tools package

Submodules

tools.build123d_retriever module

class tools.build123d_retriever.Build123dRetriever(db_file: str = 'build123d_vec.db', table: str = 'build123d_objects', embedding_model: Embeddings = None, reranking_model: Rerank = None, embedding_config: Dict = None, rerank_config: Dict = None, k: int = 10, k_rerank: int = 10)[source]

Bases: object

extract_all_objects(module_info_json: Dict) List[Dict][source]

Extract objects with type categorization from raw JSON.

generate_embedding_pairs(objects: List[Dict], only_names: bool = True)[source]
get_complete_info(query_text: str, k: int = None, k_rerank: int = None, with_docstring: bool = False, threshold: float = 0.8, distance_metrics: Literal['l2', 'cosine'] = 'cosine') List[Dict][source]

Get complete information about the queried objects in parallel.

Parameters:
  • query_text (str) – The query text.

  • k (int) – The number of results to return. Defaults to 5.

Returns:

A list of dictionaries containing complete information about the objects.

Return type:

List[Dict]

query(query_text: str, k: int = 10, k_rerank: int = 10, distance_metric: Literal['l2', 'cosine'] = 'cosine') List[Dict][source]

Query the database with a sentence or description in parallel.

Parameters:
  • query_text (str) – The query text.

  • k (int) – The number of results to return. Defaults to 5.

Returns:

A list of metadata dictionaries for the most relevant objects.

Return type:

List[Dict]

tools.code_dochelper module

class tools.code_dochelper.CodeDocHelper[source]

Bases: object

dict_to_markdown(data, show_docstring=True)[source]

Convert a dictionary containing module, class, function, or variable information into a Markdown-formatted string.

Parameters: data (dict): A dictionary containing module, class, function, or variable information. show_docstring (bool): If True, include docstrings in the output.

Returns: str: A Markdown-formatted string.

get_class_info(class_path, with_docstring=False)[source]

Retrieve the signature, methods, variables, and optionally the docstring of a class.

Parameters: class_path (str): The full import path to the class (e.g., “build123d.Box”). with_docstring (bool): If True, include the docstring in the output.

Returns: dict: A dictionary containing the class name, signature, flattened methods, variables, and docstring (optional).

get_function_info(function_path, with_docstring=False)[source]

Retrieve the signature and optionally the docstring of a function or class method.

Parameters: function_path (str): The full import path to the function (e.g., “build123d.Box.bounding_box”). with_docstring (bool): If True, include the docstring in the output.

Returns: dict: A dictionary containing the function name, signature, and docstring (optional).

get_info(path, with_docstring=True)[source]

Retrieve information about a module, class, function, or variable.

get_module_info(module_name, with_docstring=False)[source]

Retrieve the classes, functions, variables, and optionally the docstring of a module.

Parameters: module_name (str): The name of the module. with_docstring (bool): If True, include the docstring in the output.

Returns: dict: A dictionary containing the module name, classes, functions, variables, and docstring (optional).

tools.code_dochelper.doc_helper(import_path, with_docstring: bool = False)[source]

Query module, class, function, or method information and return it in Markdown format.

tools.web_reader module

tools.web_reader.fetch_content(url, return_format: Literal['text', 'markdown', 'html', 'screenshot'] = 'markdown', disable_gfm: Literal[False, True, 'table'] = False, bypass_cache=False, with_generated_alt=False, remove_images=False, timeout=None, json_response=True, with_links_summary=False, with_images_summary=False)[source]

Fetch content from a specified URL using the r.jina.ai API.

Parameters:
  • url – The URL to fetch content from.

  • return_format – The format of the returned content, options are ‘text’, ‘markdown’, ‘html’, ‘screenshot’.

  • json_response – Boolean flag to toggle JSON response format.

  • bypass_cache – Boolean flag to bypass cache.

  • disable_gfm – Boolean flag to disable Github Flavored Markdown or ‘table’ mode.

  • remove_images – Boolean flag to remove all images from the response.

  • timeout – Timeout in seconds for waiting the webpage to load.

  • with_generated_alt – Boolean flag to include generated alt text for images.

  • with_links_summary – Boolean flag to include a summary of all links at the end.

  • with_images_summary – Boolean flag to include a summary of all images at the end.

Returns:

The fetched content in the specified format.

Module contents