feat(history): Implement map_db_messages_to_langchain utility

backend/tests/test_history_mapping.py

@@ -0,0 +1,43 @@
+import pytest
+from ea_chatbot.history.models import Message
+from ea_chatbot.history.utils import map_db_messages_to_langchain
+from langchain_core.messages import HumanMessage, AIMessage
+
+def test_map_db_messages_to_langchain():
+    """
+    Test that DB Message objects are correctly mapped to LangChain message objects.
+    """
+    # 1. Setup mock DB messages
+    m1 = Message(role="user", content="Hello")
+    m2 = Message(role="assistant", content="Hi there!")
+    m3 = Message(role="user", content="How are you?")
+    
+    db_messages = [m1, m2, m3]
+    
+    # 2. Map
+    lc_messages = map_db_messages_to_langchain(db_messages)
+    
+    # 3. Assertions
+    assert len(lc_messages) == 3
+    
+    assert isinstance(lc_messages[0], HumanMessage)
+    assert lc_messages[0].content == "Hello"
+    
+    assert isinstance(lc_messages[1], AIMessage)
+    assert lc_messages[1].content == "Hi there!"
+    
+    assert isinstance(lc_messages[2], HumanMessage)
+    assert lc_messages[2].content == "How are you?"
+
+def test_map_db_messages_unknown_role():
+    """
+    Test that unknown roles are handled gracefully (e.g., defaulted to HumanMessage or raising error).
+    Let's assume we default to HumanMessage for safety, or just skip.
+    """
+    m = Message(role="system", content="System instruction")
+    lc_messages = map_db_messages_to_langchain([m])
+    
+    # Depending on implementation choice. Let's say we support 'system' too.
+    from langchain_core.messages import SystemMessage
+    assert isinstance(lc_messages[0], SystemMessage)
+    assert lc_messages[0].content == "System instruction"