Update app.py

app.py

@@ -17,6 +17,7 @@ from reportlab.lib.units import inch
 import time
 import tempfile
 import os
+from datetime import datetime
 
 # Set up logging
 logging.basicConfig(level=logging.INFO)
@@ -36,7 +37,6 @@ css = """
     --danger-color: #f44336;
     --info-color: #2196F3;
 }
-
 .dark {
     --primary-color: #4a89dc;
     --secondary-color: #3b7dd8;
@@ -49,7 +49,6 @@ css = """
     --danger-color: #F44336;
     --info-color: #2196F3;
 }
-
 body {
     background-image: url('https://images.unsplash.com/photo-1604147706283-d7119b5b822c?ixlib=rb-1.2.1&auto=format&fit=crop&w=1920&q=80');
     background-size: cover;
@@ -62,7 +61,6 @@ body {
     font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;
     color: var(--text-color);
 }
-
 .gradio-container {
     background-color: rgba(var(--bg-color), 0.97) !important;
     border-radius: 15px;
@@ -73,11 +71,9 @@ body {
     min-height: 90vh;
     border: 1px solid var(--primary-color) !important;
 }
-
 .risk-low { color: var(--success-color); font-weight: bold; }
 .risk-medium { color: var(--warning-color); font-weight: bold; }
 .risk-high { color: var(--danger-color); font-weight: bold; }
-
 .result-box {
     padding: 20px;
     border-radius: 10px;
@@ -87,7 +83,6 @@ body {
     border-left: 5px solid var(--primary-color) !important;
     color: var(--text-color);
 }
-
 .penalty-box {
     padding: 20px;
     border-radius: 10px;
@@ -96,7 +91,6 @@ body {
     background-color: var(--box-bg);
     box-shadow: 0 2px 8px rgba(0,0,0,0.08);
 }
-
 .obligation-box {
     padding: 20px;
     border-radius: 10px;
@@ -105,7 +99,6 @@ body {
     background-color: var(--box-bg);
     box-shadow: 0 2px 8px rgba(0,0,0,0.08);
 }
-
 .delay-box {
     padding: 20px;
     border-radius: 10px;
@@ -114,7 +107,6 @@ body {
     background-color: var(--box-bg);
     box-shadow: 0 2px 8px rgba(0,0,0,0.08);
 }
-
 .combined-risk-container {
     display: flex;
     flex-direction: column;
@@ -125,7 +117,6 @@ body {
     border-radius: 10px;
     box-shadow: 0 2px 8px rgba(0,0,0,0.08);
 }
-
 .risk-row {
     display: flex;
     align-items: center;
@@ -136,19 +127,16 @@ body {
     border: 1px solid var(--border-color) !important;
     transition: all 0.3s ease;
 }
-
 .risk-row:hover {
     transform: translateY(-2px);
     box-shadow: 0 4px 12px rgba(0,0,0,0.1);
 }
-
 .risk-label {
     width: 150px;
     font-weight: 600;
     font-size: 16px;
     color: var(--text-color) !important;
 }
-
 .risk-score {
     width: 120px;
     font-size: 20px;
@@ -156,7 +144,6 @@ body {
     padding: 8px 12px;
     border-radius: 6px;
 }
-
 .warning-box {
     padding: 18px;
     border-radius: 8px;
@@ -166,7 +153,6 @@ body {
     font-weight: 600;
     box-shadow: 0 2px 8px rgba(0,0,0,0.05);
 }
-
 .danger-box {
     padding: 18px;
     border-radius: 8px;
@@ -176,7 +162,6 @@ body {
     font-weight: 600;
     box-shadow: 0 2px 8px rgba(0,0,0,0.05);
 }
-
 .success-box {
     padding: 18px;
     border-radius: 8px;
@@ -186,7 +171,6 @@ body {
     font-weight: 600;
     box-shadow: 0 2px 8px rgba(0,0,0,0.05);
 }
-
 .info-box {
     padding: 18px;
     border-radius: 8px;
@@ -196,7 +180,6 @@ body {
     font-weight: 600;
     box-shadow: 0 2px 8px rgba(0,0,0,0.05);
 }
-
 .section-title {
     font-size: 22px;
     font-weight: 700;
@@ -206,7 +189,6 @@ body {
     align-items: center;
     gap: 10px;
 }
-
 .count-item {
     display: flex;
     justify-content: space-between;
@@ -214,12 +196,10 @@ body {
     border-bottom: 1px solid var(--border-color) !important;
     transition: all 0.2s ease;
 }
-
 .count-item:hover {
     background-color: rgba(0,0,0,0.05);
     transform: translateX(5px);
 }
-
 .count-label {
     font-weight: 600;
     color: var(--text-color) !important;
@@ -227,13 +207,11 @@ body {
     align-items: center;
     gap: 8px;
 }
-
 .count-value {
     color: var(--secondary-color) !important;
     font-weight: 600;
     font-size: 16px;
 }
-
 button {
     background: linear-gradient(135deg, var(--primary-color), var(--secondary-color)) !important;
     border: none !important;
@@ -244,13 +222,11 @@ button {
     transition: all 0.3s ease !important;
     box-shadow: 0 4px 6px rgba(0,0,0,0.1) !important;
 }
-
 button:hover {
     background: linear-gradient(135deg, var(--secondary-color), var(--primary-color)) !important;
     transform: translateY(-2px) !important;
     box-shadow: 0 6px 12px rgba(0,0,0,0.15) !important;
 }
-
 .upload-area {
     border: 2px dashed var(--primary-color) !important;
     background-color: rgba(240, 248, 255, 0.3) !important;
@@ -258,12 +234,10 @@ button:hover {
     padding: 30px !important;
     transition: all 0.3s ease !important;
 }
-
 .upload-area:hover {
     background-color: rgba(224, 255, 255, 0.3) !important;
     border-color: var(--secondary-color) !important;
 }
-
 .risk-meter {
     width: 100%;
     height: 20px;
@@ -272,7 +246,6 @@ button:hover {
     margin: 15px 0;
     position: relative;
 }
-
 .risk-meter-indicator {
     position: absolute;
     top: -5px;
@@ -281,7 +254,6 @@ button:hover {
     background-color: var(--text-color);
     transform: translateX(-50%);
 }
-
 .risk-meter-labels {
     display: flex;
     justify-content: space-between;
@@ -289,7 +261,6 @@ button:hover {
     font-size: 12px;
     color: var(--text-color);
 }
-
 .clause-example {
     background-color: rgba(0,0,0,0.05);
     padding: 15px;
@@ -300,13 +271,11 @@ button:hover {
     line-height: 1.5;
     color: var(--text-color);
 }
-
 .clause-number {
     font-weight: bold;
     color: var(--primary-color);
     margin-right: 8px;
 }
-
 .sentiment-meter {
     width: 100%;
     height: 20px;
@@ -314,13 +283,11 @@ button:hover {
     border-radius: 10px;
     margin: 15px 0;
 }
-
 .sentiment-score {
     height: 100%;
     border-radius: 10px;
     background-color: rgba(255,255,255,0.3);
 }
-
 /* Hide elements */
 footer, .gradio-footer, .hide, [data-testid="Use via API"], [data-testid="mmsettings"],
 #sentiment-analysis, #risk-visualization {
@@ -331,23 +298,19 @@ footer, .gradio-footer, .hide, [data-testid="Use via API"], [data-testid="mmsett
     padding: 0 !important;
     margin: 0 !important;
 }
-
 .file-info {
     margin-top: -15px;
     margin-bottom: 15px;
     color: var(--text-color);
     font-size: 13px;
 }
-
 /* Dark mode specific adjustments */
 .dark .clause-example {
     background-color: rgba(255,255,255,0.05);
 }
-
 .dark .risk-row {
     background-color: rgba(255,255,255,0.05);
 }
-
 .dark .count-item:hover {
     background-color: rgba(255,255,255,0.05);
 }
@@ -384,45 +347,180 @@ def get_hugging_face_sentiment(text: str) -> float:
         logger.error(f"Hugging Face sentiment analysis failed: {str(e)}. Using fallback score.")
         return 0.5
 
-def generate_sentiment_pdf(sentiment_score: float) -> BytesIO:
-    """Generate a PDF with sentiment analysis results in memory and return BytesIO buffer"""
+def generate_analysis_pdf(analysis_data: Dict) -> BytesIO:
+    """Generate a comprehensive PDF report with analysis results"""
     try:
         pdf_file = BytesIO()
         c = canvas.Canvas(pdf_file, pagesize=letter)
-
-        c.setFont("Helvetica-Bold", 16)
-        c.drawString(1 * inch, 10 * inch, "Sentiment Analysis Report")
-
-        c.setFont("Helvetica", 12)
-        c.drawString(1 * inch, 9.5 * inch, f"Date: {time.strftime('%Y-%m-%d')}")
-        c.drawString(1 * inch, 9.2 * inch, f"Time: {time.strftime('%H:%M:%S')}")
-
+        
+        # Header
+        c.setFont("Helvetica-Bold", 18)
+        c.drawString(1 * inch, 10.5 * inch, "Contract Risk Analysis Report")
+        c.setFont("Helvetica", 10)
+        c.drawString(1 * inch, 10.2 * inch, f"Generated on: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
+        c.drawString(1 * inch, 10 * inch, f"Document: {analysis_data.get('document_name', 'Unknown')}")
+        
+        # Add a line separator
+        c.line(1 * inch, 9.8 * inch, 7.5 * inch, 9.8 * inch)
+        
+        # Risk Summary Section
+        y_position = 9.5 * inch
         c.setFont("Helvetica-Bold", 14)
-        c.drawString(1 * inch, 8.7 * inch, f"Sentiment Score: {sentiment_score:.2f}")
-
-        y_position = 8.2 * inch
-        c.setFont("Helvetica-Bold", 12)
-        c.drawString(1 * inch, y_position, "Summary:")
+        c.drawString(1 * inch, y_position, "1. Risk Summary")
         y_position -= 0.3 * inch
-
+        
         c.setFont("Helvetica", 10)
-        summary_data = {
-            "Sentiment Score": f"{sentiment_score:.2f} (0.0 Negative to 1.0 Positive)",
-            "Interpretation": "Positive" if sentiment_score > 0.6 else "Negative" if sentiment_score < 0.4 else "Neutral",
-            "Analysis Timestamp": time.strftime("%Y-%m-%d %H:%M:%S")
+        risk_level = analysis_data['risk_level']
+        risk_color = {
+            "Low": "#4CAF50",
+            "Medium": "#FF9800",
+            "High": "#F44336"
+        }.get(risk_level, "#000000")
+        
+        c.setFillColor(risk_color)
+        c.setFont("Helvetica-Bold", 12)
+        c.drawString(1 * inch, y_position, f"Overall Risk Level: {risk_level}")
+        c.setFillColor("black")
+        
+        y_position -= 0.25 * inch
+        c.setFont("Helvetica", 10)
+        c.drawString(1 * inch, y_position, f"Risk Score: {analysis_data['risk_score']:.1f}/100")
+        y_position -= 0.25 * inch
+        
+        # Risk explanation
+        risk_explanations = {
+            "Low": "The contract appears to be low risk with favorable terms. Standard review recommended.",
+            "Medium": "The contract has moderate risk factors. Careful review of flagged clauses advised.",
+            "High": "The contract contains high-risk elements! Immediate legal review required."
         }
-
-        for key, value in summary_data.items():
-            c.drawString(1 * inch, y_position, f"{key}: {value}")
+        c.drawString(1 * inch, y_position, "Assessment:")
+        y_position -= 0.2 * inch
+        c.setFont("Helvetica", 10)
+        for line in textwrap.wrap(risk_explanations.get(risk_level, ""), width=80):
+            c.drawString(1.2 * inch, y_position, line)
+            y_position -= 0.2 * inch
+        
+        # Detailed Metrics Section
+        y_position -= 0.3 * inch
+        c.setFont("Helvetica-Bold", 14)
+        c.drawString(1 * inch, y_position, "2. Detailed Metrics")
+        y_position -= 0.3 * inch
+        
+        # Sentiment Analysis
+        c.setFont("Helvetica-Bold", 12)
+        c.drawString(1 * inch, y_position, "Sentiment Analysis:")
+        y_position -= 0.2 * inch
+        c.setFont("Helvetica", 10)
+        sentiment_score = analysis_data['sentiment_score']
+        sentiment_text = (
+            "Positive (favorable language)" if sentiment_score > 0.6 else
+            "Negative (adversarial language)" if sentiment_score < 0.4 else
+            "Neutral (balanced language)"
+        )
+        c.drawString(1.2 * inch, y_position, f"Score: {sentiment_score:.2f} - {sentiment_text}")
+        y_position -= 0.2 * inch
+        c.drawString(1.2 * inch, y_position, "Interpretation: Measures the overall tone of the contract language.")
+        y_position -= 0.25 * inch
+        
+        # Penalty Analysis
+        c.setFont("Helvetica-Bold", 12)
+        c.drawString(1 * inch, y_position, "Penalty Analysis:")
+        y_position -= 0.2 * inch
+        c.setFont("Helvetica", 10)
+        c.drawString(1.2 * inch, y_position, f"Total penalty clauses found: {analysis_data['penalty_count']}")
+        y_position -= 0.2 * inch
+        if analysis_data['penalty_values']:
+            c.drawString(1.2 * inch, y_position, f"Highest penalty amount: ${max(analysis_data['penalty_values']):,.2f}")
+            y_position -= 0.2 * inch
+            c.drawString(1.2 * inch, y_position, f"Average penalty amount: ${sum(analysis_data['penalty_values'])/len(analysis_data['penalty_values']):,.2f}")
+            y_position -= 0.2 * inch
+        c.drawString(1.2 * inch, y_position, "Interpretation: Penalties are financial consequences for non-compliance.")
+        y_position -= 0.25 * inch
+        
+        # Obligation Analysis
+        c.setFont("Helvetica-Bold", 12)
+        c.drawString(1 * inch, y_position, "Obligation Analysis:")
+        y_position -= 0.2 * inch
+        c.setFont("Helvetica", 10)
+        c.drawString(1.2 * inch, y_position, f"Total obligation clauses found: {analysis_data['obligation_count']}")
+        y_position -= 0.2 * inch
+        c.drawString(1.2 * inch, y_position, "Interpretation: Obligations are requirements that must be fulfilled.")
+        y_position -= 0.25 * inch
+        
+        # Delay Analysis
+        c.setFont("Helvetica-Bold", 12)
+        c.drawString(1 * inch, y_position, "Delay Analysis:")
+        y_position -= 0.2 * inch
+        c.setFont("Helvetica", 10)
+        c.drawString(1.2 * inch, y_position, f"Total delay clauses found: {analysis_data['delay_count']}")
+        y_position -= 0.2 * inch
+        c.drawString(1.2 * inch, y_position, "Interpretation: Delay clauses specify timelines and consequences for delays.")
+        y_position -= 0.3 * inch
+        
+        # Key Findings Section
+        if y_position < 2 * inch:
+            c.showPage()
+            y_position = 10.5 * inch
+            
+        c.setFont("Helvetica-Bold", 14)
+        c.drawString(1 * inch, y_position, "3. Key Findings")
+        y_position -= 0.3 * inch
+        c.setFont("Helvetica", 10)
+        
+        # Add key findings
+        findings = []
+        if analysis_data['risk_level'] == "High":
+            findings.append("⚠️ High-risk contract requiring immediate legal review")
+        if analysis_data['penalty_count'] > 5:
+            findings.append(f"⚠️ High number of penalty clauses ({analysis_data['penalty_count']})")
+        if analysis_data['obligation_count'] > 10:
+            findings.append(f"📝 Numerous obligations ({analysis_data['obligation_count']}) that may require tracking")
+        if analysis_data['sentiment_score'] < 0.4:
+            findings.append("🔍 Contract language appears adversarial (low sentiment score)")
+            
+        if not findings:
+            findings.append("✅ No major red flags detected in initial analysis")
+        
+        for finding in findings:
+            c.drawString(1 * inch, y_position, finding)
             y_position -= 0.25 * inch
-
+        
+        # Recommendations Section
+        y_position -= 0.3 * inch
+        c.setFont("Helvetica-Bold", 14)
+        c.drawString(1 * inch, y_position, "4. Recommendations")
+        y_position -= 0.3 * inch
+        c.setFont("Helvetica", 10)
+        
+        recommendations = []
+        if analysis_data['risk_level'] == "High":
+            recommendations.append("• Engage legal counsel for comprehensive review")
+            recommendations.append("• Negotiate penalty clauses and liability terms")
+        if analysis_data['penalty_count'] > 0:
+            recommendations.append("• Review all penalty clauses for fairness and applicability")
+        if analysis_data['obligation_count'] > 10:
+            recommendations.append("• Create an obligation tracking system")
+        if analysis_data['sentiment_score'] < 0.4:
+            recommendations.append("• Consider negotiating more balanced language")
+            
+        if not recommendations:
+            recommendations.append("• Standard contract review process sufficient")
+        
+        for rec in recommendations:
+            c.drawString(1 * inch, y_position, rec)
+            y_position -= 0.25 * inch
+        
+        # Footer
+        c.setFont("Helvetica-Oblique", 8)
+        c.drawString(1 * inch, 0.5 * inch, "Generated by Contract Risk Analyzer - Confidential")
+        
         c.save()
         pdf_file.seek(0)
-        logger.info("PDF generated in memory")
+        logger.info("PDF report generated successfully")
         return pdf_file
     except Exception as e:
-        logger.error(f"Error generating PDF: {str(e)}")
-        return None
+        logger.error(f"Error generating PDF report: {str(e)}")
+        raise Exception(f"PDF generation failed: {str(e)}")
 
 def save_to_salesforce(sf: Salesforce, data: Dict) -> str:
     """Save analysis results to Salesforce, return record ID"""
@@ -705,8 +803,21 @@ def analyze_pdf(file_obj) -> List:
             logger.error(f"Salesforce record creation failed: {str(e)}")
             salesforce_id = "N/A"
 
+        # Prepare data for PDF report
+        analysis_data = {
+            'document_name': os.path.basename(file_obj.name),
+            'sentiment_score': sentiment_score,
+            'risk_score': risk_score,
+            'risk_level': risk_level,
+            'penalty_count': total_penalties,
+            'penalty_values': penalty_values,
+            'obligation_count': total_obligations,
+            'delay_count': total_delays,
+            'record_id': record_id
+        }
+
         try:
-            pdf_buffer = generate_sentiment_pdf(sentiment_score)
+            pdf_buffer = generate_analysis_pdf(analysis_data)
             if pdf_buffer is None:
                 raise Exception("Failed to generate PDF")
             # Save to a temporary file for Gradio to serve
@@ -735,7 +846,14 @@ def analyze_pdf(file_obj) -> List:
             </div>
             {sentiment_meter}
             <div style='margin-top: 15px;'>
-                <strong>Sentiment Report:</strong> {'' if temp_file_path else 'Failed to generate PDF'}
+                <strong>Interpretation:</strong> { 
+                    "Positive (favorable language)" if sentiment_score > 0.6 else 
+                    "Negative (adversarial language)" if sentiment_score < 0.4 else 
+                    "Neutral (balanced language)" 
+                }
+            </div>
+            <div style='margin-top: 10px;'>
+                <strong>Full Report:</strong> Available for download below
             </div>
         </div>
         """
@@ -824,7 +942,7 @@ with gr.Blocks(css=css, title="PDF Contract Risk Analyzer", theme=gr.themes.Defa
     
     with gr.Row():
         sentiment_analysis = gr.HTML(label="Sentiment Analysis")
-        pdf_output = gr.File(label="Download Sentiment Report PDF", file_types=[".pdf"])
+        pdf_output = gr.File(label="Download Full Analysis Report (PDF)", file_types=[".pdf"])
     
     submit_btn.click(
         fn=analyze_pdf,