app.py

import gradio as gr
import google.generativeai as genai
import re
import html
import base64
from datetime import datetime
import json

class FermentationSEOGenerator:
    def __init__(self):
        self.model = None
        
    def configure_gemini(self, api_key):
        """Configure Gemini API with the provided key"""
        try:
            genai.configure(api_key=api_key)
            self.model = genai.GenerativeModel('gemini-1.5-pro')
            return True, "API configured successfully!"
        except Exception as e:
            return False, f"API configuration failed: {str(e)}"
    
    def parse_internal_links(self, internal_links_text):
        """Parse internal links from keyword:link format"""
        links_dict = {}
        if internal_links_text.strip():
            lines = internal_links_text.strip().split('\n')
            for line in lines:
                if ':' in line:
                    parts = line.split(':', 1)
                    if len(parts) == 2:
                        keyword = parts[0].strip()
                        link = parts[1].strip()
                        links_dict[keyword] = link
        return links_dict
    
    def parse_extra_keywords(self, extra_keywords_text):
        """Parse extra keywords from text area"""
        keywords = []
        if extra_keywords_text.strip():
            keywords = [kw.strip() for kw in extra_keywords_text.strip().split('\n') if kw.strip()]
        return keywords
    
    def calculate_readability(self, text):
        """Calculate basic readability metrics"""
        # Remove HTML tags for readability calculation
        clean_text = re.sub(r'<[^>]+>', '', text)
        
        # Count sentences, words, and syllables
        sentences = len(re.findall(r'[.!?]+', clean_text))
        words = len(clean_text.split())
        
        # Rough syllable count
        syllables = sum([max(1, len(re.findall(r'[aeiouAEIOU]', word))) for word in clean_text.split()])
        
        if sentences == 0 or words == 0:
            return {"flesch_score": 0, "grade_level": 0, "words": words, "sentences": sentences}
        
        # Flesch Reading Ease Score
        flesch_score = 206.835 - (1.015 * (words / sentences)) - (84.6 * (syllables / words))
        
        # Flesch-Kincaid Grade Level
        grade_level = (0.39 * (words / sentences)) + (11.8 * (syllables / words)) - 15.59
        
        return {
            "flesch_score": round(flesch_score, 1),
            "grade_level": round(max(0, grade_level), 1),
            "words": words,
            "sentences": sentences
        }
    
    def generate_seo_prompt(self, seo_plan, pov, tone, length, internal_links, extra_keywords, emotional_tone, generate_titles, include_readability):
        """Generate the comprehensive prompt for Gemini"""
        
        # Define content length targets
        length_targets = {
            "Short": "1500-2000 words",
            "Standard": "2500-3500 words", 
            "Long Form": "4000-6000 words"
        }
        
        internal_links_str = ""
        if internal_links:
            internal_links_str = "\n".join([f"- {keyword}: {link}" for keyword, link in internal_links.items()])
        
        extra_keywords_str = ", ".join(extra_keywords) if extra_keywords else ""
        
        prompt = f"""
You are an expert SEO content writer specializing in fermentation and gut health. Create a comprehensive, high-quality article optimized for 2025 SEO standards.

ARTICLE REQUIREMENTS:
- Topic/SEO Plan: {seo_plan}
- Target Length: {length_targets[length]}
- Point of View: {pov} person
- Tone: {tone}
- Emotional Tone: {emotional_tone}
- Niche Focus: Fermentation, gut health, probiotics, traditional food preservation

INTERNAL LINKS TO NATURALLY INCORPORATE:
{internal_links_str}

ADDITIONAL KEYWORDS TO INCLUDE:
{extra_keywords_str}

2025 SEO OPTIMIZATION REQUIREMENTS:
1. E-A-T Optimization: Include expert insights, scientific backing, personal experience elements
2. Featured Snippet Optimization: Use clear, concise answers to common questions
3. People Also Ask Integration: Address related questions naturally within content
4. Semantic SEO: Use related terms and context-rich language
5. User Intent Focus: Solve real problems and provide actionable advice

CONTENT STRUCTURE REQUIREMENTS:
- Compelling H1 with primary keyword
- Logical H2/H3 hierarchy with keyword variations
- Meta title (50-60 characters) and meta description (150-160 characters)
- Introduction hook with clear value proposition
- Step-by-step instructions where applicable
- Expert tips and troubleshooting sections
- FAQ section addressing common concerns
- Strong conclusion with call-to-action

WRITING STYLE:
- Human-like, conversational tone with {emotional_tone.lower()} emotional undertones
- Use transition words for flow
- Include personal anecdotes or expert experiences
- Balance scientific information with practical advice
- Write in {pov} person consistently
- Maintain {tone.lower()} tone throughout

HTML STRUCTURE & TECHNICAL REQUIREMENTS:
CRITICAL: Follow these HTML best practices exactly:

1. **Clean HTML Structure:**
   - Use semantic HTML5 elements: <section>, <article>, <header>
   - Wrap major content sections in <section> tags
   - No placeholder text for images/videos - use HTML comments instead
   - Clean, professional code structure

2. **Schema Markup:**
   - Complete FAQPage schema with proper opening/closing divs
   - Include at least 5-8 FAQ items with full schema markup
   - Ensure all schema containers are properly closed

3. **Content Organization:**
   - No inline questions within list items
   - Separate Q&A content into dedicated paragraphs or FAQ section
   - Use proper paragraph structure for readability

4. **Media Placeholders:**
   - Replace "(Include photos here)" with: <!-- Step-by-step photos go here -->
   - Replace "(Embed video)" with: <!-- Video tutorial embed -->
   - Use commented HTML examples for future media

5. **Links and References:**
   - Use real, working links when possible
   - If using example.com, make it clear it's a placeholder
   - Include rel="noopener" for external links

6. **Accessibility:**
   - Include lang="en" in html tag
   - Prepare alt text suggestions in comments for images
   - Use semantic heading hierarchy

EXAMPLE HTML STRUCTURE TO FOLLOW:
```html
<section>
  <h2>Section Title</h2>
  <p>Content here...</p>
  
  <h3>Subsection</h3>
  <ol>
    <li>Step one</li>
    <li>Step two</li>
  </ol>
  
  <p><strong>Common Question:</strong> Answer here...</p>
  <!-- Step-by-step photos go here -->
  <!-- <img src="path/image.jpg" alt="Description"> -->
</section>

<section>
  <h2>FAQ Section</h2>
  <div itemscope itemtype="https://schema.org/FAQPage">
    <div itemscope itemtype="https://schema.org/Question">
      <h3 itemprop="name">Question?</h3>
      <div itemprop="acceptedAnswer" itemscope itemtype="https://schema.org/Answer">
        <p itemprop="text">Answer here.</p>
      </div>
    </div>
    <!-- Add 5-8 more FAQ items -->
  </div> <!-- Close FAQPage -->
</section>
```

IMPORTANT: Generate ONLY the article content with proper HTML structure. Do NOT include any CSS styles, head tags, or full HTML document structure. Return only the article content that will be inserted into an existing page.

OUTPUT FORMAT:
Provide only the article content with:
1. Meta title and description at the top (as text, not HTML meta tags)
2. Full HTML-formatted article with proper semantic structure (content only)
3. Complete FAQ section with 6-8 questions and proper schema
4. Clean, professional HTML with no placeholder text
5. Commented placeholders for media elements
6. Natural integration of provided internal links

Generate the article now, ensuring it follows all HTML best practices and is optimized for 2025 SEO standards.
"""
        return prompt
    
    def generate_title_variations(self, main_topic):
        """Generate title variations using Gemini"""
        if not self.model:
            return ["Title generation requires API configuration"]
        
        prompt = f"""
Generate 5 SEO-optimized title variations for a fermentation article about: {main_topic}

Requirements:
- 50-60 characters each
- Include power words and emotional triggers
- Fermentation/gut health focus
- Different angles (beginner guide, expert tips, complete guide, etc.)
- Optimize for click-through rate

Format as a numbered list.
"""
        
        try:
            response = self.model.generate_content(prompt)
            return response.text.split('\n')
        except Exception as e:
            return [f"Error generating titles: {str(e)}"]
    
    def generate_article(self, api_key, seo_plan, pov, tone, length, internal_links_text, 
                        extra_keywords_text, emotional_tone, generate_titles, include_readability):
        """Main function to generate the SEO article"""
        
        # Configure API
        if not api_key.strip():
            return "❌ Error: Please enter your Gemini 1.5 Pro API key", "", "", "", ""
        
        success, message = self.configure_gemini(api_key)
        if not success:
            return f"❌ {message}", "", "", "", ""
        
        if not seo_plan.strip():
            return "❌ Error: Please provide your SEO article plan or keyword target", "", "", "", ""
        
        try:
            # Parse inputs
            internal_links = self.parse_internal_links(internal_links_text)
            extra_keywords = self.parse_extra_keywords(extra_keywords_text)
            
            # Generate main article
            prompt = self.generate_seo_prompt(
                seo_plan, pov, tone, length, internal_links, 
                extra_keywords, emotional_tone, generate_titles, include_readability
            )
            
            response = self.model.generate_content(prompt)
            article_content = response.text
            
            # Generate title variations if requested
            title_variations = ""
            if generate_titles:
                titles = self.generate_title_variations(seo_plan)
                title_variations = "🎯 **TITLE VARIATIONS:**\n" + "\n".join(titles) + "\n\n"
            
            # Calculate readability if requested
            readability_analysis = ""
            if include_readability:
                metrics = self.calculate_readability(article_content)
                readability_analysis = f"""
📊 **READABILITY ANALYSIS:**
- Word Count: {metrics['words']} words
- Sentences: {metrics['sentences']}
- Flesch Reading Ease: {metrics['flesch_score']}/100 (Higher is easier)
- Grade Level: {metrics['grade_level']} (U.S. grade level)

"""
            
            # Combine all outputs
            final_output = title_variations + readability_analysis + article_content
            
            # Create HTML version for preview - FIXED VERSION with contained CSS
            html_content = f"""
<div style="font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif; max-width: 100%; margin: 0; padding: 20px; line-height: 1.7; color: #333; background: #ffffff; box-sizing: border-box;">
    <div style="background: white; padding: 20px; border-radius: 8px; box-shadow: none; max-width: 100%; overflow-x: auto;">
        <div style="white-space: pre-wrap; word-wrap: break-word;">
            {final_output.replace(chr(10), '<br>') if not '<h1>' in final_output else final_output}
        </div>
    </div>
</div>
"""
            
            # Create downloadable HTML file (full document)
            download_html = f"""
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Generated SEO Article</title>
    <meta name="description" content="Generated fermentation SEO article optimized for 2025 standards">
    <style>
        body {{ font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif; max-width: 900px; margin: 0 auto; padding: 20px; line-height: 1.7; color: #333; background: #fafafa; }}
        .container {{ background: white; padding: 30px; border-radius: 8px; box-shadow: 0 2px 10px rgba(0,0,0,0.1); }}
        h1 {{ color: #2c3e50; border-bottom: 3px solid #27ae60; padding-bottom: 15px; font-size: 2.2em; }}
        h2 {{ color: #34495e; margin-top: 35px; margin-bottom: 15px; font-size: 1.6em; border-left: 4px solid #27ae60; padding-left: 15px; }}
        h3 {{ color: #7f8c8d; margin-top: 25px; font-size: 1.3em; }}
        section {{ margin: 30px 0; padding: 20px 0; border-bottom: 1px solid #eee; }}
        section:last-child {{ border-bottom: none; }}
        .meta-info {{ background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); color: white; padding: 20px; margin: 20px 0; border-radius: 8px; }}
        .readability {{ background: linear-gradient(135deg, #11998e 0%, #38ef7d 100%); color: white; padding: 20px; margin: 20px 0; border-radius: 8px; }}
        .titles {{ background: linear-gradient(135deg, #ffecd2 0%, #fcb69f 100%); padding: 20px; margin: 20px 0; border-radius: 8px; }}
        ul, ol {{ margin: 20px 0; padding-left: 25px; }}
        li {{ margin: 8px 0; }}
        .faq-section {{ background: #f8f9fa; padding: 20px; border-radius: 8px; margin: 20px 0; }}
        .faq-item {{ margin: 15px 0; }}
        a {{ color: #27ae60; text-decoration: none; font-weight: 500; }}
        a:hover {{ text-decoration: underline; }}
        code {{ background: #f1f2f6; padding: 2px 6px; border-radius: 3px; font-family: 'Courier New', monospace; }}
        .highlight {{ background: #fff3cd; padding: 2px 6px; border-radius: 3px; }}
        .schema-markup {{ background: #e3f2fd; border-left: 4px solid #2196f3; padding: 15px; margin: 15px 0; }}
    </style>
</head>
<body>
    <div class="container">
        {final_output}
    </div>
</body>
</html>
"""
            
            # Create markdown version
            markdown_content = final_output
            
            # Create downloadable HTML file
            html_download = base64.b64encode(download_html.encode()).decode()
            
            return (
                "✅ Article generated successfully!",
                final_output,
                html_content,
                markdown_content,
                html_download
            )
            
        except Exception as e:
            return f"❌ Error generating article: {str(e)}", "", "", "", ""

# Initialize the generator
generator = FermentationSEOGenerator()

# Create Gradio interface
def create_interface():
    with gr.Blocks(title="Fermentation AI SEO Tool", theme=gr.themes.Soft()) as app:
        gr.Markdown("""
        # 🧬 Fermentation AI SEO Tool – Powered by Gemini 1.5 Pro
        
        Generate high-quality, SEO-optimized articles focused on fermentation, gut health, and traditional food preservation.
        Optimized for 2025 SEO standards with E-A-T, featured snippets, and semantic optimization.
        """)
        
        with gr.Row():
            api_key = gr.Textbox(
                label="Enter Gemini 1.5 Pro API Key",
                placeholder="Enter your API key here...",
                type="password",
                lines=1
            )
        
        with gr.Row():
            seo_plan = gr.Textbox(
                label="Paste Your SEO Article Plan or Keyword Target",
                placeholder="Example: How to ferment vegetables at home for beginners, fermentation guide, lacto-fermentation process",
                lines=3
            )
        
        with gr.Row():
            with gr.Column():
                pov = gr.Dropdown(
                    choices=["First", "Second", "Third"],
                    value="Second",
                    label="Point of View"
                )
                tone = gr.Dropdown(
                    choices=["Friendly", "Professional", "Quirky"],
                    value="Friendly",
                    label="Blog Tone"
                )
            with gr.Column():
                length = gr.Dropdown(
                    choices=["Short", "Standard", "Long Form"],
                    value="Standard",
                    label="Content Length"
                )
                emotional_tone = gr.Dropdown(
                    choices=["Trust", "Curiosity", "Joy"],
                    value="Trust",
                    label="Emotional Tone Injection"
                )
        
        with gr.Row():
            with gr.Column():
                internal_links = gr.Textbox(
                    label="Internal Links (keyword:link format, 1 per line)",
                    placeholder="fermented garlic benefits:https://example.com/fermented-garlic\nkombucha brewing:https://example.com/kombucha-guide",
                    lines=4
                )
            with gr.Column():
                extra_keywords = gr.Textbox(
                    label="Extra Keywords (1 per line)",
                    placeholder="lacto-fermentation\nfermented vegetables\nprobiotics\ngut health",
                    lines=4
                )
        
        with gr.Row():
            generate_titles = gr.Checkbox(
                label="Generate Title Variations (3–5 options)",
                value=True
            )
            include_readability = gr.Checkbox(
                label="Include Readability Analysis (Flesch, Grade Level)",
                value=True
            )
        
        generate_btn = gr.Button("🚀 Generate Article", variant="primary", size="lg")
        
        with gr.Row():
            status = gr.Textbox(label="Status", interactive=False)
        
        with gr.Tabs():
            with gr.TabItem("📝 Article Output"):
                article_output = gr.Textbox(
                    label="Generated SEO Article",
                    lines=25,
                    max_lines=50,
                    show_copy_button=True
                )
            
            with gr.TabItem("🌐 HTML Preview"):
                html_preview = gr.HTML(label="HTML Preview")
                html_download = gr.File(
                    label="Download HTML File",
                    visible=False
                )
                download_btn = gr.Button("📥 Download HTML File", variant="secondary")
            
            with gr.TabItem("📄 Markdown"):
                markdown_output = gr.Textbox(
                    label="Markdown Format",
                    lines=20,
                    show_copy_button=True
                )
        
        # Event handlers
        def generate_and_process(*args):
            result = generator.generate_article(*args)
            status_msg, article, html_content, markdown, html_b64 = result
            
            # Create downloadable file
            if html_b64:
                timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
                filename = f"fermentation_article_{timestamp}.html"
                html_file = gr.File(value=None, visible=True)
                return status_msg, article, html_content, markdown, html_file
            
            return status_msg, article, html_content, markdown, gr.File(visible=False)
        
        def create_download_file(html_content):
            if html_content:
                timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
                filename = f"fermentation_article_{timestamp}.html"
                # Create full HTML document for download
                download_html = f"""
<!DOCTYPE html>
<html lang="en">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Generated SEO Article</title>
    <meta name="description" content="Generated fermentation SEO article optimized for 2025 standards">
    <style>
        body {{ font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif; max-width: 900px; margin: 0 auto; padding: 20px; line-height: 1.7; color: #333; background: #fafafa; }}
        .container {{ background: white; padding: 30px; border-radius: 8px; box-shadow: 0 2px 10px rgba(0,0,0,0.1); }}
        h1 {{ color: #2c3e50; border-bottom: 3px solid #27ae60; padding-bottom: 15px; font-size: 2.2em; }}
        h2 {{ color: #34495e; margin-top: 35px; margin-bottom: 15px; font-size: 1.6em; border-left: 4px solid #27ae60; padding-left: 15px; }}
        h3 {{ color: #7f8c8d; margin-top: 25px; font-size: 1.3em; }}
        section {{ margin: 30px 0; padding: 20px 0; border-bottom: 1px solid #eee; }}
        ul, ol {{ margin: 20px 0; padding-left: 25px; }}
        li {{ margin: 8px 0; }}
        a {{ color: #27ae60; text-decoration: none; font-weight: 500; }}
    </style>
</head>
<body>
    <div class="container">
        {html_content}
    </div>
</body>
</html>
"""
                with open(filename, 'w', encoding='utf-8') as f:
                    f.write(download_html)
                return gr.File(value=filename, visible=True)
            return gr.File(visible=False)
        
        generate_btn.click(
            fn=generate_and_process,
            inputs=[
                api_key, seo_plan, pov, tone, length, internal_links,
                extra_keywords, emotional_tone, generate_titles, include_readability
            ],
            outputs=[status, article_output, html_preview, markdown_output, html_download]
        )
        
        download_btn.click(
            fn=create_download_file,
            inputs=[html_preview],
            outputs=[html_download]
        )
        
        gr.Markdown("""
        ---
        ### 💡 **Updated for Clean HTML Generation:**
        - ✅ **Semantic HTML5** - Uses proper `<section>` tags and structure
        - ✅ **Complete Schema Markup** - Properly closed FAQ schema with 6-8 questions
        - ✅ **No Placeholder Text** - Clean HTML comments for media placeholders
        - ✅ **Accessibility Ready** - Proper lang attributes and semantic structure
        - ✅ **Professional Code** - No inline Q&A in lists, clean paragraph structure
        - ✅ **Fixed CSS Issues** - Contained styling that won't affect parent app
        
        ### 🎯 **Tips for Best Results:**
        - Use specific fermentation-related keywords in your SEO plan
        - Include 2-4 internal links with relevant anchor text (keyword:link format)
        - Add 5-10 related keywords for semantic SEO enhancement
        - Longer content typically performs better for competitive keywords
        - Generated HTML is production-ready with proper schema markup
        
        **Powered by Gemini 1.5 Pro API** | Built for Professional Fermentation Content
        """)
    
    return app

# Launch the app
if __name__ == "__main__":
    app = create_interface()
    app.launch(
        #server_name="0.0.0.0",
        #server_port=7860,
        #share=True,
        #show_error=True
    )