prompt_template.py

Overview

The prompt_template.py file provides a simple utility for loading text prompt templates stored as Markdown (.md) files from the same directory as the script. It manages prompt caching to avoid redundant file reads, improving efficiency when prompts are reused multiple times within an application. This functionality is commonly used in systems that dynamically retrieve and use textual templates, such as AI prompt engineering, chatbot responses, or configuration-driven text generation.

Detailed Explanation

Constants

• PROMPT_DIR (str):
  This constant holds the absolute directory path where the prompt_template.py file is located. It is used as the base directory to locate prompt Markdown files.

Variables

• _loaded_prompts (dict[str, str]):
  A private dictionary used as an in-memory cache to store the contents of loaded prompts. The keys are prompt names (without extension), and the values are the string contents of the corresponding prompt files.

Function: load_prompt

def load_prompt(name: str) -> str:

Description

Loads the content of a prompt template from a Markdown file named <name>.md located in the same directory as prompt_template.py. If the prompt content was previously loaded, it returns the cached version instead of reading from disk again.

Parameters

• name (str):
  The base name of the prompt file (without the .md extension) to be loaded. For example, if name is "welcome", the function attempts to load welcome.md.

Returns

• str:
  The content of the prompt file as a stripped Unicode string.

Raises

• FileNotFoundError:
  Raised if the file <name>.md does not exist in the prompt directory.

Usage Example

try:
    prompt_text = load_prompt("welcome_message")
    print(prompt_text)
except FileNotFoundError as e:
    print(e)

Implementation Details and Algorithms

• Caching Mechanism:
  The function implements a simple caching mechanism using the _loaded_prompts dictionary. When a prompt is loaded for the first time, its content is read from disk and stored in _loaded_prompts. Subsequent calls to load the same prompt return the cached content, reducing file I/O overhead.

• File Path Resolution:
  The prompt files are assumed to be stored in the same directory as the script (prompt_template.py). The path to the prompt file is constructed by joining PROMPT_DIR with the prompt's filename ({name}.md). This approach makes the system's prompt storage location predictable and consistent.

• File Content Handling:
  The function reads the entire content of the Markdown file at once using UTF-8 encoding and strips whitespace from the start and end of the content before caching and returning it.

Interaction with Other System Components

• Prompt Management:
  This file functions as a utility module, likely imported by other components that require prompt templates for processing user input, generating responses, or configuring AI behavior.

• File Organization:
  The file expects prompt templates to be maintained as Markdown files alongside this script or in the same directory, encouraging modular prompt management.

• Error Propagation:
  By raising FileNotFoundError when a prompt file is missing, it allows calling code to handle missing prompts explicitly, enhancing robustness.

Visual Diagram

flowchart TD
    A[load_prompt(name: str)] --> B{Is prompt in _loaded_prompts?}
    B -- Yes --> C[Return cached prompt content]
    B -- No --> D[Construct path: PROMPT_DIR + name.md]
    D --> E{Does file exist?}
    E -- No --> F[Raise FileNotFoundError]
    E -- Yes --> G[Open and read file content]
    G --> H[Strip content whitespace]
    H --> I[Cache content in _loaded_prompts]
    I --> C

Summary

The prompt_template.py file is a focused utility for loading and caching Markdown-based prompt templates from the local directory. It provides a simple, efficient way to manage prompt content for applications that rely on text templates, facilitating reusable, modular prompt management with minimal overhead.