fix: handle dataset name duplication

api/migrations/versions/2025_12_10_0925-c0dc2235f174_add_index_to_dataset_name.py

@@ -0,0 +1,31 @@
+"""add_index_to_dataset_name
+
+Revision ID: c0dc2235f174
+Revises: 09cfdda155d1
+Create Date: 2025-12-10 09:25:54.480660
+
+"""
+from alembic import op
+import models as models
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = 'c0dc2235f174'
+down_revision = '09cfdda155d1'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('datasets', schema=None) as batch_op:
+        batch_op.create_index('dataset_tenant_name_idx', ['tenant_id', 'name'], unique=False)
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('datasets', schema=None) as batch_op:
+        batch_op.drop_index('dataset_tenant_name_idx')
+    # ### end Alembic commands ###

api/models/dataset.py

@@ -46,6 +46,7 @@ class Dataset(Base):
     __table_args__ = (
         sa.PrimaryKeyConstraint("id", name="dataset_pkey"),
         sa.Index("dataset_tenant_idx", "tenant_id"),
+        sa.Index("dataset_tenant_name_idx", "tenant_id", "name"),
         adjusted_json_index("retrieval_model_idx", "retrieval_model"),
     )
 

api/services/dataset_service.py

@@ -2356,7 +2356,15 @@ class DocumentService:
 
         cut_length = 18
         cut_name = documents[0].name[:cut_length]
-        dataset.name = cut_name + "..."
+        proposed_name = cut_name + "..."
+
+        # Check for duplicate names and auto-increment if needed
+        existing_dataset_names = db.session.scalars(
+            select(Dataset.name).where(Dataset.tenant_id == tenant_id, Dataset.name.like(f"{proposed_name}%"))
+        ).all()
+        if existing_dataset_names:
+            proposed_name = generate_incremental_name(existing_dataset_names, proposed_name)
+        dataset.name = proposed_name
         dataset.description = "useful for when you want to answer queries about the " + documents[0].name
         db.session.commit()