Create app.py

app.py

@@ -0,0 +1,164 @@
+import gradio as gr
+import requests
+from bs4 import BeautifulSoup
+import pandas as pd
+import plotly.express as px
+from langchain_google_genai import ChatGoogleGenerativeAI
+from langchain_core.messages import HumanMessage
+import os
+import re
+
+# --- Configuration & Initialization ---
+# For deployment on Hugging Face, set GEMINI_API_KEY in the Space's secrets
+llm = ChatGoogleGenerativeAI(model="gemini-1.5-flash", google_api_key=os.environ.get("GEMINI_API_KEY"))
+
+# --- Core Scraping & Analysis Functions ---
+
+def fetch_html(url):
+    """Fetches HTML content from a URL with a browser-like user-agent."""
+    try:
+        headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'}
+        response = requests.get(url, headers=headers, timeout=10)
+        response.raise_for_status()
+        return response.text
+    except requests.RequestException as e:
+        print(f"Error fetching {url}: {e}")
+        return None
+
+def analyze_onpage_seo(soup):
+    """Analyzes the on-page SEO elements of a webpage."""
+    title = soup.find('title').get_text(strip=True) if soup.find('title') else "N/A"
+    description = soup.find('meta', attrs={'name': 'description'})['content'] if soup.find('meta', attrs={'name': 'description'}) else "N/A"
+    
+    headings = {'h1': [], 'h2': [], 'h3': []}
+    for h_tag in ['h1', 'h2', 'h3']:
+        for tag in soup.find_all(h_tag):
+            headings[h_tag].append(tag.get_text(strip=True))
+            
+    # Simple word count
+    word_count = len(soup.get_text().split())
+    
+    return {
+        "title": title,
+        "description": description,
+        "headings": headings,
+        "word_count": word_count
+    }
+
+def analyze_tech_stack(soup, html):
+    """Performs a very basic tech stack analysis."""
+    tech = set()
+    if "react.js" in html or 'data-reactroot' in html: tech.add("React")
+    if "vue.js" in html: tech.add("Vue.js")
+    if "angular.js" in html: tech.add("Angular")
+    if "wp-content" in html: tech.add("WordPress")
+    if "gtag('config'" in html: tech.add("Google Analytics (GA4)")
+    if "GTM-" in html: tech.add("Google Tag Manager")
+    if soup.find('link', href=lambda x: x and 'tailwind' in x): tech.add("Tailwind CSS")
+        
+    return list(tech) if tech else ["Basic HTML/CSS"]
+
+def generate_ai_summary(url, seo_data, tech_data):
+    """Generates a high-level strategic summary using an LLM."""
+    prompt = f"""
+    You are a world-class marketing strategist. Analyze the following data for the website {url} and provide a concise, actionable summary.
+
+    ## On-Page SEO Analysis:
+    - **Title:** {seo_data['title']}
+    - **Meta Description:** {seo_data['description']}
+    - **H1 Headings:** {', '.join(seo_data['headings']['h1'])}
+    - **Page Word Count:** {seo_data['word_count']}
+
+    ## Technology Stack:
+    - {', '.join(tech_data)}
+
+    ## Your Analysis:
+    Provide a 3-bullet point summary covering:
+    1.  **Their Primary Goal:** What is this page trying to achieve based on its language and structure?
+    2.  **Their Target Audience:** Who are they talking to?
+    3.  **A Key Strategic Insight:** What is one clever thing they are doing, or one major missed opportunity?
+    """
+    try:
+        response = llm.invoke([HumanMessage(content=prompt)])
+        return response.content
+    except Exception as e:
+        return f"Could not generate AI summary: {e}"
+
+# --- The Main Orchestrator ---
+
+def competitor_teardown(url):
+    """The main function that runs the entire analysis pipeline."""
+    if not url.startswith(('http://', 'https://')):
+        url = 'https://' + url
+    
+    yield "Fetching website...", None, None, None, None, gr.Button("Analyzing...", interactive=False)
+    
+    html = fetch_html(url)
+    if not html:
+        yield "Failed to fetch URL. Please check the address and try again.", None, None, None, None, gr.Button("Analyze", interactive=True)
+        return
+
+    soup = BeautifulSoup(html, 'html.parser')
+    
+    yield "Analyzing SEO & Tech...", None, None, None, None, gr.Button("Analyzing...", interactive=False)
+    seo_data = analyze_onpage_seo(soup)
+    tech_data = analyze_tech_stack(soup, html)
+    
+    yield "Generating AI Summary...", None, None, None, None, gr.Button("Analyzing...", interactive=False)
+    ai_summary = generate_ai_summary(url, seo_data, tech_data)
+
+    # --- Prepare Outputs ---
+    
+    # SEO Output
+    seo_md = f"""
+    ### 👑 SEO & Content
+    | Metric | Value |
+    | :--- | :--- |
+    | **Page Title** | `{seo_data['title']}` |
+    | **Meta Description** | `{seo_data['description']}` |
+    | **Word Count** | `{seo_data['word_count']}` |
+    
+    #### Heading Structure:
+    - **H1 Tags ({len(seo_data['headings']['h1'])}):** {', '.join(f'`{h}`' for h in seo_data['headings']['h1'])}
+    - **H2 Tags ({len(seo_data['headings']['h2'])}):** {len(seo_data['headings']['h2'])} found
+    """
+    
+    # Tech Stack Output
+    tech_md = "### stack Tech Stack\n\n" + "\n".join([f"- `{t}`" for t in tech_data])
+    
+    yield ai_summary, seo_md, tech_md, "Analysis Complete", "More features coming soon!", gr.Button("Analyze", interactive=True)
+
+# --- Gradio UI ---
+with gr.Blocks(theme=gr.themes.Soft(), css="footer {display: none !important;}") as demo:
+    gr.Markdown("# 🕵️ Gumbo Board: The Instant Competitor Teardown")
+    gr.Markdown("Enter a competitor's website to get an instant analysis of their online strategy. *Powered by Gumbo (BeautifulSoup) & AI.*")
+
+    with gr.Row():
+        url_input = gr.Textbox(label="Enter Competitor URL", placeholder="e.g., notion.so or mailchimp.com", scale=4)
+        submit_btn = gr.Button("Analyze", variant="primary", scale=1)
+
+    with gr.Tabs():
+        with gr.TabItem("🧠 AI Summary"):
+            summary_output = gr.Markdown("Your AI-powered strategic summary will appear here.")
+        with gr.TabItem("🔍 On-Page SEO"):
+            seo_output = gr.Markdown()
+        with gr.TabItem("⚙️ Tech Stack"):
+            tech_output = gr.Markdown()
+        with gr.TabItem("📢 Ads & Keywords (Coming Soon)"):
+            ads_output = gr.Markdown()
+        with gr.TabItem("Social Presence (Coming Soon)"):
+            social_output = gr.Markdown()
+
+    # Link the button click to the main function
+    submit_btn.click(
+        fn=competitor_teardown,
+        inputs=[url_input],
+        outputs=[summary_output, seo_output, tech_output, ads_output, social_output, submit_btn]
+    )
+    
+    gr.Markdown("---")
+    gr.Markdown("### Ready for More? \n Get unlimited reports, save projects, and export to PDF with our Pro plan. \n **[🚀 Launching Soon - Sign up on Gumroad!](https://gumroad.com/)**")
+
+
+# Launch the app
+demo.launch()