modules/pdf_report.py

import io
from datetime import datetime
from reportlab.lib.pagesizes import letter
from reportlab.lib import colors
from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle, PageBreak, HRFlowable

def build_pdf(ai_report_text, summary_data, flagged_df):
    """
    Generate the AML Compliance Monitoring PDF using ReportLab
    Returns PDF file as bytes
    """
    buffer = io.BytesIO()
    doc = SimpleDocTemplate(buffer, pagesize=letter, rightMargin=72, leftMargin=72, topMargin=72, bottomMargin=18)
    
    styles = getSampleStyleSheet()
    
    title_style = ParagraphStyle(
        "CoverTitle",
        parent=styles['Heading1'],
        fontName="Helvetica-Bold",
        fontSize=24,
        textColor=colors.HexColor("#0a1628"),
        alignment=1, # Center
        spaceAfter=20
    )
    
    subtitle_style = ParagraphStyle(
        "CoverSubtitle",
        parent=styles['Heading2'],
        fontName="Helvetica",
        fontSize=16,
        textColor=colors.HexColor("#e63946"),
        alignment=1,
        spaceAfter=40
    )
    
    normal_center = ParagraphStyle(
        "NormalCenter",
        parent=styles['Normal'],
        fontName="Helvetica",
        fontSize=12,
        alignment=1,
        spaceAfter=10
    )
    
    confidential_style = ParagraphStyle(
        "Confidential",
        parent=styles['Normal'],
        fontName="Helvetica-Bold",
        fontSize=12,
        textColor=colors.HexColor("#e63946"),
        alignment=1,
        spaceBefore=100
    )
    
    # Body styles
    section_header_style = ParagraphStyle(
        "SectionHeader",
        parent=styles['Heading2'],
        fontName="Helvetica-Bold",
        fontSize=13,
        textColor=colors.HexColor("#e63946"),
        spaceBefore=15,
        spaceAfter=10
    )
    
    body_style = ParagraphStyle(
        "ReportBody",
        parent=styles['Normal'],
        fontName="Helvetica",
        fontSize=10,
        textColor=colors.HexColor("#1a1a2e"),
        leading=14 # line spacing 1.4 (10 * 1.4)
    )
    
    story = []
    
    # ------------------
    # Page 1: Cover Page
    # ------------------
    story.append(Spacer(1, 100))
    story.append(Paragraph("AML COMPLIANCE MONITORING REPORT", title_style))
    story.append(Paragraph("AML Shield — Powered by AI", subtitle_style))
    
    story.append(HRFlowable(width="80%", thickness=2, color=colors.HexColor("#e63946"), spaceBefore=20, spaceAfter=20))
    
    story.append(Paragraph(f"Date: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}", normal_center))
    story.append(Paragraph(f"Dataset: {summary_data.get('filename', 'Unknown')}", normal_center))
    story.append(Paragraph(f"Date Range: {summary_data.get('date_range', 'Unknown')}", normal_center))
    story.append(Paragraph(f"Total Transactions: {summary_data.get('total_transactions', 0)}", normal_center))
    
    story.append(Paragraph("CONFIDENTIAL — For Internal Use Only.<br/>This report contains sensitive AML analysis.", confidential_style))
    story.append(PageBreak())
    
    # ------------------
    # Pages 2+: AI Report Body
    # ------------------
    sections = [
        "1. EXECUTIVE SUMMARY",
        "2. KEY FINDINGS",
        "3. HIGH RISK TRANSACTIONS",
        "4. CUSTOMER RISK ASSESSMENT (KYC)",
        "5. REGULATORY IMPLICATIONS",
        "6. RECOMMENDATIONS",
        "7. CONCLUSION"
    ]
    
    # Simple parser to split by sections
    # Find all instances of these headers in the text and format them
    text = ai_report_text
    
    # Just a basic approach: we find the positions of headers, and slice
    # However, text may not perfectly match these if AI varies slightly.
    # We will just split paragraphs and check if they start with a number + section name logic
    paragraphs = text.replace('\r\n', '\n').split('\n\n')
    
    for p in paragraphs:
        p = p.strip()
        if not p: continue
        
        # Check if it looks like a header (starts with number and is short, or matches our section list)
        is_header = False
        for sec in sections:
            if p.upper().startswith(sec) or p.startswith(f"**{sec}"):
                is_header = True
                p_clean = p.replace("**", "") # Remove markdown bold
                story.append(HRFlowable(width="100%", thickness=1, color=colors.lightgrey, spaceBefore=10, spaceAfter=5))
                story.append(Paragraph(p_clean, section_header_style))
                break
        
        if not is_header:
            # Clean up markdown specifics for body
            p_clean = p.replace("**", "")
            # Markdown bullet points to text representation
            if p_clean.startswith("- ") or p_clean.startswith("* "):
                p_clean = p_clean.replace("- ", r"&bull; ", 1)
                p_clean = p_clean.replace("* ", r"&bull; ", 1)
                
            story.append(Paragraph(p_clean, body_style))
            story.append(Spacer(1, 10))

    story.append(PageBreak())
    
    # ------------------
    # Final Page: Top 20 Flagged Transactions Table
    # ------------------
    story.append(Paragraph("Top 20 Flagged Transactions", section_header_style))
    story.append(Spacer(1, 15))
    
    if not flagged_df.empty:
        top_20 = flagged_df.sort_values(by='risk_score', ascending=False).head(20)
        
        table_data = []
        headers = ['Transaction ID', 'Customer', 'Amount', 'Type', 'Risk Score', 'Risk Level', 'Rules Triggered']
        table_data.append(headers)
        
        for idx, row in top_20.iterrows():
            rules = row['rule_flags']
            rule_str = ", ".join(rules) if isinstance(rules, list) else str(rules)
            amount = f"${row['amount']:,.2f}"
            row_data = [
                str(row['transaction_id']),
                str(row['customer_id']),
                amount,
                str(row['transaction_type']),
                f"{row['risk_score']:.1f}",
                str(row['risk_level']),
                rule_str
            ]
            table_data.append(row_data)
            
        # Table styling
        ts = TableStyle([
            ('BACKGROUND', (0,0), (-1,0), colors.HexColor("#0a1628")),
            ('TEXTCOLOR', (0,0), (-1,0), colors.white),
            ('ALIGN', (0,0), (-1,-1), 'LEFT'),
            ('FONTNAME', (0,0), (-1,0), 'Helvetica-Bold'),
            ('FONTSIZE', (0,0), (-1,0), 10),
            ('BOTTOMPADDING', (0,0), (-1,0), 12),
            ('GRID', (0,0), (-1,-1), 1, colors.black),
            ('FONTNAME', (0,1), (-1,-1), 'Helvetica'),
            ('FONTSIZE', (0,1), (-1,-1), 8),
            ('VALIGN', (0,0), (-1,-1), 'MIDDLE'),
        ])
        
        # Add alternating row colors based on risk
        for i, row in enumerate(top_20.itertuples(), start=1):
            if row.risk_level == 'High':
                ts.add('BACKGROUND', (0,i), (-1,i), colors.HexColor("#ffe0e0"))
            elif row.risk_level == 'Medium':
                ts.add('BACKGROUND', (0,i), (-1,i), colors.HexColor("#fff9c4"))
        
        t = Table(table_data, repeatRows=1)
        t.setStyle(ts)
        story.append(t)
    else:
        story.append(Paragraph("No flagged transactions to display.", body_style))

    doc.build(story)
    pdf_bytes = buffer.getvalue()
    buffer.close()
    
    return pdf_bytes