openperplex 0.1.0

A Python client for the Openperplex API

Openperplex

Openperplex is a Python library that provides a simple interface to interact with the Openperplex API. It allows users to perform searches and retrieve information using various endpoints, with support for both streaming and non-streaming responses.

Installation

You can install Openperplex using pip:

pip install openperplex

Requirements

• Python 3.6+
• httpx
• typing

Usage

Initialization

To use the Openperplex library, you need to initialize the Openperplex class with your API key, provider, and provider key:

from openperplex import Openperplex

client = Openperplex(
    api_key="your_api_key",
    provider="openai",  # or "groq"
    provider_key="your_provider_key"
)

Methods

1. Search Stream

The search_stream method allows you to perform a search and receive results as a stream of events.

for event in client.search_stream(
    query="Your search query",
    date_context="optional_date_context",
    location="us",
    pro_mode=False
):
    print(event)

2. Search Simple Stream

The search_simple_stream method provides a simplified streaming search.

for event in client.search_simple_stream("Your search query"):
    print(event)

3. Search

The search method performs a search and returns the complete results as a list.

results = client.search(
    query="Your search query",
    date_context="optional_date_context",
    location="us", // can be any country code
    pro_mode=False
)
print(results)

4. Search Simple

The search_simple method provides a simplified search that returns results as a string.

result = client.search_simple("Your search query")
print(result)

Error Handling

The library uses custom OpenperplexError exceptions to handle errors. You can catch these exceptions to handle errors in your code:

from openperplex import Openperplex, OpenperplexError

client = Openperplex(api_key="your_api_key", provider="openai", provider_key="your_provider_key")

try:
    results = client.search("Your query")
except OpenperplexError as e:
    print(f"An error occurred: {e.status_code} - {e.detail}")

Streaming Responses

For methods that return a stream (search_stream and search_simple_stream), the library yields dictionaries with different types of events:

• {"type": "llm", "text": "..."}: LLM-generated text
• {"type": "sources", ...}: Source information
• {"type": "relevant", ...}: Relevant questions
• {"type": "finished", ...}: Indicates the end of the stream

You can process these events as needed in your application.

Configuration Options

• api_key (required): Your Openperplex API key
• provider (required): The provider to use ("openai" or "groq")
• provider_key (required): Your provider API key
• base_url (optional): The base URL for the Openperplex API (defaults to the provided URL)

Notes

• The library automatically closes the HTTP client when the Openperplex instance is deleted.
• Make sure to handle rate limiting and other API-specific requirements as per the Openperplex API documentation.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License