Move python tests to root (#242)

* Move tests to root

* Move pytest ini

.github/workflows/unittests.yml

@@ -0,0 +1,25 @@
+name: Tests
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+  workflow_call:
+
+jobs:
+  test_package:
+    name: Unit Tests
+    runs-on: "ubuntu-20.04"
+    steps:
+        - uses: actions/checkout@v4
+        - name: Setup python
+          uses: actions/setup-python@v5
+          with:
+            python-version: "3.11"
+            architecture: x64
+        - name: Install dependencies
+          run: |
+            pip install -r code/requirements.txt
+        - name: Run Python tests
+          run: python -m pytest --rootdir=code -m "not azure"

code/requirements.txt

@@ -25,4 +25,4 @@ pandas==1.5.1
 python-docx==1.1.0
 
 # Add dev dependencies here - this will be refactored out by Poetry
-pytest==7.4.4
+pytest==8.0.0

pytest.ini

@@ -0,0 +1,4 @@
+[pytest]
+markers =
+    unittest: Unit Tests (relatively fast)
+    azure: marks tests as extended (run less frequently, relatively slow)

code/utilities/tests/test_AzureBlobStorage.py → tests/test_AzureBlobStorage.py

@@ -1,10 +1,11 @@
 import pytest
-from ..helpers.AzureBlobStorageHelper import AzureBlobStorageClient 
+from code.utilities.helpers.AzureBlobStorageHelper import AzureBlobStorageClient
 
 @pytest.fixture
 def blob_client():
     return AzureBlobStorageClient()
 
+@pytest.mark.azure("This test requires Azure Blob Storage")
 def test_upload_and_download_file(blob_client):
     # Upload a file
     file_name = "test_file.txt"

code/utilities/tests/test_ContentSafetyChecker.py → tests/test_ContentSafetyChecker.py

@@ -1,6 +1,7 @@
 import pytest
-from ..tools.ContentSafetyChecker import ContentSafetyChecker
+from code.utilities.tools.ContentSafetyChecker import ContentSafetyChecker
 
+@pytest.mark.azure("This test requires Azure Content Safety configured")   
 def test_document_chunking_layout():
 
     cut = ContentSafetyChecker()

code/utilities/tests/test_DocumentChunking.py → tests/test_DocumentChunking.py

@@ -1,7 +1,7 @@
 import pytest
 from typing import List
-from ..common.SourceDocument import SourceDocument
-from ..helpers.DocumentChunkingHelper import DocumentChunking, ChunkingSettings, ChunkingStrategy
+from code.utilities.common.SourceDocument import SourceDocument
+from code.utilities.helpers.DocumentChunkingHelper import DocumentChunking, ChunkingSettings, ChunkingStrategy
 
 # Create a sample document
 documents = [
@@ -19,6 +19,7 @@
     )
 ]
 
+@pytest.mark.unittest("DocumentChunking")
 def test_document_chunking_layout():
     # Test layout chunking strategy
     chunking = ChunkingSettings({"strategy": ChunkingStrategy.LAYOUT, "size": 10, "overlap": 5})

code/utilities/tests/test_DocumentLoading.py → tests/test_DocumentLoading.py

@@ -1,8 +1,8 @@
 import pytest
 from typing import List
-from ..helpers.DocumentLoadingHelper import DocumentLoading, LoadingSettings
-
+from code.utilities.helpers.DocumentLoadingHelper import DocumentLoading, LoadingSettings
 
+@pytest.mark.azure("This test requires Azure Document Intelligence configured")
 def test_document_loading_layout():
     # Azure Form Recognizer Layout
     document_loading = DocumentLoading()
@@ -15,6 +15,7 @@ def test_document_loading_layout():
     assert data[4].page_number == 4
     assert data[4].source == url
 
+@pytest.mark.azure("This test requires Azure Document Intelligence configured")
 def test_document_loading_read():
     # Azure Form Recognizer Read
     document_loading = DocumentLoading()
@@ -34,7 +35,8 @@ def test_document_loading_web():
     data = document_loading.load(url, LoadingSettings({"strategy": "web"}))
     assert len(data) == 1
     assert data[0].source == url
-
+
+@pytest.mark.azure("This test requires Azure Document Intelligence configured")
 def test_document_loading_docx():
     document_loading = DocumentLoading()
     url = "https://csciblob.blob.core.windows.net/rag-sol-acc/What is Azure OpenAI Service.docx"

code/utilities/tests/test_DocumentProcessor.py → tests/test_DocumentProcessor.py

@@ -1,12 +1,12 @@
 import pytest
-from ..helpers.DocumentProcessorHelper import DocumentProcessor
-from ..helpers.ConfigHelper import ConfigHelper
+from code.utilities.helpers.DocumentProcessorHelper import DocumentProcessor
+from code.utilities.helpers.ConfigHelper import ConfigHelper
 
 document_url = "https://csciblob.blob.core.windows.net/rag-sol-acc/cognitive-services.pdf"
 url = "https://learn.microsoft.com/en-us/azure/search/search-what-is-azure-search"
 docx_url = "https://csciblob.blob.core.windows.net/rag-sol-acc/What is Azure OpenAI Service.docx"
 
-
+@pytest.mark.azure("This test requires Azure")
 def test_document_processor_layout():
     document_processor = DocumentProcessor()
     processors = list(filter(lambda x : x.document_type == 'pdf', ConfigHelper.get_active_config_or_default().document_processors))
@@ -15,6 +15,7 @@ def test_document_processor_layout():
     assert len(keys) > 0
 
 
+@pytest.mark.azure("This test requires Azure")
 def test_document_processor_read():
     document_processor = DocumentProcessor()
     processors = list(filter(lambda x : x.document_type == 'pdf', ConfigHelper.get_active_config_or_default().document_processors))
@@ -23,13 +24,15 @@ def test_document_processor_read():
     assert len(keys) > 0
 
 
+@pytest.mark.azure("This test requires Azure")
 def test_document_processor_web():
     document_processor = DocumentProcessor()
     processors = list(filter(lambda x : x.document_type == 'url', ConfigHelper.get_active_config_or_default().document_processors))
     keys = document_processor.process(source_url= url, processors=processors)    
     print(keys)
     assert len(keys) > 0
 
+@pytest.mark.azure("This test requires Azure")
 def test_document_processor_docx():
     document_processor = DocumentProcessor()
     processors = list(filter(lambda x : x.document_type == 'docx', ConfigHelper.get_active_config_or_default().document_processors))

code/utilities/tests/test_Orchestrator.py → tests/test_Orchestrator.py

@@ -1,13 +1,15 @@
 import pytest
-from ..helpers.OrchestratorHelper import Orchestrator, OrchestrationSettings
+from code.utilities.helpers.OrchestratorHelper import Orchestrator, OrchestrationSettings
 
+@pytest.mark.azure("This test requires Azure Open AI configured")
 def test_orchestrator_openai_function():
     message_orchestrator = Orchestrator()
     strategy = "openai_function"
     messages = message_orchestrator.handle_message(user_message="What's Azure AI Search?", chat_history=[], conversation_id="test_openai_function", orchestrator = OrchestrationSettings({"strategy" : strategy}))    
     assert messages[-1]['role'] == 'assistant'
     assert messages[-1]['content'] != ''
 
+@pytest.mark.azure("This test requires Azure Open AI configured")
 def test_orchestrator_langchain():
     message_orchestrator = Orchestrator()
     strategy = "langchain"

code/utilities/tests/test_OutputParserTool.py → tests/test_OutputParserTool.py

@@ -3,8 +3,8 @@
 
 from typing import List
 
-from utilities.parser.OutputParserTool import OutputParserTool
-from ..common.SourceDocument import SourceDocument
+from code.utilities.parser.OutputParserTool import OutputParserTool
+from code.utilities.common.SourceDocument import SourceDocument
 
 
 def test_returns_parsed_messages():