set_label will INSERT if label not already presented. Fixes #788 (#789)

src/oaklib/implementations/aggregator/aggregator_implementation.py

@@ -1,3 +1,4 @@
+import logging
 from collections import defaultdict
 from dataclasses import dataclass
 from io import TextIOWrapper
@@ -140,6 +141,10 @@ def sssom_mappings(self, *args, **kwargs) -> Iterable[Mapping]:
  def label(self, curie: CURIE, **kwargs) -> str:
  return self._delegate_first(lambda i: i.label(curie, **kwargs))
 
+ def set_label(self, curie: CURIE, label: str) -> None:
+ logging.debug(f"Assuming {curie} is in first aggregated resource, label={label}")
+ return self._delegate_first(lambda i: i.set_label(curie, label))
+
  def curies_by_label(self, label: str) -> List[CURIE]:
  return list(self._delegate_iterator(lambda i: i.curies_by_label(label)))
 

src/oaklib/implementations/sqldb/sql_implementation.py

@@ -788,11 +788,19 @@ def _execute(self, stmt):
  self.save()
 
  def set_label(self, curie: CURIE, label: str) -> bool:
- stmt = (
- update(Statements)
- .where(and_(Statements.subject == curie, Statements.predicate == LABEL_PREDICATE))
- .values(value=label)
- )
+ existing_label = self.label(curie)
+ if existing_label:
+ stmt = (
+ update(Statements)
+ .where(and_(Statements.subject == curie, Statements.predicate == LABEL_PREDICATE))
+ .values(value=label)
+ )
+ else:
+ stmt = (
+ insert(Statements)
+ .values(subject=curie, predicate=LABEL_PREDICATE, value=label)
+ .execution_options(autocommit=True)
+ )
  self._execute(stmt)
 
  def basic_search(self, search_term: str, config: SearchConfiguration = None) -> Iterable[CURIE]:

src/oaklib/utilities/lexical/lexical_indexer.py

@@ -79,9 +79,8 @@ def add_labels_from_uris(oi: BasicOntologyInterface):
  if curie.startswith("<http"):
  label = label.replace(">", "")
  label = " ".join(label.split("_"))
- # print(f'{curie} ==> {label} // {type(oi)}')
  oi.set_label(curie, label)
- # print(oi.get_label_by_curie(curie))
+
 
 
 def create_lexical_index(
@@ -127,6 +126,7 @@ def _invert_mapping_pred(mapping_pred: PRED_CURIE) -> PRED_CURIE:
  logging.info("Creating mapping index")
  mapping_pairs_by_curie = defaultdict(list)
  for curie in oi.entities():
+ logging.debug(f"Finding mappings for {curie}")
  pairs = list(oi.simple_mappings_by_curie(curie))
  for pred, object_id in pairs:
  mapping_pairs_by_curie[curie].append((pred, object_id))