welearn-database 0.2.6__tar.gz → 0.2.7__tar.gz

{welearn_database-0.2.6 → welearn_database-0.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: welearn-database
-Version: 0.2.6
+Version: 0.2.7
 Summary: All stuff related to relationnal database from the WeLearn project
 License: cc-by-sa-nc
 Author: Théo

{welearn_database-0.2.6 → welearn_database-0.2.7}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "welearn-database"
-version = "0.2.6"
+version = "0.2.7"
 description = "All stuff related to relationnal database from the WeLearn project"
 authors = [
     {name = "Théo",email = "theo.nardin@cri-paris.org"}

{welearn_database-0.2.6 → welearn_database-0.2.7}/welearn_database/data/models/document_related.py

@@ -54,7 +54,7 @@ class WeLearnDocument(Base):
     description: Mapped[str | None]
     full_content: Mapped[str | None]
     details: Mapped[dict[str, Any] | None]
-    _trace: Mapped[int | None] = mapped_column(types.BIGINT)
+    trace: Mapped[int | None] = mapped_column(types.BIGINT)
     corpus_id: Mapped[UUID] = mapped_column(
         types.Uuid,
         ForeignKey(f"{DbSchemaEnum.CORPUS_RELATED.value}.corpus.id"),
@@ -101,10 +101,15 @@ class WeLearnDocument(Base):
         :raises ValueError: If the full content is too short.
         """
         if not value:
+            self.trace = None
             return value
+        cleaned = clean_text(value)
         if len(value) < 25:
             raise ValueError(f"Content is too short : {len(value)}")
-        return clean_text(value)
+
+        # Hash compute and db storage
+        self.trace = adler32(cleaned.encode("utf-8"))
+        return cleaned
 
     @validates("description")
     def validate_description(self, key, value):
@@ -118,13 +123,6 @@ class WeLearnDocument(Base):
             return value
         return clean_text(value)
 
-    @property
-    def trace(self):
-        if self.full_content:
-            return adler32(bytes(self.full_content, "utf-8"))
-        else:
-            return None
-
 
 class ProcessState(Base):
     """