The type hinting for allowed_tools variable is unnecessary as it is already defined in the previous line. Removing the type hinting improves the readability of the code.
The display name for the SupabaseVectorStore is now set to "Supabase". This improves the user experience by providing a more descriptive name for the vector store.
The `sort_fields` method has been added to the `Template` class to sort fields based on the `DIRECT_TYPES` constant. Fields that have a `field_type` in `DIRECT_TYPES` are sorted first, followed by the remaining fields. This ensures that fields that have a direct type are processed first, which is important for the correct functioning of the template.
The DIRECT_TYPES constant was removed from the vertex and graph modules as it is now defined in the utils module. This change improves code organization and reduces duplication.
The format_dict function was updated to set a default value for the model_name key in the value dictionary for the OpenAI, ChatOpenAI, and Anthropic models. This ensures that the model_name key always has a value, even if the options list is empty.
✨ feat(chains.py): add TemplateField 'chain_type' to support different types of QA chains
The 'memory' field was previously set to required=False, but it is actually required for the chain to function properly. This fix sets required=True for the 'memory' field.
A new TemplateField 'chain_type' has been added to support different types of QA chains. This field is of type 'str', is required, and is a list of options. It allows the user to select the type of QA chain they want to use.
The advanced field was set to True for input_key and output_key fields, which made them appear in the advanced section of the UI. This was not intended, so the advanced field is now set to False for these fields.
🐛 fix(process.py): make inputs optional in process_graph_cached function
The inputs and tweaks parameters in the process_flow endpoint are now optional, which allows for more flexibility in the API. The inputs parameter in the process_graph_cached function is now optional, which prevents a ValueError from being raised when a Chain object is processed without inputs.
The memory_key field is now set to "chat_history" by default. This change ensures that the memory_key field is always initialized with a default value, which is useful for the application's functionality.
The import statements for the weaviate and pinecone libraries were causing import errors. Adding the `# type: ignore` comment suppresses these errors and allows the code to run without issues.
This commit adds type hinting to the function arguments of `get_result_and_thought` and `process_graph_cached` functions in `process.py` file. This improves code readability and maintainability.
The `initialize_chroma` function had redundant code that was removed. The `embedding_function` parameter was renamed to `embedding` to match the parameter name used in the `class_object` constructor. The `documents` and `texts` parameters were being used interchangeably, so the code was updated to use only `documents`.
The `name` parameter in the `format_field` method of the `DocumentLoaderFrontNode` class now uses the `Optional` type hint to indicate that it can be None. This improves the code's readability and makes it easier to understand the expected behavior of the method.
The function now checks if "texts" is in the params dictionary and renames it to "documents" for consistency. The "supabase_url" and "supabase_service_key" are now removed from the params dictionary and passed directly to the create_client function. The function also removes the "documents" and "texts" keys from the params dictionary if there are no documents in the params. This improves the readability of the code and removes redundant code.
The log level of the sorted vertices in the generator_build method of the Graph class has been changed from info to debug. This is because the sorted vertices are not critical information and are only useful for debugging purposes.
✨ feat(vector_store.py): add support for Supabase client object to be passed in as a parameter
The 'texts' parameter has been renamed to 'documents' to improve semantics. This change makes it clearer that the parameter is a list of documents. Additionally, support for a Supabase client object has been added as a parameter. This allows for more flexibility in the initialization of the SupabaseVectorStore class.
The SupabaseVectorStore is now supported in the vectorstores configuration in the config.yaml file. The SupabaseIcon has been added to the nodeIcons object in the utils.ts file to be able to display the icon for the SupabaseVectorStore.
The SupabaseVectorStore type requires additional fields such as table_name, query_name, supabase_url, and supabase_service_key. These fields are added to the extra_fields list and are then added to the list of fields that are displayed to the user. This change allows the user to create a SupabaseVectorStore with the required fields.
🚀 feat(vector_store.py): add support for initializing SupabaseVectorStore
This commit adds support for initializing new vector store types in the loading.py file. Specifically, the initialize_weaviate, initialize_faiss, and initialize_supabase functions were added to support the Weaviate, FAISS, and SupabaseVectorStore vector stores, respectively. The vector_store.py file was also updated to include the SupabaseVectorStore class and the initialize_supabase function. This allows for more flexibility in choosing vector stores for the application.
This commit adds support for Weaviate and FAISS vector stores to the existing Pinecone, Qdrant, and Chroma vector stores. The `initialize_weaviate` function initializes Weaviate and returns the class object. The `initialize_faiss` function initializes FAISS and returns the class object. These functions are used to initialize the respective vector stores. The `docs_in_params` function is used to check if there are documents in the parameters.
The FAISS vector store type is now supported and can be configured with the folder_path and index_name fields. The folder_path field specifies the local path to the FAISS index, while the index_name field specifies the name of the index. Additionally, a new client_kwargs field has been added to support additional client configuration options.
✨ feat(schemas.py): add StreamData class to handle SSE messages
✨ feat(frontend): add RadialProgress and ProgressBar components
🔧 chore(frontend): add daisyui to project dependencies
🎨 style(headerComponent): comment out unused code
✨ feat(ui/progress.tsx): add Progress component to reuse in other components
✨ feat(types/components): add ProgressBarType and RadialProgressType to use in components
🎨 style(tailwind.config.js): add daisyui plugin to tailwind configuration to use in the project
The return_direct field is a boolean field that allows the user to specify whether the output of the Python function should be returned directly or wrapped in a dictionary. This field is added to the TemplateField list of the PythonFunctionToolNode class.
✨ feat(schemas.py): add ProcessResponse schema to represent the response of the process endpoint
The PredictRequest schema has been renamed to InputRequest to better reflect its purpose. The message field has been removed as it is not needed. TweaksRequest has been added to represent the request body for the tweaks endpoint. UpdateTemplateRequest has been added to represent the request body for updating a template. ProcessResponse has been added to represent the response of the process endpoint, which returns a dictionary as the result.
The code has been refactored to improve readability by adding type hints and reformatting the code. The initialize_pinecone and initialize_chroma functions have been updated to support initializing objects from existing indexes. If there are no docs in the params, the functions will return an existing index. If there are docs in the params, the functions will create a new index. The initialize_qdrant function has not been changed.
🔨 refactor(vector_store.py): improve code readability by adding type hints and reformatting code
This commit removes unused imports and variables from the process.py file. The get_result_and_thought function has been simplified to take a dictionary of inputs instead of a single message string. The function now returns the output of the langchain_object instead of a tuple of result and thought.
The MemoryFrontendNode template now includes two new fields, input_key and output_key, which are string fields that are not required and are shown in the advanced section of the template. These fields can be used to specify the input and output keys for the memory node.
🔀 refactor(constants.tsx): update BASE_API_URL to reflect the changes in the predict_flow endpoint name
The predict_flow endpoint has been renamed to process_flow to better reflect its functionality. The inputs are now passed as a dictionary instead of a PredictRequest object. The tweaks parameter is now optional and is passed as a dictionary. The response model has been updated to reflect the changes in the response. The BASE_API_URL constant in constants.tsx has been updated to reflect the changes in the endpoint name.
The VectorStoreFrontendNode class now supports the Pinecone vector store type. Two new fields, pinecone_api_key and pinecone_env, have been added to the extra_fields list. The add_extra_fields method has been updated to add these fields to the template if the type_name is Pinecone. The add_field method has been updated to add all fields in the extra_fields list to the template. The add_field method has also been updated to remove the text_key field if the field name is "text_key". Finally, the fields list has been updated to include the new fields.
GitLoader is added to documentloaders to allow loading of documents from a Git repository. Pinecone is added to vectorstores to allow for vector similarity search. These additions improve the functionality of the application by allowing for more diverse sources of data and more efficient search capabilities.
🚀 feat(base.py): update import statement for loading module
🚀 feat(vector_store.py): add initialization functions for pinecone, chroma, and qdrant
The unused imports and endpoint for node validation were removed from validate.py. The import statement for the loading module was updated in base.py. Initialization functions for pinecone, chroma, and qdrant were added to vector_store.py to allow for the creation of objects from the respective services.
The `return_source_documents` field is added to the template as a boolean field. It is not a required field and is set to `True` by default. This field is used to determine whether the source documents should be returned or not. It is displayed in the frontend and can be toggled by the user.
The function build_template was renamed to build_file_field to improve naming consistency. A metadata field of type dict was added to the document loader template to allow for additional information to be passed along with the document. The format_field method was updated to show the metadata field when it is present.
🔨 refactor(documentloaders.py): rename build_template function to build_file_field for better semantics
📝 docs(initialize): add empty __init__.py file to initialize directory
The initialize directory was added to the project, but it was missing an __init__.py file. This file is necessary to make the directory a package and allow importing modules from it. An empty __init__.py file was added to the directory to fix this issue.
🐛 fix(process.py): correctly update template_data with tweak_value
The file extension is no longer included in the saved file name to avoid issues with file extensions that may not be supported. In process.py, the tweak_value is now correctly updated in the template_data dictionary. The key is now set to "value" for all tweaks except for "file_path" where the key is set to the name of the tweak.
✨ feat(loading.py): add persist parameter to Chroma vector store instantiation to enable persistence of vector store
The persist parameter is now extracted from the params dictionary before instantiating the Chroma vector store. This improves readability and reduces the complexity of the code. The persist parameter is now added to the Chroma vector store instantiation to enable persistence of the vector store. This allows the vector store to be reused across multiple sessions, improving performance and reducing the time required to load the vector store.
🐛 fix(vectorstores.py): add persist field to Weaviate vector store type
The `VectorStoreFrontendNode` class now supports the Chroma vector store type. A new boolean field `persist` has been added to the Chroma vector store type. The `add_extra_fields` method has been updated to add the `persist` field to the Chroma vector store type. The `format_field` method has been updated to include the `persist` field in the basic fields. Additionally, the `add_extra_fields` method has been updated to add the `weaviate_url` field to the Weaviate vector store type.
The changes in this commit are purely cosmetic. The indentation and formatting issues in the code have been fixed to improve readability and maintainability.
The QueryCheckerTool class was renamed to QuerySQLCheckerTool to match the correct class name. This change ensures that the correct class is being used and avoids any potential errors that may arise from using the wrong class name.
The save_uploaded_file function now uses the folder_name parameter instead of file_name to improve semantics. The appdirs library is now used to get the user cache directory and a folder for langflow cache is created. The sha256 hash of the file content is now used as the file name to avoid collisions and improve security. A folder is now created for each flow_id in the save_uploaded_file function. The code that deletes files after loading in the instantiate_documentloader function has been removed as it is unnecessary and can cause issues.
🐛 fix(endpoints.py): change file_name parameter to folder_name in save_uploaded_file function
🔒 chore(utils.py): use appdirs to get user cache directory and create a folder for langflow cache
🔒 chore(utils.py): use sha256 hash of file content as file name to avoid collisions and improve security
🔒 chore(utils.py): create folder for each flow_id in save_uploaded_file function
