temporal-normalization-spacy 2.0.1__tar.gz → 2.0.2__tar.gz

{temporal_normalization_spacy-2.0.1/temporal_normalization_spacy.egg-info → temporal_normalization_spacy-2.0.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: temporal_normalization_spacy
-Version: 2.0.1
+Version: 2.0.2
 Summary: A spaCy plugin for identifying and parsing historical data in Romanian texts
 Home-page: https://github.com/iliedorobat/timespan-normalization-spacy
 Author: Ilie Cristian Dorobat
@@ -22,7 +22,7 @@ Dynamic: requires-dist
 Dynamic: requires-python
 Dynamic: summary
 
-# Temporal Expressions Normalization spaCy (TeNs)
+# Temporal Expressions Normalization for spaCy (TeNs)
 
 <b>Temporal Expressions Normalization spaCy (TeNs)</b> is a powerful pipeline component for spaCy that seamlessly
 identifies and parses date entities in text. It leverages the <b>[Temporal Expressions Normalization Framework](

{temporal_normalization_spacy-2.0.1 → temporal_normalization_spacy-2.0.2}/README.md

@@ -1,4 +1,4 @@
-# Temporal Expressions Normalization spaCy (TeNs)
+# Temporal Expressions Normalization for spaCy (TeNs)
 
 <b>Temporal Expressions Normalization spaCy (TeNs)</b> is a powerful pipeline component for spaCy that seamlessly
 identifies and parses date entities in text. It leverages the <b>[Temporal Expressions Normalization Framework](

{temporal_normalization_spacy-2.0.1 → temporal_normalization_spacy-2.0.2}/setup.py

@@ -2,10 +2,10 @@ from setuptools import setup, find_packages
 
 setup(
     name="temporal_normalization_spacy",
-    version="2.0.1",
+    version="2.0.2",
     author="Ilie Cristian Dorobat",
     description="A spaCy plugin for identifying and parsing historical data "
-                "in Romanian texts",
+    "in Romanian texts",
     long_description=open("README.md").read(),
     long_description_content_type="text/markdown",
     url="https://github.com/iliedorobat/timespan-normalization-spacy",

{temporal_normalization_spacy-2.0.1 → temporal_normalization_spacy-2.0.2}/temporal_normalization/commons/temporal_models.py

@@ -29,7 +29,13 @@ class TemporalExpression:
             TimeSeries(item) for item in json_obj["timeSeries"]
         ] if self.is_valid else []
         self.matches: list[str] = list(
-            set([matched_value for ts in self.time_series for matched_value in ts.matches])
+            set(
+                [
+                    matched_value
+                    for ts in self.time_series
+                    for matched_value in ts.matches
+                ]
+            )
         )
 
     def __str__(self):

{temporal_normalization_spacy-2.0.1 → temporal_normalization_spacy-2.0.2}/temporal_normalization/index.py

@@ -129,10 +129,10 @@ def _retokenize(
             if start_token is not None and end_token is not None:
                 # use exact token boundaries to create a custom `Span` for well-defined
                 # time expressions with known character offsets.
-                entity, exists = _create_span(doc, start_char, end_char, start_token, end_token)
+                entity, existed_entity = _create_span(doc, start_char, end_char, start_token, end_token)
                 time_series: list[TimeSeries] = [ts for expression in expressions for ts in expression.time_series]
                 matched_ts = [ts for ts in time_series if _matched(entity.text, ts.matches)]
-                _assign_time_series(matched_ts, retokenized_entities, retokenizer, doc, entity, exists)
+                _retokenize_entity(doc, matched_ts, entity, existed_entity, retokenized_entities, retokenizer)
             else:
                 # For more ambiguous or loosely defined expressions, such as "martie -iunie 2013"
                 # or "dintre secolele al XV-lea și al XVIII-lea", iterates through existing entities
@@ -142,24 +142,26 @@ def _retokenize(
                     if entity not in retokenized_entities:
                         time_series: list[TimeSeries] = [ts for expression in expressions for ts in expression.time_series]
                         matched_ts = [ts for ts in time_series if _is_substring(entity.text, ts.matches)]
-                        _assign_time_series(matched_ts, retokenized_entities, retokenizer, doc, entity, True)
+                        _retokenize_entity(doc, matched_ts, entity, True, retokenized_entities, retokenizer)
 
 
-def _assign_time_series(
+def _retokenize_entity(
+    doc: Doc,
     matched_ts: list[TimeSeries],
+    entity: Span,
+    existed_entity: bool,
     retokenized_entities: list[Span],
     retokenizer: Doc.retokenize,
-    doc: Doc,
-    entity: Span,
-    exists: bool
 ) -> None:
     """
-    Attaches matched TimeSeries to a given entity and updates the doc's entity list.
+    Retokenizes and enriches a temporal entity span with matched time series data.
+    Updates the Doc with the new entity and merges it if needed.
 
     Args:
         doc (Doc): The processed spaCy document.
+        matched_ts (list[TimeSeries]): The matched time series.
         entity (Span): The named entity to enrich.
-        exists (bool): Whether the entity already exists in doc.ents.
+        existed_entity (bool): Whether the entity already exists in doc.ents.
         retokenized_entities (list): Accumulator for entities that require retokenization.
         retokenizer (Doc.retokenize): The spaCy retokenizer context.
     """
@@ -167,11 +169,38 @@ def _assign_time_series(
     if not len(matched_ts):
         return None
 
-    if exists:
+    _assign_time_series(matched_ts, entity, existed_entity)
+    _update_doc_ents(doc, entity)
+    _merge_entity(doc, entity, retokenized_entities, retokenizer)
+
+
+def _assign_time_series(
+    matched_ts: list[TimeSeries], entity: Span, existed_entity: bool
+) -> None:
+    """
+    Attaches matched TimeSeries to a given entity.
+
+    Args:
+        matched_ts (list[TimeSeries]): The matched time series.
+        entity (Span): The named entity to enrich.
+        existed_entity (bool): Whether the entity already exists in doc.ents.
+    """
+
+    if existed_entity:
         entity._.time_series = matched_ts
     else:
         entity._.set("time_series", matched_ts)
 
+
+def _update_doc_ents(doc: Doc, entity: Span) -> None:
+    """
+    Updates the doc's entity list
+
+    Args:
+        doc (Doc): The processed spaCy document.
+        entity (Span): The named entity to enrich.
+    """
+
     all_ents = list(doc.ents)
     if entity not in all_ents:
         # E.g.: entity in all_ents => "Ecaterina Balș ( 22 iulie 1814 - august 1887 ) - născută în familia Dimachi , a doua soție a generalului Teodor Balș ( 1805-1857 ) , caimacam al Moldovei în perioada 1856-1857 , cu care se căsătorise la 30 iunie 1846 la Dimăcheni ( Dorohoi ) ."
@@ -180,7 +209,24 @@ def _assign_time_series(
 
     doc.ents = filter_spans(all_ents)
 
-    if entity not in all_ents:
+
+def _merge_entity(
+    doc: Doc,
+    entity: Span,
+    retokenized_entities: list[Span],
+    retokenizer: Doc.retokenize,
+) -> None:
+    """
+    Merges a custom entity span into the spaCy Doc if it is not already part of
+    doc.ents, and tracks it in a list of retokenized entities.
+
+    Args:
+        entity (Span): The named entity to enrich.
+        retokenized_entities (list): Accumulator for entities that require retokenization.
+        retokenizer (Doc.retokenize): The spaCy retokenizer context.
+    """
+
+    if entity not in doc.ents:
         retokenized_entities.append(entity)
         retokenizer.merge(entity)
 
@@ -226,7 +272,9 @@ def _is_substring(text: str, matches: list[str]) -> bool:
     return False
 
 
-def _create_span(doc: Doc, start_char: int, end_char: int, start_token: int, end_token: int) -> tuple[Span, bool]:
+def _create_span(
+    doc: Doc, start_char: int, end_char: int, start_token: int, end_token: int
+) -> tuple[Span, bool]:
     """
     Creates a new span for a temporal expression or returns an existing overlapping entity.
 

{temporal_normalization_spacy-2.0.1 → temporal_normalization_spacy-2.0.2/temporal_normalization_spacy.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: temporal_normalization_spacy
-Version: 2.0.1
+Version: 2.0.2
 Summary: A spaCy plugin for identifying and parsing historical data in Romanian texts
 Home-page: https://github.com/iliedorobat/timespan-normalization-spacy
 Author: Ilie Cristian Dorobat
@@ -22,7 +22,7 @@ Dynamic: requires-dist
 Dynamic: requires-python
 Dynamic: summary
 
-# Temporal Expressions Normalization spaCy (TeNs)
+# Temporal Expressions Normalization for spaCy (TeNs)
 
 <b>Temporal Expressions Normalization spaCy (TeNs)</b> is a powerful pipeline component for spaCy that seamlessly
 identifies and parses date entities in text. It leverages the <b>[Temporal Expressions Normalization Framework](