Update app.py

app.py

@@ -1,8 +1,10 @@
 import gradio as gr
 import requests
+import socket
 import logging
 import time
 from bs4 import BeautifulSoup
+from urllib.parse import urlparse
 from google import genai
 
 # -------------------------------------------------
@@ -16,170 +18,174 @@ logger = logging.getLogger(__name__)
 
 logger.info("AI Website Review Tool starting up")
 
-# -------------------------------------------------
-# Website Fetching (Safe + Resilient)
-# -------------------------------------------------
+# -----------------------------
+# URL Normalization
+# -----------------------------
+def normalize_url(url: str) -> str:
+    parsed = urlparse(url)
+    if not parsed.scheme:
+        return "https://" + url
+    return url
+
+
+# -----------------------------
+# Fetch & Parse Website (Hardened)
+# -----------------------------
 def fetch_website_text(url: str) -> str:
+    socket.setdefaulttimeout(10)
+
     headers = {
-        "User-Agent": "Mozilla/5.0 (compatible; AIWebsiteAnalyzer/1.0)"
+        "User-Agent": (
+            "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
+            "AppleWebKit/537.36 (KHTML, like Gecko) "
+            "Chrome/121.0 Safari/537.36"
+        ),
+        "Accept": "text/html,application/xhtml+xml",
+        "Accept-Language": "en-US,en;q=0.9",
     }
 
-    response = requests.get(url, headers=headers, timeout=15)
+    response = requests.get(
+        url,
+        headers=headers,
+        timeout=10,
+        allow_redirects=True,
+    )
     response.raise_for_status()
 
     soup = BeautifulSoup(response.text, "html.parser")
 
-    # Remove scripts/styles
+    # Remove noisy tags
     for tag in soup(["script", "style", "noscript"]):
         tag.decompose()
 
-    text = soup.get_text(separator=" ")
-    return " ".join(text.split())[:8000]
+    title = soup.title.string.strip() if soup.title else ""
+    h1 = soup.find("h1").get_text(strip=True) if soup.find("h1") else ""
 
+    body_text = " ".join(soup.stripped_strings)
+    body_text = body_text[:8000]  # token safety
+
+    return f"""
+PAGE TITLE:
+{title}
+PRIMARY H1:
+{h1}
+VISIBLE CONTENT:
+{body_text}
+"""
 
+
+# -----------------------------
+# Safe Wrapper (Never Crash)
+# -----------------------------
 def fetch_website_text_safe(url: str) -> str:
-    logger.info(f"Attempting to fetch: {url}")
     try:
-        logger.info(f"Fetching website content | url={url}")
         return fetch_website_text(url)
     except Exception as e:
-        logger.warning(
-            f"Website fetch failed | url={url} | error={str(e)}"
-        )
-        return """
+        return f"""
 ⚠️ Unable to fully fetch website content.
-
-Please analyze this website based on:
-- URL structure
-- Homepage intent
-- General UX, SEO, and conversion best practices
+Error:
+{str(e)}
+Fallback:
+Analyze based on URL structure, homepage intent, and general best practices.
 """
 
 
-# -------------------------------------------------
+# -----------------------------
 # Gemini Analysis
-# -------------------------------------------------
+# -----------------------------
 def analyze_website(api_key, url, industry, goal):
-    client = genai.Client(api_key=api_key)
-    logging.info("Initialized AI client")
-
-    website_text = fetch_website_text_safe(url)
-
-    prompt = f"""
-You are an expert AI consultant helping small businesses improve their websites.
+    if not api_key:
+        return "❌ Please enter your Gemini API key."
 
-Website URL:
-{url}
+    if not url:
+        return "❌ Please enter a website URL."
 
-Industry:
-{industry}
-
-Primary Business Goal:
-{goal}
-
-Website Content (may be partial):
+    try:
+        url = normalize_url(url)
+
+        client = genai.Client(api_key=api_key)
+
+        website_text = fetch_website_text_safe(url)
+
+        prompt = f"""
+You are an AI consultant helping small businesses improve their websites.
+Business context:
+- Industry: {industry}
+- Primary goal: {goal}
+Analyze the website content below and provide recommendations in this structure:
+1. Messaging Clarity (score 1–10)
+   - Main issue
+   - 2–3 actionable recommendations
+2. Conversion Effectiveness (score 1–10)
+   - Main issue
+   - 2–3 actionable recommendations
+3. Trust & Credibility (score 1–10)
+   - Main issue
+   - 2–3 actionable recommendations
+4. User Experience Issues
+   - Bullet list of issues
+5. AI & Automation Opportunities
+   - 3 concrete ideas a small business could implement
+End with:
+- Overall score out of 100
+- Top 3 fixes to prioritize this week
+Use clear, non-technical business language.
+Website content:
 {website_text}
-
-Please provide:
-1. Overall assessment
-2. UX & design improvements
-3. SEO recommendations
-4. Conversion & messaging improvements
-5. AI & automation opportunities
-6. Top 3 prioritized action items
-
-Be concise, practical, and actionable.
 """
 
-    try:
         response = client.models.generate_content(
             model="gemini-2.5-flash",
             contents=prompt,
         )
-        return response.text
-
-    except Exception as e:
-        logger.exception("Gemini API error")
-        raise
-
-
-# -------------------------------------------------
-# Analysis Wrapper (UX + Logging)
-# -------------------------------------------------
-def analyze_with_status(api_key, url, industry, goal):
-    start_time = time.time()
-    safe_url = url[:120] if url else "EMPTY"
-
-    logger.info(
-        f"Analysis started | url={safe_url} | industry={industry} | goal={goal}"
-    )
-
-    # Disable button + show status
-    yield (
-        "🔍 **Analyzing website… please wait**",
-        "",
-        gr.update(interactive=False),
-    )
-
-    try:
-        logging.info("Starting analysis")
-        result = analyze_website(api_key, url, industry, goal)
-
-        duration = round(time.time() - start_time, 2)
-        logger.info(
-            f"Analysis completed | url={safe_url} | duration={duration}s"
-        )
 
-        yield (
-            "✅ **Analysis complete**",
-            "## ✅ Website Analysis Complete\n\n" + result,
-            gr.update(interactive=True),
-        )
+        return response.text
 
     except Exception as e:
-        duration = round(time.time() - start_time, 2)
-        logger.exception(
-            f"Analysis failed | url={safe_url} | duration={duration}s"
-        )
-
-        yield (
-            "❌ **Analysis failed**",
-            "Something went wrong while analyzing this website. Please try again.",
-            gr.update(interactive=True),
-        )
+        return f"❌ Error during analysis: {str(e)}"
 
 
-# -------------------------------------------------
+# -----------------------------
 # Gradio UI
-# -------------------------------------------------
+# -----------------------------
 with gr.Blocks(title="AI Website Review Tool") as demo:
+    gr.Markdown("## 🔍 AI Website Review Tool")
     gr.Markdown(
-        """
-# 🔍 AI Website Review Tool
-AI-powered website audits for small businesses  
-*Built by Esquire IT*
-"""
+        "Analyze any website and receive practical, business-focused recommendations."
     )
 
     api_key = gr.Textbox(
         label="Gemini API Key",
-        placeholder="Enter your Gemini API key",
+        placeholder="Paste your Gemini API key here",
         type="password",
     )
 
     url = gr.Textbox(
         label="Website URL",
-        placeholder="https://www.example.com",
+        placeholder="https://example.com",
     )
 
-    industry = gr.Textbox(
+    industry = gr.Dropdown(
         label="Industry",
-        placeholder="e.g. Law Firm, Restaurant, SaaS, Consultant",
+        choices=[
+            "General SMB",
+            "Law Firm",
+            "Hospitality",
+            "Healthcare",
+            "Real Estate",
+        ],
+        value="General SMB",
     )
 
-    goal = gr.Textbox(
-        label="Primary Goal",
-        placeholder="e.g. Get more leads, increase bookings, improve credibility",
+    goal = gr.Dropdown(
+        label="Primary Website Goal",
+        choices=[
+            "Generate leads",
+            "Sell services",
+            "Build credibility",
+            "Educate visitors",
+        ],
+        value="Generate leads",
     )
 
     analyze_btn = gr.Button("Analyze Website")
@@ -187,13 +193,28 @@ AI-powered website audits for small businesses
     status = gr.Markdown("")
     output = gr.Markdown()
 
+    def analyze_with_status(api_key, url, industry, goal):
+        # Disable button + show analyzing status
+        yield (
+            "🔍 **Analyzing website… please wait**",
+            "",
+            gr.update(interactive=False)
+        )
+
+        result = analyze_website(api_key, url, industry, goal)
+
+        # Re-enable button + show results
+        yield (
+            "✅ **Analysis complete**",
+            "## ✅ Website Analysis Complete\n\n" + result,
+            gr.update(interactive=True)
+        )
+
     analyze_btn.click(
         fn=analyze_with_status,
         inputs=[api_key, url, industry, goal],
-        outputs=[status, output, analyze_btn],
+        outputs=[status, output],
         queue=False,
     )
 
-logger.info("AI Website Review Tool ready")
-
-demo.launch()
+demo.launch()