Update app.py

app.py

@@ -6,8 +6,6 @@ import pickle
 from huggingface_hub import hf_hub_download
 from sklearn.metrics.pairwise import cosine_similarity
 import httpx
-from bs4 import BeautifulSoup
-import re
 
 # ============================================
 # CONFIGURATION
@@ -42,7 +40,8 @@ class KnowledgeBase:
             self.embeddings = data['embeddings']
             self.loaded = True
             
-            return True, f"✅ Loaded {len(self.knowledge_base)} searchable paragraphs from {len(set(p['url'] for p in self.knowledge_base))} blog posts"
+            num_posts = len(set(p['url'] for p in self.knowledge_base))
+            return True, f"✅ Loaded {len(self.knowledge_base)} paragraphs from {num_posts} blog posts"
             
         except Exception as e:
             return False, f"❌ Error: {str(e)}"
@@ -123,7 +122,6 @@ class OrphanPageAnalyzer:
     def analyze(self, orphan_url, num_sources=3):
         """
         Complete analysis: Find sources, placements, and generate report
-        Returns: markdown report with implementation details
         """
         
         # Get orphan page metadata
@@ -249,23 +247,174 @@ Provide ONLY the modified sentence with the anchor text naturally integrated."""
     def generate_report(self, orphan_url, orphan_title, results):
         """Generate markdown report"""
         
-        report = f"""# 🔗 Internal Linking Report
+        report = f"# 🔗 Internal Linking Report\n\n"
+        report += f"**Orphan Page:** {orphan_title}\n"
+        report += f"**Target URL:** `{orphan_url}`\n"
+        report += f"**Links Found:** {len(results)}\n\n"
+        report += "---\n\n"
+        
+        for i, result in enumerate(results, 1):
+            report += f"## Link {i}: {result['source_title']}\n\n"
+            report += f"**Source URL:** `{result['source_url']}`\n"
+            report += f"**Paragraph #:** {result['paragraph_index']}\n"
+            report += f"**Relevance Score:** {result['score']}/100\n"
+            report += f"**Anchor Text:** \"{result['anchor_text']}\"\n\n"
+            
+            report += "### Current Sentence:\n"
+            report += "```\n"
+            report += result['current_sentence'] + "\n"
+            report += "```\n\n"
+            
+            report += "### New Sentence (with link):\n"
+            report += "```\n"
+            report += result['new_sentence'] + "\n"
+            report += "```\n\n"
+            
+            report += "### HTML Code:\n"
+            report += "```html\n"
+            html_code = result['new_sentence'].replace(
+                result['anchor_text'],
+                f'<a href="{result["target_url"]}">{result["anchor_text"]}</a>'
+            )
+            report += html_code + "\n"
+            report += "```\n\n"
+            report += "---\n\n"
+        
+        return report
 
-**Orphan Page:** {orphan_title}  
-**Target URL:** `{orphan_url}`  
-**Links Found:** {len(results)}
+# ============================================
+# GLOBAL STATE
+# ============================================
+kb = KnowledgeBase()
+analyzer = None
 
----
+# ============================================
+# GRADIO FUNCTIONS
+# ============================================
+def setup(api_key, hf_token):
+    """Setup API and load knowledge base"""
+    global analyzer
+    
+    status = []
+    
+    # Setup API
+    if not api_key or not api_key.strip():
+        return "❌ Please enter your OpenRouter API key", None
+    
+    try:
+        client = OpenRouterClient(api_key)
+        status.append("✅ API key configured")
+    except Exception as e:
+        return f"❌ API Error: {str(e)}", None
+    
+    # Load knowledge base
+    token = hf_token.strip() if hf_token else None
+    success, message = kb.load_from_huggingface(HF_DATASET_REPO, token)
+    
+    if not success:
+        return f"✅ API key configured\n{message}", None
+    
+    status.append(message)
+    
+    # Create analyzer
+    analyzer = OrphanPageAnalyzer(kb, client)
+    status.append("✅ System ready!")
+    
+    return "\n".join(status), None
 
-"""
-        
-        for i, result in enumerate(results, 1):
-            report += f"""
-## Link {i}: {result['source_title']}
+def analyze_orphan(orphan_url, num_sources):
+    """Analyze orphan page and generate report"""
+    
+    if not analyzer:
+        return "❌ Please complete setup first", None
+    
+    if not orphan_url or not orphan_url.strip():
+        return "❌ Please enter an orphan page URL", None
+    
+    try:
+        report, table = analyzer.analyze(orphan_url, num_sources)
+        return report, table
+    except Exception as e:
+        return f"❌ Error: {str(e)}", None
 
-**Source URL:** `{result['source_url']}`  
-**Paragraph #:** {result['paragraph_index']}  
-**Relevance Score:** {result['score']}/100  
-**Anchor Text:** "{result['anchor_text']}"
+# ============================================
+# INTERFACE
+# ============================================
+with gr.Blocks(title="Edstellar Internal Linking Tool", theme=gr.themes.Soft()) as app:
+    
+    gr.Markdown("# 🔗 Edstellar Internal Linking Tool")
+    gr.Markdown("Enter an orphan page URL to get instant internal linking recommendations")
+    
+    # Setup Section
+    with gr.Accordion("⚙️ Setup (Click to expand - Do this once)", open=True):
+        gr.Markdown("### Step 1: Configure API Keys")
+        
+        with gr.Row():
+            api_key = gr.Textbox(
+                label="OpenRouter API Key",
+                placeholder="sk-or-v1-...",
+                type="password",
+                scale=2
+            )
+            hf_token = gr.Textbox(
+                label="Hugging Face Token (optional)",
+                placeholder="hf_...",
+                type="password",
+                scale=2
+            )
+        
+        setup_btn = gr.Button("🚀 Setup System", variant="primary", size="lg")
+        setup_status = gr.Textbox(label="Setup Status", lines=3, interactive=False)
+    
+    gr.Markdown("---")
+    
+    # Analysis Section
+    gr.Markdown("### 📊 Analyze Orphan Page")
+    
+    with gr.Row():
+        orphan_url_input = gr.Textbox(
+            label="Orphan Page URL",
+            placeholder="https://edstellar.com/blog/your-orphan-page",
+            scale=3
+        )
+        num_sources_input = gr.Slider(
+            label="Number of Sources",
+            minimum=3,
+            maximum=5,
+            value=3,
+            step=1,
+            scale=1
+        )
+    
+    analyze_btn = gr.Button("🔍 Analyze & Generate Report", variant="primary", size="lg")
+    
+    gr.Markdown("---")
+    
+    # Results Section
+    gr.Markdown("### 📄 Report")
+    
+    report_output = gr.Markdown()
+    
+    gr.Markdown("### 📊 Summary Table")
+    table_output = gr.Dataframe(
+        label="Quick Overview",
+        wrap=True,
+        interactive=False
+    )
+    
+    # Wire up events
+    setup_btn.click(
+        setup,
+        inputs=[api_key, hf_token],
+        outputs=[setup_status, table_output]
+    )
+    
+    analyze_btn.click(
+        analyze_orphan,
+        inputs=[orphan_url_input, num_sources_input],
+        outputs=[report_output, table_output]
+    )
 
-### Current Sentence:
+# Launch
+if __name__ == "__main__":
+    app.launch()