File size: 4,072 Bytes
c7aa14e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 |
# app/mobile_usability/service.py
import os
import logging
from typing import Dict, Any
# Optional project settings import (falls back to env var)
try:
from app.page_speed.config import settings # if present in your project
GEMINI_KEY = getattr(settings, "gemini_api_key", None)
except Exception:
GEMINI_KEY = None
from app.mobile_usability.models import Recommendation, PrioritySuggestions
from app.mobile_usability.prompts import MobilePrompts
# LangChain / Gemini wrapper imports (same style as your SEO module)
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import PydanticOutputParser
glogger = logging.getLogger(__name__)
class MobileUsabilityService:
"""
LLM-only service for generating mobile usability report and prioritized suggestions.
This class requires a Gemini API key to be available (env var or settings).
"""
def __init__(self):
# require Gemini key
key = GEMINI_KEY or os.getenv("GEMINI_API_KEY")
if not key:
msg = "Gemini API key not configured. Set settings.gemini_api_key or GEMINI_API_KEY env var."
glogger.error(msg)
raise RuntimeError(msg)
self.gemini_api_key = key
# initialize LLM wrapper (Gemini)
self.llm = ChatGoogleGenerativeAI(
model="gemini-2.5-flash",
temperature=0,
max_tokens=None,
timeout=None,
max_retries=3,
api_key=self.gemini_api_key
)
# Prompt template for generating the multi-line mobile usability report
self.report_prompt = ChatPromptTemplate.from_messages([
("system", MobilePrompts.Report_PROMPT),
("human", "Please generate a comprehensive mobile usability audit based on this data:\n\n{mobile_data}")
])
# Parser + priority prompt for structured priority suggestions
self.parser = PydanticOutputParser(pydantic_object=Recommendation)
self.priority_chain = (
ChatPromptTemplate.from_messages([
("system", MobilePrompts.SYSTEM_PROMPT),
("human", "{report}")
]).partial(format_instructions=self.parser.get_format_instructions())
| self.llm
| self.parser
)
def generate_mobile_report(self, mobile_data: Dict[str, Any]) -> str:
"""
Generate a mobile usability audit report using Gemini LLM (raw multi-line string).
Raises on failure.
"""
glogger.info("Invoking Gemini to generate mobile usability report.")
prompt_input = {"mobile_data": mobile_data}
try:
# Use the prompt template piped into the LLM (same pattern as your SEO module)
prompt_with_llm = self.report_prompt | self.llm
response = prompt_with_llm.invoke(prompt_input)
if not response:
raise RuntimeError("Empty response from Gemini for mobile usability report.")
# response is typically an object with .content in your setup; keep same handling
return getattr(response, "content", str(response)).strip()
except Exception as e:
glogger.error("Error generating mobile report via Gemini: %s", e, exc_info=True)
raise
def generate_mobile_priority(self, report: str) -> PrioritySuggestions:
"""
Generate prioritized mobile-usability suggestions using LLM + PydanticOutputParser.
Returns PrioritySuggestions Pydantic model instance.
"""
glogger.info("Invoking Gemini to generate priority suggestions.")
try:
rec: Recommendation = self.priority_chain.invoke({"report": report})
# `rec` is a Pydantic model (Recommendation); return the nested PrioritySuggestions
return rec.priority_suggestions
except Exception as e:
glogger.error("Error generating priorities via Gemini: %s", e, exc_info=True)
raise
|