ok

Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.9
+
+WORKDIR /code
+
+COPY ./requirements.txt /code/requirements.txt
+
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+COPY . /code/
+
+CMD ["python", "app.py"]

README.md

@@ -3,9 +3,8 @@ title: SEO Report Generator
 emoji: 🔍
 colorFrom: blue
 colorTo: green
-sdk: streamlit
-sdk_version: 1.28.1
-app_file: app.py
+sdk: docker
+app_port: 7860
 pinned: false
 license: mit
 ---

requirements.txt

@@ -7,4 +7,5 @@ jinja2
 validators
 urllib3
 lxml
-uuid
+uuid
+reportlab

simple_pdf_generator.py

@@ -17,7 +17,7 @@ class SimplePDFGenerator:
     
     def generate_pdf(self, html_content: str) -> bytes:
         """
-        Generate PDF from HTML content using simple text-based approach
+        Generate PDF from HTML content with better formatting
         """
         if not self.available:
             raise ImportError("PDF generation requires reportlab: pip install reportlab")
@@ -25,59 +25,78 @@ class SimplePDFGenerator:
         # Import reportlab components
         from reportlab.pdfgen import canvas
         from reportlab.lib.pagesizes import letter, A4
-        from reportlab.lib.styles import getSampleStyleSheet
-        from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer
+        from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
+        from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle
         from reportlab.lib.units import inch
+        from reportlab.lib.colors import Color, black, blue, green, red
         from bs4 import BeautifulSoup
+        import re
         
-        # Parse HTML and extract text content
+        # Parse HTML and extract content
         soup = BeautifulSoup(html_content, 'html.parser')
         
-        # Remove style and script tags
-        for tag in soup(["style", "script"]):
-            tag.decompose()
-        
         # Create PDF buffer
         buffer = io.BytesIO()
         
-        # Create PDF document
-        doc = SimpleDocTemplate(buffer, pagesize=A4)
+        # Create PDF document with margins
+        doc = SimpleDocTemplate(
+            buffer, 
+            pagesize=A4,
+            topMargin=0.75*inch,
+            bottomMargin=0.75*inch,
+            leftMargin=0.75*inch,
+            rightMargin=0.75*inch
+        )
+        
+        # Define custom styles
         styles = getSampleStyleSheet()
-        story = []
         
-        # Extract title
-        title_tag = soup.find('title')
-        title = title_tag.text if title_tag else "SEO Report"
+        # Custom styles
+        title_style = ParagraphStyle(
+            'CustomTitle',
+            parent=styles['Heading1'],
+            fontSize=24,
+            textColor=black,
+            spaceAfter=20,
+            alignment=1  # Center
+        )
         
-        # Add title
-        story.append(Paragraph(title, styles['Title']))
-        story.append(Spacer(1, 12))
+        header_style = ParagraphStyle(
+            'CustomHeader',
+            parent=styles['Heading2'],
+            fontSize=16,
+            textColor=blue,
+            spaceBefore=15,
+            spaceAfter=10
+        )
         
-        # Extract main content sections
-        sections = soup.find_all(['h1', 'h2', 'h3', 'p', 'div'])
+        subheader_style = ParagraphStyle(
+            'CustomSubHeader',
+            parent=styles['Heading3'],
+            fontSize=14,
+            textColor=black,
+            spaceBefore=10,
+            spaceAfter=8
+        )
         
-        for section in sections:
-            if section.name in ['h1', 'h2', 'h3']:
-                # Headers
-                text = section.get_text().strip()
-                if text:
-                    if section.name == 'h1':
-                        story.append(Paragraph(text, styles['Heading1']))
-                    elif section.name == 'h2':
-                        story.append(Paragraph(text, styles['Heading2']))
-                    else:
-                        story.append(Paragraph(text, styles['Heading3']))
-                    story.append(Spacer(1, 6))
-            
-            elif section.name in ['p', 'div']:
-                # Paragraphs
-                text = section.get_text().strip()
-                if text and len(text) > 20:  # Skip very short text
-                    try:
-                        story.append(Paragraph(text[:500], styles['Normal']))  # Limit length
-                        story.append(Spacer(1, 6))
-                    except:
-                        pass  # Skip problematic content
+        story = []
+        
+        # Add report title
+        title = "SEO Analysis Report"
+        url_elem = soup.find(string=re.compile(r'https?://'))
+        if url_elem:
+            url = re.search(r'https?://[^\s]+', str(url_elem))
+            if url:
+                title = f"SEO Analysis Report - {url.group()}"
+        
+        story.append(Paragraph(title, title_style))
+        story.append(Spacer(1, 20))
+        
+        # Extract and format content systematically
+        self._extract_executive_summary(soup, story, header_style, styles['Normal'])
+        self._extract_technical_seo(soup, story, header_style, subheader_style, styles['Normal'])
+        self._extract_content_audit(soup, story, header_style, subheader_style, styles['Normal'])
+        self._extract_recommendations(soup, story, header_style, styles['Normal'])
         
         # Build PDF
         doc.build(story)
@@ -85,6 +104,67 @@ class SimplePDFGenerator:
         # Get PDF data
         buffer.seek(0)
         return buffer.getvalue()
+    
+    def _extract_executive_summary(self, soup, story, header_style, normal_style):
+        """Extract executive summary section"""
+        exec_section = soup.find(string=re.compile(r'Executive Summary', re.I))
+        if exec_section:
+            story.append(Paragraph("Executive Summary", header_style))
+            
+            # Look for health score
+            health_text = soup.find(string=re.compile(r'Overall SEO Health', re.I))
+            if health_text:
+                parent = health_text.find_parent()
+                if parent:
+                    text = parent.get_text().strip()
+                    story.append(Paragraph(text, normal_style))
+                    story.append(Spacer(1, 10))
+    
+    def _extract_technical_seo(self, soup, story, header_style, subheader_style, normal_style):
+        """Extract technical SEO section"""
+        tech_section = soup.find(string=re.compile(r'Technical SEO', re.I))
+        if tech_section:
+            story.append(Paragraph("Technical SEO Analysis", header_style))
+            
+            # Look for performance scores
+            perf_elements = soup.find_all(string=re.compile(r'Performance Score|Mobile|Desktop', re.I))
+            for elem in perf_elements[:3]:  # Limit results
+                parent = elem.find_parent()
+                if parent:
+                    text = parent.get_text().strip()
+                    if len(text) > 10 and len(text) < 200:
+                        story.append(Paragraph(text, normal_style))
+            story.append(Spacer(1, 10))
+    
+    def _extract_content_audit(self, soup, story, header_style, subheader_style, normal_style):
+        """Extract content audit section"""
+        content_section = soup.find(string=re.compile(r'Content Audit', re.I))
+        if content_section:
+            story.append(Paragraph("Content Audit", header_style))
+            
+            # Look for content metrics
+            content_elements = soup.find_all(string=re.compile(r'Pages Analyzed|Metadata|Word Count', re.I))
+            for elem in content_elements[:3]:  # Limit results
+                parent = elem.find_parent()
+                if parent:
+                    text = parent.get_text().strip()
+                    if len(text) > 10 and len(text) < 200:
+                        story.append(Paragraph(text, normal_style))
+            story.append(Spacer(1, 10))
+    
+    def _extract_recommendations(self, soup, story, header_style, normal_style):
+        """Extract recommendations section"""
+        rec_section = soup.find(string=re.compile(r'Recommendation', re.I))
+        if rec_section:
+            story.append(Paragraph("Recommendations", header_style))
+            
+            # Look for recommendation items
+            rec_elements = soup.find_all('li')
+            for elem in rec_elements[:5]:  # Top 5 recommendations
+                text = elem.get_text().strip()
+                if len(text) > 15:
+                    story.append(Paragraph(f"• {text}", normal_style))
+            story.append(Spacer(1, 10))
 
 def create_browser_pdf_instructions() -> str:
     """

templates/index.html

@@ -1,5 +1,6 @@
 <!DOCTYPE html>
 <html lang="en">
+
 <head>
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
@@ -17,7 +18,7 @@
             font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
             min-height: 100vh;
             background-image: radial-gradient(circle at 20% 20%, rgba(120, 119, 198, 0.3) 0%, transparent 50%),
-                              radial-gradient(circle at 80% 80%, rgba(255, 119, 198, 0.3) 0%, transparent 50%);
+                radial-gradient(circle at 80% 80%, rgba(255, 119, 198, 0.3) 0%, transparent 50%);
         }
 
         .container {
@@ -42,7 +43,7 @@
             justify-content: center;
             gap: 12px;
         }
-        
+
         .logo-icon {
             background: rgba(255, 255, 255, 0.1);
             border-radius: 12px;
@@ -53,13 +54,13 @@
             backdrop-filter: blur(5px);
             border: 1px solid rgba(255, 255, 255, 0.2);
         }
-        
+
         .logo-svg {
             width: 32px;
             height: 32px;
             filter: brightness(0) invert(1);
         }
-        
+
         .logo-text {
             background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
             -webkit-background-clip: text;
@@ -107,7 +108,11 @@
         .form-input:focus {
             outline: none;
             border-color: #374151;
-            background: rgba(255, 255, 255, 0.08);
+            background: rgba(255, 255, 255, 0.15);
+        }
+
+        .form-input:not(:placeholder-shown) {
+            background: rgba(255, 255, 255, 0.15);
         }
 
         .form-input::placeholder {
@@ -220,8 +225,13 @@
         }
 
         @keyframes spin {
-            0% { transform: rotate(0deg); }
-            100% { transform: rotate(360deg); }
+            0% {
+                transform: rotate(0deg);
+            }
+
+            100% {
+                transform: rotate(360deg);
+            }
         }
 
         .error-message {
@@ -238,17 +248,18 @@
             .container {
                 padding: 20px 16px;
             }
-            
+
             .main-card {
                 padding: 24px;
             }
-            
+
             .logo {
                 font-size: 36px;
             }
         }
     </style>
 </head>
+
 <body>
     <div class="container">
         <div class="header">
@@ -256,7 +267,7 @@
                 <span class="logo-icon">
                     <img src="/static/logo.svg" alt="Thinkly Logo" class="logo-svg">
                 </span>
-                <span class="logo-text">Thinkly SEO</span>
+                <span class="logo-text">Thinkly Labs SEO</span>
             </h1>
             <p class="subtitle">Professional SEO Analysis & Reporting</p>
         </div>
@@ -288,14 +299,13 @@
             <form id="seoForm">
                 <div class="form-group">
                     <label class="form-label" for="url">Website URL</label>
-                    <input type="url" id="url" name="url" class="form-input" 
-                           placeholder="https://example.com" required>
+                    <input type="url" id="url" name="url" class="form-input" placeholder="https://example.com" required>
                 </div>
 
                 <div class="form-group">
                     <label class="form-label" for="competitors">Competitor URLs (Optional)</label>
-                    <textarea id="competitors" name="competitors" class="form-input textarea" 
-                              placeholder="https://competitor1.com&#10;https://competitor2.com&#10;One URL per line"></textarea>
+                    <textarea id="competitors" name="competitors" class="form-input textarea"
+                        placeholder="https://competitor1.com&#10;https://competitor2.com&#10;One URL per line"></textarea>
                 </div>
 
                 <button type="submit" class="generate-btn" id="generateBtn">
@@ -314,30 +324,30 @@
     </div>
 
     <script>
-        document.getElementById('seoForm').addEventListener('submit', async function(e) {
+        document.getElementById('seoForm').addEventListener('submit', async function (e) {
             e.preventDefault();
-            
+
             const url = document.getElementById('url').value.trim();
             const competitors = document.getElementById('competitors').value
                 .split('\n')
                 .map(c => c.trim())
                 .filter(c => c);
-            
+
             const loadingOverlay = document.getElementById('loadingOverlay');
             const errorMessage = document.getElementById('errorMessage');
             const loadingText = document.getElementById('loadingText');
             const loadingSubtext = document.getElementById('loadingSubtext');
-            
+
             errorMessage.style.display = 'none';
             loadingOverlay.style.display = 'flex';
-            
+
             const loadingMessages = [
                 { text: 'Analyzing technical SEO...', subtext: 'Checking PageSpeed insights' },
                 { text: 'Performing content audit...', subtext: 'Crawling website content' },
                 { text: 'Analyzing competitors...', subtext: 'Comparing performance metrics' },
                 { text: 'Generating professional report...', subtext: 'Creating charts and visualizations' }
             ];
-            
+
             let messageIndex = 0;
             const messageInterval = setInterval(() => {
                 if (messageIndex < loadingMessages.length) {
@@ -346,7 +356,7 @@
                     messageIndex++;
                 }
             }, 3000);
-            
+
             try {
                 const response = await fetch('/generate', {
                     method: 'POST',
@@ -358,12 +368,12 @@
                         competitors: competitors
                     })
                 });
-                
+
                 const data = await response.json();
-                
+
                 clearInterval(messageInterval);
                 loadingOverlay.style.display = 'none';
-                
+
                 if (data.success) {
                     window.location.href = data.redirect_url;
                 } else {
@@ -379,4 +389,5 @@
         });
     </script>
 </body>
+
 </html>

templates/report.html

@@ -20,7 +20,7 @@
         }
 
         .header {
-            background: rgba(10, 10, 10, 0.4);
+            background: #00000066;
             border-bottom: 1px solid rgba(255, 255, 255, 0.1);
             padding: 20px 0;
             position: sticky;