Feature/add prompt tests and cleanup (#1523)

* add extra parsers

* add prompt tests, cleanup

* add prompt tests, cleanup

* merge

* set mock console as default

* set mock console as default

.github/actions/run-sdk-prompt-management-tests/action.yml

@@ -0,0 +1,42 @@
+name: 'Run SDK Prompt Management Tests'
+description: 'Runs SDK prompt management tests for R2R'
+runs:
+  using: "composite"
+  steps:
+    # First run basic prompt operations
+    - name: Add prompt test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_add_prompt
+
+    - name: Get prompt test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_get_prompt
+
+    - name: Get all prompts test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_get_all_prompts
+
+    - name: Update prompt test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_update_prompt
+
+    # Then run error handling and access control tests
+    - name: Prompt error handling test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_prompt_error_handling
+
+    - name: Prompt access control test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_prompt_access_control
+
+    # Finally run deletion test
+    - name: Delete prompt test (SDK)
+      working-directory: ./py
+      shell: bash
+      run: poetry run python tests/integration/runner_sdk.py test_delete_prompt

.github/workflows/r2r-full-py-integration-tests.yml

@@ -27,6 +27,7 @@ jobs:
           - sdk-retrieval
           - sdk-auth
           - sdk-collections
+          - sdk-prompts
     env:
       OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       TELEMETRY_ENABLED: 'false'
@@ -82,3 +83,8 @@ jobs:
         if: matrix.test_category == 'sdk-collections'
         uses: ./.github/actions/run-sdk-collections-tests
         continue-on-error: true
+
+      - name: Run SDK Prompt Tests
+        if: matrix.test_category == 'sdk-prompts'
+        uses: ./.github/actions/run-sdk-prompt-management-tests
+        continue-on-error: true

.github/workflows/r2r-light-py-integration-tests.yml

@@ -29,6 +29,7 @@ jobs:
           - sdk-retrieval
           - sdk-auth
           - sdk-collections
+          - sdk-prompts
     env:
       OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       TELEMETRY_ENABLED: 'false'
@@ -85,3 +86,8 @@ jobs:
         if: matrix.test_category == 'sdk-collections'
         uses: ./.github/actions/run-sdk-collections-tests
         continue-on-error: true
+
+      - name: Run SDK Prompt Tests
+        if: matrix.test_category == 'sdk-prompts'
+        uses: ./.github/actions/run-sdk-prompt-management-tests
+        continue-on-error: true

py/core/base/providers/email.py

@@ -12,7 +12,7 @@ class EmailConfig(ProviderConfig):
     smtp_username: Optional[str] = None
     smtp_password: Optional[str] = None
     from_email: Optional[str] = None
-    use_tls: bool = True
+    use_tls: Optional[bool] = True
 
     @property
     def supported_providers(self) -> list[str]:

py/core/base/providers/ingestion.py

@@ -20,13 +20,13 @@ class IngestionConfig(ProviderConfig):
     )
     extra_parsers: dict[str, str] = {}
 
-    audio_transcription_model: str
+    audio_transcription_model: str = "openai/whisper-1"
 
-    vision_img_prompt_name: Optional[str] = None
-    vision_img_model: str
+    vision_img_prompt_name: str = "vision_img"
+    vision_img_model: str = "openai/gpt-4-mini"
 
-    vision_pdf_prompt_name: Optional[str] = None
-    vision_pdf_model: str
+    vision_pdf_prompt_name: str = "vision_pdf"
+    vision_pdf_model: str = "openai/gpt-4-mini"
 
     @property
     def supported_providers(self) -> list[str]:

py/core/parsers/media/img_parser.py

@@ -16,8 +16,6 @@
 class ImageParser(AsyncParser[DataType]):
     """A parser for image data using vision models."""
 
-    DEFAULT_IMG_VISION_PROMPT_NAME = "vision_img"
-
     def __init__(
         self,
         config: IngestionConfig,
@@ -56,7 +54,6 @@ async def ingest(  # type: ignore
         if not self.vision_prompt_text:
             self.vision_prompt_text = await self.database_provider.get_prompt(  # type: ignore
                 prompt_name=self.config.vision_img_prompt_name
-                or self.DEFAULT_IMG_VISION_PROMPT_NAME
             )
         try:
             # Verify model supports vision

py/core/parsers/media/pdf_parser.py

@@ -25,8 +25,6 @@
 class VLMPDFParser(AsyncParser[DataType]):
     """A parser for PDF documents using vision models for page processing."""
 
-    DEFAULT_PDF_VISION_PROMPT_NAME = "vision_pdf"
-
     def __init__(
         self,
         config: IngestionConfig,
@@ -142,7 +140,6 @@ async def ingest(
         if not self.vision_prompt_text:
             self.vision_prompt_text = await self.database_provider.get_prompt(  # type: ignore
                 prompt_name=self.config.vision_pdf_prompt_name
-                or self.DEFAULT_PDF_VISION_PROMPT_NAME
             )
 
         temp_dir = None

py/r2r.toml

@@ -70,9 +70,13 @@ batch_size = 256
 
 [embedding]
 provider = "litellm"
-base_model = "openai/text-embedding-3-large"
-base_dimension = 3072
-quantization_settings = { quantization_type = "INT1" }
+# RECOMMENDED - For advanced applications,
+# use `openai/text-embedding-3-large` with `base_dimension = 3072` and binary quantization
+# base_model = "openai/text-embedding-3-large"
+# base_dimension = 3072
+# quantization_settings = { quantization_type = "INT1" }
+base_model = "openai/text-embedding-3-small"
+base_dimension = 512
 batch_size = 128
 add_title_as_prefix = false
 rerank_model = "None"
@@ -116,4 +120,4 @@ provider = "simple"
 provider = "r2r"
 
 [email]
-provider = "smtp"
+provider = "mock_console"

py/tests/conftest.py

@@ -12,6 +12,7 @@
     BCryptConfig,
     CompletionConfig,
     DatabaseConfig,
+    EmailConfig,
     EmbeddingConfig,
     PersistentLoggingConfig,
     SqlitePersistentLoggingProvider,
@@ -29,6 +30,7 @@
 )
 from core.providers import (
     BCryptProvider,
+    ConsoleMockEmailProvider,
     LiteLLMCompletionProvider,
     LiteLLMEmbeddingProvider,
     PostgresDBProvider,
@@ -169,12 +171,25 @@ def auth_config(app_config):
     )
 
 
+@pytest.fixture(scope="function")
+def email_provider(app_config):
+    return ConsoleMockEmailProvider(
+        EmailConfig(provider="console_mock", app=app_config)
+    )
+
+
 @pytest.fixture(scope="function")
 async def r2r_auth_provider(
-    auth_config, crypto_provider, temporary_postgres_db_provider
+    auth_config,
+    crypto_provider,
+    temporary_postgres_db_provider,
+    email_provider,
 ):
     auth_provider = R2RAuthProvider(
-        auth_config, crypto_provider, temporary_postgres_db_provider
+        auth_config,
+        crypto_provider,
+        temporary_postgres_db_provider,
+        email_provider,
     )
     await auth_provider.initialize()
     yield auth_provider

py/tests/core/providers/database/relational/test_document_db.py

@@ -45,7 +45,7 @@ async def test_upsert_documents_overview(temporary_postgres_db_provider):
     assert inserted_document.id == document_info.id
     assert inserted_document.collection_ids == document_info.collection_ids
     assert inserted_document.user_id == document_info.user_id
-    assert inserted_document.type == document_info.document_type
+    assert inserted_document.document_type == document_info.document_type
     assert inserted_document.metadata == document_info.metadata
     assert inserted_document.title == document_info.title
     assert inserted_document.version == document_info.version