364 lines
16 KiB
Python
364 lines
16 KiB
Python
import logging
|
|
import httpx
|
|
import asyncio
|
|
from typing import Dict, Any, List, Union, Optional
|
|
from fastapi import HTTPException
|
|
from modules.shared.configuration import APP_CONFIG
|
|
from modules.aicore.aicoreBase import BaseConnectorAi
|
|
from modules.datamodels.datamodelAi import AiModel, ModelCapabilitiesEnum, PriorityEnum, ProcessingModeEnum, OperationTypeEnum
|
|
|
|
# Configure logger
|
|
logger = logging.getLogger(__name__)
|
|
|
|
def loadConfigData():
|
|
"""Load configuration data for Perplexity connector"""
|
|
return {
|
|
"apiKey": APP_CONFIG.get('Connector_AiPerplexity_API_SECRET'),
|
|
"apiUrl": APP_CONFIG.get('Connector_AiPerplexity_API_URL'),
|
|
"modelName": APP_CONFIG.get('Connector_AiPerplexity_MODEL_NAME'),
|
|
"temperature": float(APP_CONFIG.get('Connector_AiPerplexity_TEMPERATURE')),
|
|
}
|
|
|
|
class AiPerplexity(BaseConnectorAi):
|
|
"""Connector for communication with the Perplexity API."""
|
|
|
|
def __init__(self):
|
|
super().__init__()
|
|
# Load configuration
|
|
self.config = loadConfigData()
|
|
self.apiKey = self.config["apiKey"]
|
|
self.apiUrl = self.config["apiUrl"]
|
|
self.modelName = self.config["modelName"]
|
|
|
|
# HttpClient for API calls
|
|
self.httpClient = httpx.AsyncClient(
|
|
timeout=120.0, # Longer timeout for complex requests
|
|
headers={
|
|
"Authorization": f"Bearer {self.apiKey}",
|
|
"Content-Type": "application/json",
|
|
"Accept": "application/json"
|
|
}
|
|
)
|
|
|
|
logger.info(f"Perplexity Connector initialized with model: {self.modelName}")
|
|
|
|
def getConnectorType(self) -> str:
|
|
"""Get the connector type identifier."""
|
|
return "perplexity"
|
|
|
|
def getModels(self) -> List[AiModel]:
|
|
"""Get all available Perplexity models."""
|
|
return [
|
|
AiModel(
|
|
name="perplexity_callAiBasic",
|
|
displayName="Llama 3.1 Sonar Large 128k",
|
|
connectorType="perplexity",
|
|
maxTokens=128000,
|
|
contextLength=128000,
|
|
costPer1kTokensInput=0.005,
|
|
costPer1kTokensOutput=0.005,
|
|
speedRating=8,
|
|
qualityRating=8,
|
|
capabilities=[ModelCapabilitiesEnum.TEXT_GENERATION, ModelCapabilitiesEnum.CHAT, ModelCapabilitiesEnum.REASONING, ModelCapabilitiesEnum.WEB_SEARCH],
|
|
functionCall=self.callAiBasic,
|
|
priority=PriorityEnum.BALANCED,
|
|
processingMode=ProcessingModeEnum.ADVANCED,
|
|
operationTypes=[OperationTypeEnum.GENERAL, OperationTypeEnum.PLAN, OperationTypeEnum.ANALYSE, OperationTypeEnum.GENERATE, OperationTypeEnum.WEB_RESEARCH],
|
|
version="llama-3.1-sonar-large-128k-online",
|
|
calculatePriceUsd=lambda processingTime, bytesSent, bytesReceived: (bytesSent / 4 / 1000) * 0.005 + (bytesReceived / 4 / 1000) * 0.005
|
|
),
|
|
AiModel(
|
|
name="perplexity_callAiWithWebSearch",
|
|
displayName="Sonar Pro",
|
|
connectorType="perplexity",
|
|
maxTokens=128000,
|
|
contextLength=128000,
|
|
costPer1kTokensInput=0.01,
|
|
costPer1kTokensOutput=0.01,
|
|
speedRating=7,
|
|
qualityRating=9,
|
|
capabilities=[ModelCapabilitiesEnum.TEXT_GENERATION, ModelCapabilitiesEnum.WEB_SEARCH, ModelCapabilitiesEnum.RESEARCH],
|
|
functionCall=self.callAiWithWebSearch,
|
|
priority=PriorityEnum.QUALITY,
|
|
processingMode=ProcessingModeEnum.DETAILED,
|
|
operationTypes=[OperationTypeEnum.WEB_RESEARCH],
|
|
version="sonar-pro",
|
|
calculatePriceUsd=lambda processingTime, bytesSent, bytesReceived: (bytesSent / 4 / 1000) * 0.01 + (bytesReceived / 4 / 1000) * 0.01
|
|
),
|
|
AiModel(
|
|
name="perplexity_researchTopic",
|
|
displayName="Mistral 7B Instruct",
|
|
connectorType="perplexity",
|
|
maxTokens=32000,
|
|
contextLength=32000,
|
|
costPer1kTokensInput=0.002,
|
|
costPer1kTokensOutput=0.002,
|
|
speedRating=8,
|
|
qualityRating=8,
|
|
capabilities=[ModelCapabilitiesEnum.WEB_SEARCH, ModelCapabilitiesEnum.RESEARCH, ModelCapabilitiesEnum.INFORMATION_GATHERING],
|
|
functionCall=self.researchTopic,
|
|
priority=PriorityEnum.COST,
|
|
processingMode=ProcessingModeEnum.BASIC,
|
|
operationTypes=[OperationTypeEnum.WEB_RESEARCH],
|
|
version="mistral-7b-instruct",
|
|
calculatePriceUsd=lambda processingTime, bytesSent, bytesReceived: (bytesSent / 4 / 1000) * 0.002 + (bytesReceived / 4 / 1000) * 0.002
|
|
),
|
|
AiModel(
|
|
name="perplexity_answerQuestion",
|
|
displayName="Mistral 7B Instruct QA",
|
|
connectorType="perplexity",
|
|
maxTokens=32000,
|
|
contextLength=32000,
|
|
costPer1kTokensInput=0.002,
|
|
costPer1kTokensOutput=0.002,
|
|
speedRating=8,
|
|
qualityRating=8,
|
|
capabilities=[ModelCapabilitiesEnum.WEB_SEARCH, ModelCapabilitiesEnum.QUESTION_ANSWERING, ModelCapabilitiesEnum.RESEARCH],
|
|
functionCall=self.answerQuestion,
|
|
priority=PriorityEnum.COST,
|
|
processingMode=ProcessingModeEnum.BASIC,
|
|
operationTypes=[OperationTypeEnum.WEB_RESEARCH],
|
|
version="mistral-7b-instruct",
|
|
calculatePriceUsd=lambda processingTime, bytesSent, bytesReceived: (bytesSent / 4 / 1000) * 0.002 + (bytesReceived / 4 / 1000) * 0.002
|
|
),
|
|
AiModel(
|
|
name="perplexity_getCurrentNews",
|
|
displayName="Mistral 7B Instruct News",
|
|
connectorType="perplexity",
|
|
maxTokens=32000,
|
|
contextLength=32000,
|
|
costPer1kTokensInput=0.002,
|
|
costPer1kTokensOutput=0.002,
|
|
speedRating=8,
|
|
qualityRating=8,
|
|
capabilities=[ModelCapabilitiesEnum.WEB_SEARCH, ModelCapabilitiesEnum.NEWS, ModelCapabilitiesEnum.CURRENT_EVENTS],
|
|
functionCall=self.getCurrentNews,
|
|
priority=PriorityEnum.COST,
|
|
processingMode=ProcessingModeEnum.BASIC,
|
|
operationTypes=[OperationTypeEnum.WEB_RESEARCH],
|
|
version="mistral-7b-instruct",
|
|
calculatePriceUsd=lambda processingTime, bytesSent, bytesReceived: (bytesSent / 4 / 1000) * 0.002 + (bytesReceived / 4 / 1000) * 0.002
|
|
)
|
|
]
|
|
|
|
async def callAiBasic(self, messages: List[Dict[str, Any]], temperature: float = None, maxTokens: int = None) -> str:
|
|
"""
|
|
Calls the Perplexity API with the given messages.
|
|
|
|
Args:
|
|
messages: List of messages in OpenAI format (role, content)
|
|
temperature: Temperature for response generation (0.0-1.0)
|
|
maxTokens: Maximum number of tokens in the response
|
|
|
|
Returns:
|
|
The response from the Perplexity API
|
|
|
|
Raises:
|
|
HTTPException: For errors in API communication
|
|
"""
|
|
try:
|
|
# Use parameters from configuration if none were overridden
|
|
if temperature is None:
|
|
temperature = self.config.get("temperature", 0.2)
|
|
|
|
# Don't set maxTokens from config - let the model use its full context length
|
|
# Our continuation system handles stopping early via prompt engineering
|
|
|
|
payload = {
|
|
"model": self.modelName,
|
|
"messages": messages,
|
|
"temperature": temperature
|
|
}
|
|
|
|
# Add max_tokens - use provided value or throw error
|
|
if maxTokens is None:
|
|
raise ValueError("maxTokens must be provided for Perplexity API calls")
|
|
payload["max_tokens"] = maxTokens
|
|
|
|
response = await self.httpClient.post(
|
|
self.apiUrl,
|
|
json=payload
|
|
)
|
|
|
|
if response.status_code != 200:
|
|
error_detail = f"Perplexity API error: {response.status_code} - {response.text}"
|
|
logger.error(error_detail)
|
|
|
|
# Provide more specific error messages based on status code
|
|
if response.status_code == 429:
|
|
error_message = "Rate limit exceeded. Please wait before making another request."
|
|
elif response.status_code == 401:
|
|
error_message = "Invalid API key. Please check your Perplexity API configuration."
|
|
elif response.status_code == 400:
|
|
error_message = f"Invalid request to Perplexity API: {response.text}"
|
|
else:
|
|
error_message = f"Perplexity API error ({response.status_code}): {response.text}"
|
|
|
|
raise HTTPException(status_code=500, detail=error_message)
|
|
|
|
responseJson = response.json()
|
|
content = responseJson["choices"][0]["message"]["content"]
|
|
return content
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error calling Perplexity API: {str(e)}")
|
|
raise HTTPException(status_code=500, detail=f"Error calling Perplexity API: {str(e)}")
|
|
|
|
async def callAiWithWebSearch(self, query: str, temperature: float = None, maxTokens: int = None) -> str:
|
|
"""
|
|
Calls Perplexity API with web search capabilities for research.
|
|
|
|
Args:
|
|
query: The research query or question
|
|
temperature: Temperature for response generation (0.0-1.0)
|
|
maxTokens: Maximum number of tokens in the response
|
|
|
|
Returns:
|
|
The response from Perplexity with web search context
|
|
"""
|
|
try:
|
|
# Use parameters from configuration if none were overridden
|
|
if temperature is None:
|
|
temperature = self.config.get("temperature", 0.2)
|
|
|
|
# Don't set maxTokens from config - let the model use its full context length
|
|
# Our continuation system handles stopping early via prompt engineering
|
|
|
|
# For web search, we use the configured model name
|
|
webSearchModel = self.modelName
|
|
|
|
payload = {
|
|
"model": webSearchModel,
|
|
"messages": [
|
|
{
|
|
"role": "user",
|
|
"content": query
|
|
}
|
|
],
|
|
"temperature": temperature
|
|
}
|
|
|
|
# Add max_tokens - use provided value or throw error
|
|
if maxTokens is None:
|
|
raise ValueError("maxTokens must be provided for Perplexity API calls")
|
|
payload["max_tokens"] = maxTokens
|
|
|
|
response = await self.httpClient.post(
|
|
self.apiUrl,
|
|
json=payload
|
|
)
|
|
|
|
if response.status_code != 200:
|
|
error_detail = f"Perplexity Web Search API error: {response.status_code} - {response.text}"
|
|
logger.error(error_detail)
|
|
|
|
if response.status_code == 429:
|
|
error_message = "Rate limit exceeded for web search. Please wait before making another request."
|
|
elif response.status_code == 401:
|
|
error_message = "Invalid API key for web search. Please check your Perplexity API configuration."
|
|
elif response.status_code == 400:
|
|
error_message = f"Invalid request to Perplexity Web Search API: {response.text}"
|
|
else:
|
|
error_message = f"Perplexity Web Search API error ({response.status_code}): {response.text}"
|
|
|
|
raise HTTPException(status_code=500, detail=error_message)
|
|
|
|
responseJson = response.json()
|
|
content = responseJson["choices"][0]["message"]["content"]
|
|
return content
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error calling Perplexity Web Search API: {str(e)}")
|
|
raise HTTPException(status_code=500, detail=f"Error calling Perplexity Web Search API: {str(e)}")
|
|
|
|
async def researchTopic(self, topic: str, depth: str = "basic") -> str:
|
|
"""
|
|
Research a topic using Perplexity's web search capabilities.
|
|
|
|
Args:
|
|
topic: The topic to research
|
|
depth: Research depth - "basic", "detailed", or "comprehensive"
|
|
|
|
Returns:
|
|
Comprehensive research results on the topic
|
|
"""
|
|
try:
|
|
# Create research prompts based on depth
|
|
if depth == "basic":
|
|
prompt = f"Provide a basic overview of: {topic}"
|
|
elif depth == "detailed":
|
|
prompt = f"Provide a detailed analysis of: {topic}. Include recent developments, key facts, and important information."
|
|
else: # comprehensive
|
|
prompt = f"Provide a comprehensive research report on: {topic}. Include recent developments, key facts, statistics, expert opinions, and current trends."
|
|
|
|
return await self.callAiWithWebSearch(prompt)
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error researching topic: {str(e)}")
|
|
raise HTTPException(status_code=500, detail=f"Error researching topic: {str(e)}")
|
|
|
|
async def answerQuestion(self, question: str, context: str = None) -> str:
|
|
"""
|
|
Answer a question using web search for current information.
|
|
|
|
Args:
|
|
question: The question to answer
|
|
context: Optional context to provide
|
|
|
|
Returns:
|
|
Answer with web search context
|
|
"""
|
|
try:
|
|
if context:
|
|
prompt = f"Context: {context}\n\nQuestion: {question}\n\nPlease provide a comprehensive answer using current information from the web."
|
|
else:
|
|
prompt = f"Question: {question}\n\nPlease provide a comprehensive answer using current information from the web."
|
|
|
|
return await self.callAiWithWebSearch(prompt)
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error answering question: {str(e)}")
|
|
raise HTTPException(status_code=500, detail=f"Error answering question: {str(e)}")
|
|
|
|
async def getCurrentNews(self, topic: str = None, limit: int = 5) -> str:
|
|
"""
|
|
Get current news on a specific topic.
|
|
|
|
Args:
|
|
topic: The topic to get news about (optional)
|
|
limit: Number of news items to retrieve
|
|
|
|
Returns:
|
|
Current news information
|
|
"""
|
|
try:
|
|
if topic:
|
|
prompt = f"Get the latest news about {topic}. Provide {limit} recent news items with sources and dates."
|
|
else:
|
|
prompt = f"Get the latest news. Provide {limit} recent news items with sources and dates."
|
|
|
|
return await self.callAiWithWebSearch(prompt)
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error getting current news: {str(e)}")
|
|
raise HTTPException(status_code=500, detail=f"Error getting current news: {str(e)}")
|
|
|
|
async def _testConnection(self) -> bool:
|
|
"""
|
|
Tests the connection to the Perplexity API.
|
|
|
|
Returns:
|
|
True if connection is successful, False otherwise
|
|
"""
|
|
try:
|
|
# Try a simple test message
|
|
testMessages = [
|
|
{"role": "user", "content": "Hello, please respond with just 'OK' to confirm the connection works."}
|
|
]
|
|
|
|
response = await self.callAiBasic(testMessages)
|
|
return response and len(response.strip()) > 0
|
|
|
|
except Exception as e:
|
|
logger.error(f"Perplexity connection test failed: {str(e)}")
|
|
return False
|