app.py

import streamlit as st
import requests
import json
import plotly.graph_objects as go
import plotly.express as px
from typing import Dict, List, Any
import time
import pandas as pd

# Page configuration
st.set_page_config(
    page_title="Market Research & Analysis Platform",
    layout="wide",
    initial_sidebar_state="expanded"
)

# Custom styling (updated for dark mode)
st.markdown("""
    <style>
    .main {
        background-color: #121212;
        color: #ffffff;
    }
    .insight-card {
        background-color: #1e1e1e;
        padding: 1.5rem;
        border-radius: 8px;
        box-shadow: 0 2px 4px rgba(255,255,255,0.1);
        margin: 1rem 0;
    }
    .metric-card {
        background-color: #1e1e1e;
        padding: 1rem;
        border-radius: 4px;
        margin: 0.5rem 0;
    }
    .source-card {
        background-color: #2e2e2e;
        padding: 0.5rem;
        border-radius: 4px;
        font-size: 0.9rem;
        margin-top: 0.5rem;
        color: #ffffff;
    }
    .highlight-text {
        color: #4c6ef5;
        font-weight: bold;
    }
    </style>
""", unsafe_allow_html=True)

def query_perplexity(query: str, context: Dict) -> Dict:
    url = "https://api.perplexity.ai/chat/completions"
    headers = {
        "Authorization": f"Bearer {st.secrets['PERPLEXITY_API_KEY']}",
        "Content-Type": "application/json"
    }
    payload = {
        "model": "llama-3.1-sonar-small-128k-online",
        "messages": [
            {"role": "system", "content": get_system_prompt(context)},
            {"role": "user", "content": query}
        ],
        "temperature": 0.2,
        "max_tokens": 4096,
        "top_p": 0.9,
        "search_domain_filter": ["perplexity.ai"],
        "return_images": False,
        "return_related_questions": False,
        "search_recency_filter": context.get('timeframe', 'month')
    }
    try:
        response = requests.post(url, headers=headers, json=payload)
        if response.status_code == 200:
            return {
                "status": "success",
                "data": response.json(),
                "citations": response.json().get("citations", [])
            }
        else:
            return {"status": "error", "message": f"API Error: {response.status_code}"}
    except Exception as e:
        return {"status": "error", "message": str(e)}

def get_system_prompt(context: Dict) -> str:
    # Adjust the prompt if Executive style and Comprehensive depth are selected.
    if context.get("style", "").lower() == "executive" and context.get("depth", "").lower() == "comprehensive":
        return f"""You are an expert market research analyst focusing on {context['focus_area']}.
Provide a high-level executive summary with key insights and concrete metrics.
Analysis style: Executive
Depth: Comprehensive
Timeline: Past {context['timeframe']}

Format your response with clear sections:
1. Executive Summary
2. Key Metrics
3. Market Position
4. Growth Analysis
5. Strategic Recommendations

Include specific numbers, percentages, and actionable insights."""
    else:
        return f"""You are an expert market research analyst focusing on {context['focus_area']}.
Provide detailed analysis with concrete metrics and specific insights.
Analysis style: {context['style']}
Depth: {context['depth']}
Timeline: Past {context['timeframe']}

Format your response with clear sections:
1. Key Metrics
2. Market Position
3. Growth Analysis
4. Competitive Insights
5. Strategic Recommendations

Include specific numbers, percentages, and actionable insights."""

def parse_perplexity_response(response: Dict) -> Dict:
    try:
        content = response['data']['choices'][0]['message']['content']
        citations = response.get('citations', [])
        sections = {
            'key_metrics': [],
            'market_position': [],
            'growth_analysis': [],
            'competitive_insights': [],
            'recommendations': []
        }
        current_section = None
        for line in content.split('\n'):
            line = line.strip()
            if not line:
                continue
            lower_line = line.lower()
            if 'key metric' in lower_line:
                current_section = 'key_metrics'
            elif 'market position' in lower_line:
                current_section = 'market_position'
            elif 'growth' in lower_line:
                current_section = 'growth_analysis'
            elif 'competiti' in lower_line:
                current_section = 'competitive_insights'
            elif 'recommend' in lower_line:
                current_section = 'recommendations'
            elif current_section and line.startswith(('-', '•', '*')):
                sections[current_section].append(line.lstrip('-•* '))
        return {"status": "success", "sections": sections, "citations": citations}
    except Exception as e:
        return {"status": "error", "message": str(e)}

def extract_metrics(content: Dict) -> Dict:
    metrics = {
        'market_share': [],
        'growth_rate': [],
        'competitive_position': [],
        'innovation_score': []
    }
    try:
        for section in content['sections'].values():
            for line in section:
                if '%' in line or any(char.isdigit() for char in line):
                    if 'market share' in line.lower():
                        metrics['market_share'].append(extract_number(line))
                    elif 'growth' in line.lower():
                        metrics['growth_rate'].append(extract_number(line))
                    elif 'position' in line.lower():
                        metrics['competitive_position'].append(extract_number(line))
                    elif 'innovation' in line.lower():
                        metrics['innovation_score'].append(extract_number(line))
        return metrics
    except Exception as e:
        st.error(f"Error extracting metrics: {str(e)}")
        return metrics

def extract_number(text: str) -> float:
    import re
    numbers = re.findall(r'[-+]?\d*\.?\d+%?', text)
    if numbers:
        number = numbers[0]
        return float(number.replace('%', '')) if '%' in number else float(number)
    return 0.0

def create_visualizations(metrics: Dict, context: Dict) -> Dict:
    charts = {}
    if metrics['market_share'] and metrics['competitive_position']:
        fig = go.Figure()
        categories = ['Market Share', 'Growth Rate', 'Competitive Position', 'Innovation Score']
        values = [
            metrics['market_share'][0] if metrics['market_share'] else 0,
            metrics['growth_rate'][0] if metrics['growth_rate'] else 0,
            metrics['competitive_position'][0] if metrics['competitive_position'] else 0,
            metrics['innovation_score'][0] if metrics['innovation_score'] else 0
        ]
        fig.add_trace(go.Scatterpolar(
            r=values,
            theta=categories,
            fill='toself',
            name=context['company_name']
        ))
        fig.update_layout(
            polar=dict(radialaxis=dict(visible=True, range=[0, 100])),
            showlegend=False,
            title=f"Market Position Analysis - {context['company_name']}"
        )
        charts['market_position'] = fig
    if metrics['growth_rate']:
        fig = go.Figure()
        fig.add_trace(go.Scatter(
            y=metrics['growth_rate'],
            mode='lines+markers',
            name='Growth Rate'
        ))
        fig.update_layout(
            title=f"Growth Trend Analysis - {context['company_name']}",
            yaxis_title='Growth Rate (%)',
            showlegend=True
        )
        charts['growth_trend'] = fig
    return charts

def format_insights(content: Dict) -> str:
    """Format detailed analysis with sub-titles for each section"""
    formatted = ""
    section_titles = {
        'key_metrics': '📊 Key Metrics',
        'market_position': '🎯 Market Position',
        'growth_analysis': '📈 Growth Analysis',
        'competitive_insights': '🔍 Competitive Insights',
        'recommendations': '💡 Strategic Recommendations'
    }
    for section, title in section_titles.items():
        if content['sections'].get(section):
            formatted += f"\n### {title}\n\n"
            for idx, point in enumerate(content['sections'][section], start=1):
                formatted += f"- **{idx}.** {point}\n"
    return formatted

def main():
    st.title("Market Research & Analysis Platform")
    st.markdown("Real-time market insights with data-driven analysis")

    with st.sidebar:
        st.header("Analysis Parameters")
        company_name = st.text_input("Company/Product Name", placeholder="e.g., Tesla, OpenAI, Snowflake")
        industry = st.selectbox("Industry", ["Technology", "AI/ML", "SaaS", "Fintech", "E-commerce", "Healthcare", "Energy", "Other"])
        st.markdown("### Analysis Configuration")
        focus_area = st.multiselect("Focus Areas", ["Market Position", "Growth Trajectory", "Technology Stack", "Competitive Analysis", "Innovation Trends", "Investment Outlook"], default=["Market Position", "Growth Trajectory"])
        timeframe = st.select_slider("Analysis Timeframe", options=["week", "month", "quarter", "year"], value="month")
        depth = st.select_slider("Analysis Depth", options=["Brief", "Detailed", "Comprehensive"], value="Detailed")
        style = st.selectbox("Report Style", ["Technical", "Business", "Executive"], index=1)
        competitors = st.text_input("Key Competitors (optional)", placeholder="Comma-separated names")

    if st.button("Generate Analysis", type="primary"):
        if not company_name:
            st.warning("Please enter a company name.")
            return

        analysis_context = {
            "company_name": company_name,
            "industry": industry,
            "focus_area": ", ".join(focus_area),
            "timeframe": timeframe,
            "depth": depth,
            "style": style,
            "competitors": competitors
        }
        
        with st.spinner("Generating market analysis..."):
            progress_bar = st.progress(0)
            status_text = st.empty()
            try:
                status_text.text("Gathering market intelligence...")
                progress_bar.progress(20)
                research_response = query_perplexity(
                    f"Provide a detailed market analysis for {company_name} in the {industry} industry, focusing on {', '.join(focus_area)}",
                    analysis_context
                )
                if research_response["status"] != "success":
                    st.error("Failed to gather market intelligence.")
                    return

                status_text.text("Processing insights...")
                progress_bar.progress(40)
                parsed_content = parse_perplexity_response(research_response)
                if parsed_content["status"] != "success":
                    st.error("Failed to process insights.")
                    return

                status_text.text("Generating visualizations...")
                progress_bar.progress(60)
                metrics = extract_metrics(parsed_content)
                charts = create_visualizations(metrics, analysis_context)

                status_text.text("Preparing analysis report...")
                progress_bar.progress(80)
                tabs = st.tabs(["Overview", "Detailed Analysis", "Visualizations"])
                
                with tabs[0]:
                    st.markdown("## Executive Summary", unsafe_allow_html=True)
                    if metrics:
                        cols = st.columns(len(metrics))
                        for col, (metric, values) in zip(cols, metrics.items()):
                            if values:
                                col.metric(metric.replace('_', ' ').title(), f"{values[0]:.1f}%")
                    if parsed_content["citations"]:
                        st.markdown("### Sources")
                        for citation in parsed_content["citations"]:
                            st.markdown(f'<div class="source-card">{citation}</div>', unsafe_allow_html=True)
                
                with tabs[1]:
                    st.markdown(format_insights(parsed_content), unsafe_allow_html=True)
                
                with tabs[2]:
                    for chart_name, fig in charts.items():
                        st.plotly_chart(fig, use_container_width=True)
                
                progress_bar.progress(100)
                status_text.text("Analysis complete!")
                
                st.download_button(
                    "Download Analysis",
                    data=json.dumps({
                        "context": analysis_context,
                        "insights": parsed_content["sections"],
                        "metrics": metrics,
                        "citations": parsed_content["citations"]
                    }, indent=2),
                    file_name=f"market_analysis_{company_name}.json",
                    mime="application/json"
                )
            except Exception as e:
                st.error(f"An error occurred: {str(e)}")
                return

if __name__ == "__main__":
    main()