.cat = CAT(cdb, vocab)
+
+    def test_gets_entities(self):
+        ents = self.cat.get_entities(self.TEXT)
+        self.assertTrue(ents)
+        self.assertTrue(len(ents["entities"]))