from transformers import pipeline
import streamlit as st

@st.cache_resource
def get_llm_pipelines():
    classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli", truncation=True)
    sentiment_pipe = pipeline("sentiment-analysis", model="distilbert-base-uncased-finetuned-sst-2-english", truncation=True)
    summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-6-6", truncation=True)
    return classifier, sentiment_pipe, summarizer

def analyze_post(text, classifier, sentiment_pipe, summarizer):
    if not text or not text.strip():
        return {"category": "N/A", "sentiment": "N/A", "summary": "Empty"}

    result = {}
    labels = ["Bug Report", "Feature Request", "Competitor Mention", "Positive Feedback"]
    hypothesis = "This text is about a {}."
    result["category"] = classifier(text, labels, hypothesis_template=hypothesis)["labels"][0]
    result["sentiment"] = sentiment_pipe(text)[0]['label']

    if len(text.split()) > 40:
        result["summary"] = summarizer(text, max_length=60, min_length=20, do_sample=False)[0]['summary_text']
    else:
        result["summary"] = text
    return result