📎 Source:openclaw-sumopod — view on GitHub & star ⭐

📧 Smart Email Forward with PDF Data Extraction

Automate Invoice Processing with OpenClaw Sumopod

🎯 TL;DR (Too Long; Didn't Read)

What: Automatically find emails with PDF attachments, extract key data, and forward formatted summaries to your team. Time: 15 minutes setup Result: No more manual copy-paste from invoices! 🤖

┌─────────────┐    ┌─────────────┐    ┌─────────────┐    ┌─────────────┐
│   Gmail     │───▶│  Download   │───▶│   Extract   │───▶│   Forward   │
│  Invoice    │    │  PDF Attach │    │ Invoice Data│    │ Formatted   │
│   Email     │    │   to /tmp   │    │ (amount, #) │    │   Email     │
└─────────────┘    └─────────────┘    └─────────────┘    └─────────────┘

💡 Why Use This?

Real-World Scenario

Before (Manual Hell): 😫

1. Open Gmail
2. Find invoice email from vendor
3. Download PDF attachment
4. Open PDF, squint at numbers
5. Copy invoice number
6. Copy total amount  
7. Copy due date
8. Open new email
9. Type everything manually
10. Send to accounting@company.com
11. Repeat 50x per day... 💀

After (Automation Magic): ✨

1. Run workflow once
2. All invoices processed automatically
3. Accounting gets clean data
4. You get coffee ☕

Perfect for:

📊 Invoice processing
📑 Receipt collection
🧾 Expense report automation
📋 Purchase order tracking

🗺️ Visual Workflow Map

                    ┌─────────────────────────────────────┐
                    │         WORKFLOW OVERVIEW           │
                    └─────────────────────────────────────┘
                                      │
                                      ▼
    ┌───────────────────────────────────────────────────────────────┐
    │  STEP 1: SEARCH                                               │
    │  ┌─────────────┐                                              │
    │  │ Gmail Query │──▶ "from:vendor@acme.com has:attachment"    │
    │  └─────────────┘                                              │
    └───────────────────────────────────────────────────────────────┘
                                      │
                                      ▼
    ┌───────────────────────────────────────────────────────────────┐
    │  STEP 2: DOWNLOAD                                             │
    │  ┌─────────────┐    ┌─────────────┐                          │
    │  │ Get Email   │──▶ │ Save PDF    │                          │
    │  │ Attachments │    │ to /tmp/    │                          │
    │  └─────────────┘    └─────────────┘                          │
    └───────────────────────────────────────────────────────────────┘
                                      │
                                      ▼
    ┌───────────────────────────────────────────────────────────────┐
    │  STEP 3: EXTRACT                                              │
    │  ┌─────────────┐    ┌─────────────────────┐                  │
    │  │ Parse PDF   │──▶ │ Extract:            │                  │
    │  │ Text/Tables │    │ • Invoice Number    │                  │
    │  └─────────────┘    │ • Total Amount      │                  │
    │                     │ • Due Date          │                  │
    │                     │ • Vendor Name       │                  │
    │                     └─────────────────────┘                  │
    └───────────────────────────────────────────────────────────────┘
                                      │
                                      ▼
    ┌───────────────────────────────────────────────────────────────┐
    │  STEP 4: FORMAT & FORWARD                                     │
    │  ┌─────────────┐    ┌─────────────┐    ┌─────────────┐       │
    │  │ Build Clean │──▶ │ Format as   │──▶ │ Send to     │       │
    │  │ Data Object │    │ HTML/Markdown│    │ Recipient   │       │
    │  └─────────────┘    └─────────────┘    └─────────────┘       │
    └───────────────────────────────────────────────────────────────┘
                                      │
                                      ▼
                          ┌─────────────────────┐
                          │   ☕ COFFEE TIME!   │
                          └─────────────────────┘

🛠️ Prerequisites

Before we start, make sure you have:

Requirement	Status	How to Check
OpenClaw Gateway	✅ Required	`openclaw gateway status`
Gmail Access	✅ Required	Connected via OAuth
Python 3.8+	✅ Required	`python3 --version`
`pdftotext` or `PyPDF2`	✅ Required	`which pdftotext`

Install Required Tools

# Option 1: Install pdftotext (recommended, faster)
sudo apt-get install poppler-utils

# Option 2: Install Python PDF library
pip install PyPDF2 pdfplumber

🚀 Step-by-Step Implementation

Step 1: Create the Workflow Directory

mkdir -p ~/workflows/email-pdf-processor
cd ~/workflows/email-pdf-processor

Step 2: Create the Main Script

Create file: process_invoices.py

#!/usr/bin/env python3
"""
📧 Smart Email Forward with PDF Data Extraction
Automates invoice processing from Gmail
"""

import os
import re
import json
import tempfile
from datetime import datetime
from pathlib import Path

# ============================================================
# CONFIGURATION - Edit these for your needs
# ============================================================

CONFIG = {
    # Gmail search query - customize this!
    "search_query": "from:vendor@example.com has:attachment filename:pdf newer_than:1d",
    
    # Where to forward extracted data
    "forward_to": "accounting@acmecorp.com",
    
    # Email subject prefix
    "subject_prefix": "[AUTO-INVOICE]",
    
    # PDF download directory (auto-cleanup)
    "download_dir": "/tmp/invoice_pdfs",
    
    # Debug mode (prints extra info)
    "debug": True
}

# ============================================================
# STEP 1: Search Gmail for Invoice Emails
# ============================================================

def search_emails(query):
    """
    🔍 Search Gmail using gog CLI
    Returns list of emails matching query
    """
    import subprocess
    
    print(f"🔍 Searching Gmail: {query}")
    
    cmd = ["gog", "gmail", "search", query, "--max=10", "--json"]
    
    try:
        result = subprocess.run(cmd, capture_output=True, text=True)
        emails = json.loads(result.stdout)
        print(f"✅ Found {len(emails)} email(s)")
        return emails
    except Exception as e:
        print(f"❌ Search failed: {e}")
        return []

# ============================================================
# STEP 2: Download PDF Attachments
# ============================================================

def download_attachments(email_id, download_dir):
    """
    📥 Download all PDF attachments from an email
    Returns list of downloaded file paths
    """
    import subprocess
    import shutil
    
    os.makedirs(download_dir, exist_ok=True)
    downloaded = []
    
    # Get email details
    cmd = ["gog", "gmail", "get", email_id, "--include-attachments", "--save-to", download_dir]
    
    try:
        result = subprocess.run(cmd, capture_output=True, text=True)
        
        # Find PDF files in download directory
        for file in os.listdir(download_dir):
            if file.lower().endswith('.pdf'):
                full_path = os.path.join(download_dir, file)
                downloaded.append(full_path)
                print(f"  📄 Downloaded: {file}")
        
        return downloaded
    except Exception as e:
        print(f"❌ Download failed: {e}")
        return []

# ============================================================
# STEP 3: Extract Data from PDF
# ============================================================

def extract_pdf_data(pdf_path):
    """
    📊 Extract invoice data from PDF
    Uses regex patterns to find common invoice fields
    """
    import subprocess
    
    print(f"  🔍 Extracting: {os.path.basename(pdf_path)}")
    
    # Try pdftotext first (most reliable)
    try:
        result = subprocess.run(
            ["pdftotext", "-layout", pdf_path, "-"],
            capture_output=True,
            text=True,
            timeout=30
        )
        text = result.stdout
    except:
        # Fallback: Try PyPDF2
        try:
            import PyPDF2
            with open(pdf_path, 'rb') as f:
                reader = PyPDF2.PdfReader(f)
                text = ""
                for page in reader.pages:
                    text += page.extract_text() or ""
        except Exception as e:
            print(f"    ⚠️ Could not extract text: {e}")
            return None
    
    # Clean up text
    text = text.replace('\n', ' ').replace('\r', ' ')
    text = ' '.join(text.split())  # Remove extra spaces
    
    # ========================================================
    # PATTERN MATCHING - Customize these for your invoices!
    # ========================================================
    
    data = {
        "filename": os.path.basename(pdf_path),
        "extracted_text_snippet": text[:500] + "..." if len(text) > 500 else text,
        "fields": {}
    }
    
    # Pattern 1: Invoice Number (various formats)
    invoice_patterns = [
        r'[Ii]nvoice\s*[Nn]o\.?\s*:?\s*([A-Z0-9\-]+)',
        r'[Ii]nvoice\s*#\s*:?\s*([A-Z0-9\-]+)',
        r'[Nn]o\.?\s*[Ff]aktur\s*:?\s*([A-Z0-9\-]+)',
        r'[Ff]aktur\s*:?\s*([A-Z0-9\-]+)'
    ]
    
    for pattern in invoice_patterns:
        match = re.search(pattern, text)
        if match:
            data["fields"]["invoice_number"] = match.group(1).strip()
            break
    
    # Pattern 2: Total Amount (various formats)
    amount_patterns = [
        r'[Tt]otal\s*:?\s*Rp\.?\s*([\d.,]+)',
        r'[Tt]otal\s+[Aa]mount\s*:?\s*Rp\.?\s*([\d.,]+)',
        r'[Jj]umlah\s*:?\s*Rp\.?\s*([\d.,]+)',
        r'[Gg]rand\s+[Tt]otal\s*:?\s*Rp\.?\s*([\d.,]+)',
        r'[Tt]otal\s+[Pp]embayaran\s*:?\s*Rp\.?\s*([\d.,]+)'
    ]
    
    for pattern in amount_patterns:
        match = re.search(pattern, text)
        if match:
            amount_str = match.group(1).replace('.', '').replace(',', '.')
            data["fields"]["total_amount"] = f"Rp {amount_str}"
            break
    
    # Pattern 3: Due Date
    date_patterns = [
        r'[Dd]ue\s+[Dd]ate\s*:?\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
        r'[Tt]anggal\s+[Jj]atuh\s+[Tt]empo\s*:?\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
        r'[Jj]atuh\s+[Tt]empo\s*:?\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})'
    ]
    
    for pattern in date_patterns:
        match = re.search(pattern, text)
        if match:
            data["fields"]["due_date"] = match.group(1)
            break
    
    # Pattern 4: Vendor/Company Name
    vendor_patterns = [
        r'[Ff]rom\s*:?\s*([A-Z][A-Za-z\s]+(?:PT|CV|Ltd|Inc)?\.?)',
        r'[Dd]ari\s*:?\s*([A-Z][A-Za-z\s]+(?:PT|CV)?\.?)',
        r'([A-Z][A-Za-z\s]+(?:PT|CV)\s+[A-Za-z\s]+)'
    ]
    
    for pattern in vendor_patterns:
        match = re.search(pattern, text)
        if match:
            data["fields"]["vendor"] = match.group(1).strip()
            break
    
    # Pattern 5: Description/Items (first line)
    desc_match = re.search(r'[Dd]escription\s*:?\s*([^:]+?)(?=[Qq]ty|[Uu]nit|[Pp]rice)', text)
    if desc_match:
        data["fields"]["description"] = desc_match.group(1).strip()[:100]
    
    print(f"    ✅ Extracted {len(data['fields'])} fields")
    return data

# ============================================================
# STEP 4: Format and Forward Email
# ============================================================

def format_and_forward(extracted_data, original_email):
    """
    📤 Format extracted data and send forward email
    """
    import subprocess
    
    # Build HTML email body
    html_body = f"""
    <html>
    <head>
        <style>
            body {{ font-family: Arial, sans-serif; line-height: 1.6; }}
            .header {{ background: #4CAF50; color: white; padding: 20px; }}
            .content {{ padding: 20px; }}
            .invoice-card {{ border: 1px solid #ddd; border-radius: 8px; margin: 15px 0; padding: 15px; }}
            .field {{ margin: 8px 0; }}
            .label {{ font-weight: bold; color: #555; display: inline-block; width: 150px; }}
            .value {{ color: #333; }}
            .amount {{ font-size: 1.2em; color: #4CAF50; font-weight: bold; }}
            .footer {{ background: #f5f5f5; padding: 15px; font-size: 0.9em; color: #666; }}
        </style>
    </head>
    <body>
        <div class="header">
            <h2>📧 Automated Invoice Processing</h2>
            <p>Extracted from: {original_email.get('subject', 'Unknown')}</p>
        </div>
        <div class="content">
    """
    
    # Add each invoice
    for invoice in extracted_data:
        fields = invoice.get('fields', {})
        
        html_body += f"""
            <div class="invoice-card">
                <h3>📄 {invoice['filename']}</h3>
                <div class="field">
                    <span class="label">Invoice Number:</span>
                    <span class="value">{fields.get('invoice_number', 'N/A')}</span>
                </div>
                <div class="field">
                    <span class="label">Vendor:</span>
                    <span class="value">{fields.get('vendor', 'N/A')}</span>
                </div>
                <div class="field">
                    <span class="label">Total Amount:</span>
                    <span class="amount">{fields.get('total_amount', 'N/A')}</span>
                </div>
                <div class="field">
                    <span class="label">Due Date:</span>
                    <span class="value">{fields.get('due_date', 'N/A')}</span>
                </div>
                <div class="field">
                    <span class="label">Description:</span>
                    <span class="value">{fields.get('description', 'N/A')}</span>
                </div>
            </div>
        """
    
    html_body += f"""
        </div>
        <div class="footer">
            <p>⏰ Processed at: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}</p>
            <p>🤖 Automated by OpenClaw Sumopod</p>
        </div>
    </body>
    </html>
    """
    
    # Build plain text version
    text_body = f"""
AUTOMATED INVOICE PROCESSING
============================
Extracted from: {original_email.get('subject', 'Unknown')}

INVOICE DETAILS:
"""
    for invoice in extracted_data:
        fields = invoice.get('fields', {})
        text_body += f"""
---
File: {invoice['filename']}
Invoice Number: {fields.get('invoice_number', 'N/A')}
Vendor: {fields.get('vendor', 'N/A')}
Total Amount: {fields.get('total_amount', 'N/A')}
Due Date: {fields.get('due_date', 'N/A')}
Description: {fields.get('description', 'N/A')}
"""
    
    text_body += f"""
---
Processed at: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
Automated by OpenClaw Sumopod
"""
    
    # Send email using gog
    subject = f"{CONFIG['subject_prefix']} Invoice Data Extraction"
    
    print(f"\n📤 Forwarding to: {CONFIG['forward_to']}")
    
    # Save HTML to temp file for attachment
    with tempfile.NamedTemporaryFile(mode='w', suffix='.html', delete=False) as f:
        f.write(html_body)
        html_file = f.name
    
    try:
        cmd = [
            "gog", "gmail", "send",
            "--to", CONFIG['forward_to'],
            "--subject", subject,
            "--body", text_body,
            "--html-file", html_file
        ]
        
        result = subprocess.run(cmd, capture_output=True, text=True)
        
        if result.returncode == 0:
            print("✅ Email forwarded successfully!")
            return True
        else:
            print(f"❌ Failed to send: {result.stderr}")
            return False
    finally:
        os.unlink(html_file)

# ============================================================
# MAIN WORKFLOW
# ============================================================

def main():
    """
    🚀 Main execution flow
    """
    print("=" * 60)
    print("📧 Smart Email Forward with PDF Data Extraction")
    print("=" * 60)
    print()
    
    # Step 1: Search
    emails = search_emails(CONFIG["search_query"])
    
    if not emails:
        print("📭 No emails found matching criteria")
        return
    
    # Process each email
    for email in emails:
        print(f"\n{'─' * 50}")
        print(f"📨 Processing: {email.get('subject', 'No Subject')}")
        print(f"   From: {email.get('from', 'Unknown')}")
        print(f"   ID: {email.get('id', 'N/A')}")
        
        # Step 2: Download
        download_dir = os.path.join(CONFIG["download_dir"], email.get('id', 'unknown'))
        pdf_files = download_attachments(email.get('id'), download_dir)
        
        if not pdf_files:
            print("  ⚠️ No PDF attachments found")
            continue
        
        # Step 3: Extract
        extracted_data = []
        for pdf_file in pdf_files:
            data = extract_pdf_data(pdf_file)
            if data:
                extracted_data.append(data)
        
        if not extracted_data:
            print("  ⚠️ Could not extract data from any PDF")
            continue
        
        # Step 4: Forward
        format_and_forward(extracted_data, email)
        
        # Cleanup
        import shutil
        if os.path.exists(download_dir):
            shutil.rmtree(download_dir)
            print(f"  🧹 Cleaned up: {download_dir}")
    
    print(f"\n{'=' * 60}")
    print("✨ Processing complete!")
    print("=" * 60)

if __name__ == "__main__":
    main()

Step 3: Create Configuration File

Create file: config.json

{
  "search_query": "from:vendor@acmecorp.com has:attachment filename:pdf newer_than:1d",
  "forward_to": "accounting@acmecorp.com",
  "subject_prefix": "[AUTO-INVOICE]",
  "download_dir": "/tmp/invoice_pdfs",
  "debug": true,
  "extraction_patterns": {
    "invoice_number": [
      "[Ii]nvoice\\s*[Nn]o\\.?\\s*:?\\s*([A-Z0-9\\-]+)",
      "[Nn]o\\.?\\s*[Ff]aktur\\s*:?\\s*([A-Z0-9\\-]+)"
    ],
    "total_amount": [
      "[Tt]otal\\s*:?\\s*Rp\\.?\\s*([\\d.,]+)",
      "[Jj]umlah\\s*:?\\s*Rp\\.?\\s*([\\d.,]+)"
    ],
    "due_date": [
      "[Dd]ue\\s+[Dd]ate\\s*:?\\s*(\\d{1,2}[/-]\\d{1,2}[/-]\\d{2,4})",
      "[Jj]atuh\\s+[Tt]empo\\s*:?\\s*(\\d{1,2}[/-]\\d{1,2}[/-]\\d{2,4})"
    ]
  }
}

Step 4: Create Helper Scripts

Script 1: Test Mode (No Sending)

Create file: test_extract.py

#!/usr/bin/env python3
"""
🧪 Test PDF extraction without sending emails
Usage: python3 test_extract.py /path/to/invoice.pdf
"""

import sys
import json
import re
import subprocess

def extract_text_from_pdf(pdf_path):
    """Extract text from PDF using pdftotext"""
    try:
        result = subprocess.run(
            ["pdftotext", "-layout", pdf_path, "-"],
            capture_output=True,
            text=True,
            timeout=30
        )
        return result.stdout
    except Exception as e:
        print(f"❌ Error: {e}")
        return None

def extract_invoice_data(text):
    """Extract invoice fields from text"""
    text = text.replace('\n', ' ').replace('\r', ' ')
    text = ' '.join(text.split())
    
    data = {}
    
    # Invoice Number
    patterns = [
        r'[Ii]nvoice\s*[Nn]o\.?\s*:?\s*([A-Z0-9\-]+)',
        r'[Ff]aktur\s*:?\s*([A-Z0-9\-]+)'
    ]
    for p in patterns:
        m = re.search(p, text)
        if m:
            data['invoice_number'] = m.group(1)
            break
    
    # Amount
    patterns = [
        r'[Tt]otal\s*:?\s*Rp\.?\s*([\d.,]+)',
        r'[Jj]umlah\s*:?\s*Rp\.?\s*([\d.,]+)'
    ]
    for p in patterns:
        m = re.search(p, text)
        if m:
            data['total_amount'] = f"Rp {m.group(1)}"
            break
    
    # Date
    patterns = [
        r'[Dd]ue\s+[Dd]ate\s*:?\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
        r'[Jj]atuh\s+[Tt]empo\s*:?\s*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})'
    ]
    for p in patterns:
        m = re.search(p, text)
        if m:
            data['due_date'] = m.group(1)
            break
    
    return data

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print("Usage: python3 test_extract.py <pdf_file>")
        sys.exit(1)
    
    pdf_file = sys.argv[1]
    print(f"🔍 Testing extraction: {pdf_file}")
    print("-" * 50)
    
    text = extract_text_from_pdf(pdf_file)
    if text:
        print(f"📄 Text extracted ({len(text)} chars)")
        print("\n📝 First 500 characters:")
        print(text[:500])
        print("\n" + "-" * 50)
        
        data = extract_invoice_data(text)
        print("\n✅ Extracted Data:")
        print(json.dumps(data, indent=2, ensure_ascii=False))
    else:
        print("❌ Failed to extract text")

Make it executable:

chmod +x test_extract.py

Script 2: Dry Run Mode

Create file: dry_run.py

#!/usr/bin/env python3
"""
🧪 Dry Run - Test workflow without sending emails
Shows what WOULD be sent
"""

import subprocess
import json

# Same search as main script
SEARCH_QUERY = "from:vendor@example.com has:attachment filename:pdf newer_than:1d"

print("🔍 DRY RUN MODE - No emails will be sent")
print("=" * 50)

# Search emails
cmd = ["gog", "gmail", "search", SEARCH_QUERY, "--max=5", "--json"]
result = subprocess.run(cmd, capture_output=True, text=True)

try:
    emails = json.loads(result.stdout)
    print(f"\n📧 Would process {len(emails)} email(s):\n")
    
    for i, email in enumerate(emails, 1):
        print(f"{i}. {email.get('subject', 'No Subject')}")
        print(f"   From: {email.get('from', 'Unknown')}")
        print(f"   Date: {email.get('date', 'Unknown')}")
        print()
    
    print("✅ Dry run complete. No actions taken.")
    print("\nTo actually process, run: python3 process_invoices.py")
    
except json.JSONDecodeError:
    print("❌ Error parsing Gmail response")
    print(result.stderr)

🧪 Testing & Verification

Test 1: Verify PDF Extraction

# Test with a sample PDF
python3 test_extract.py /path/to/sample_invoice.pdf

Expected Output:

🔍 Testing extraction: sample_invoice.pdf
--------------------------------------------------
📄 Text extracted (2847 chars)

📝 First 500 characters:
INVOICE
From: PT Example Sejahtera
Invoice No: INV-2024-001
...

--------------------------------------------------

✅ Extracted Data:
{
  "invoice_number": "INV-2024-001",
  "total_amount": "Rp 15.000.000",
  "due_date": "15/03/2024"
}

Test 2: Dry Run

python3 dry_run.py

Expected Output:

🔍 DRY RUN MODE - No emails will be sent
==================================================

📧 Would process 3 email(s):

1. Invoice March 2024 - PT Example Sejahtera
   From: vendor@acmecorp.com
   Date: 2024-03-01

2. Service Invoice #245
   From: vendor@acmecorp.com
   Date: 2024-03-02

✅ Dry run complete. No actions taken.

Test 3: Full Run (with test email)

# First, modify config to forward to yourself
# Edit config.json: "forward_to": "your-email@example.com"

# Then run
python3 process_invoices.py

⚠️ Troubleshooting

Common Issues & Solutions

Issue 1: `pdftotext: command not found`

Error:

❌ Error: [Errno 2] No such file or directory: 'pdftotext'

Solution:

# Install poppler-utils (Linux)
sudo apt-get install poppler-utils

# Or use PyPDF2 fallback
pip install PyPDF2

Issue 2: Gmail Authentication Failed

Error:

❌ Search failed: Authentication required

Solution:

# Re-authenticate with gog
gog auth login

# Or check status
gog auth status

Issue 3: No Data Extracted from PDF

Error:

✅ Extracted 0 fields

Solutions:

Check if PDF is scanned (image-based)
# Install OCR tools sudo apt-get install tesseract-ocr pip install pytesseract pdf2image
Adjust regex patterns for your invoice format
# Add custom pattern in process_invoices.py custom_patterns = [ r'Your\s*Pattern\s*Here\s*:?\s*([A-Z0-9]+)', ]
Debug: Print full extracted text
# Add this to extract_pdf_data() if CONFIG["debug"]: print(f"Full text: {text[:2000]}")

Issue 4: Permission Denied on Download Directory

Error:

❌ Download failed: [Errno 13] Permission denied

Solution:

# Create directory with proper permissions
mkdir -p /tmp/invoice_pdfs
chmod 755 /tmp/invoice_pdfs

# Or change config to use different path
# Edit config.json: "download_dir": "/home/user/invoices"

Issue 5: Email Not Sending

Error:

❌ Failed to send: Error sending email

Solutions:

Check recipient email
# Verify email format

echo "accounting@acmecorp.com" | grep -E "^a-zA-Z0-9._%+-+@a-zA-Z0-9.-+.a-zA-Z{2,}$"


2. **Test gog send manually**
```bash
gog gmail send \
  --to "test@example.com" \
  --subject "Test" \
  --body "Hello World"

Check rate limits
# Add delay between sends import time time.sleep(2) # 2 second delay

🔧 Customization Guide

Adding Custom Extraction Fields

Edit process_invoices.py and add to the extract_pdf_data() function:

# Example: Extract PO Number
po_pattern = r'[Pp]\.?[Oo]\.?\s*[Nn]o\.?\s*:?\s*([A-Z0-9\-]+)'
match = re.search(po_pattern, text)
if match:
    data["fields"]["po_number"] = match.group(1)

# Example: Extract Tax Amount
tax_pattern = r'[Pp][Pp][Nn]\s*:?\s*Rp\.?\s*([\d.,]+)'
match = re.search(tax_pattern, text)
if match:
    data["fields"]["tax_amount"] = f"Rp {match.group(1)}"

Changing Email Template

Find the format_and_forward() function and modify:

# Custom subject
subject = f"🚨 URGENT: Invoice {fields.get('invoice_number')} requires payment"

# Custom body
html_body = f"""
<h1>Your Custom Template Here</h1>
<p>Invoice: {fields.get('invoice_number')}</p>
"""

Filtering by Attachment Name

Add to download_attachments():

# Only process files matching pattern
if not re.match(r'^INV-\d+\.pdf$', file, re.IGNORECASE):
    print(f"  ⏭️ Skipping: {file} (doesn't match pattern)")
    continue

📊 Monitoring & Logging

Add Logging to Your Script

import logging
from datetime import datetime

# Setup logging
logging.basicConfig(
    filename=f'invoice_processor_{datetime.now():%Y%m%d}.log',
    level=logging.INFO,
    format='%(asctime)s - %(levelname)s - %(message)s'
)

# Use in code
logging.info(f"Processing email: {email_id}")
logging.warning(f"No PDF found in email: {email_id}")
logging.error(f"Failed to extract: {str(e)}")

Create Processing Report

def generate_report(processed_emails, success_count, fail_count):
    report = f"""
    📊 Processing Report
    ===================
    Date: {datetime.now()}
    Total Emails: {len(processed_emails)}
    Successful: {success_count} ✅
    Failed: {fail_count} ❌
    
    Processed:
    {chr(10).join(f"  - {e}" for e in processed_emails)}
    """
    return report

🎯 Quick Reference Card

┌─────────────────────────────────────────────────────────┐
│           SMART EMAIL FORWARD - CHEATSHEET             │
├─────────────────────────────────────────────────────────┤
│  Install deps: sudo apt-get install poppler-utils      │
│  Test extract: python3 test_extract.py invoice.pdf      │
│  Dry run:      python3 dry_run.py                       │
│  Full run:     python3 process_invoices.py              │
├─────────────────────────────────────────────────────────┤
│  Common gog commands:                                   │
│  • gog gmail search "query" --json                     │
│  • gog gmail get EMAIL_ID --include-attachments        │
│  • gog gmail send --to X --subject Y --body Z          │
├─────────────────────────────────────────────────────────┤
│  Regex patterns for invoices:                           │
│  • INV-001: [A-Z]{2,3}-\d+                             │
│  • Rp 1.000.000: Rp\.?\s*[\d.,]+                       │
│  • 15/03/2024: \d{1,2}[/-]\d{1,2}[/-]\d{2,4}           │
└─────────────────────────────────────────────────────────┘

✅ Checklist Before Going Live

Test PDF extraction with sample invoices
Verify Gmail authentication working
Run dry_run.py to confirm email selection
Set forward_to to your email for testing
Test full workflow with 1-2 emails
Review extracted data accuracy
Update regex patterns if needed
Set up logging
Change forward_to to production email
Schedule with cron or n8n

🚀 Next Steps

Schedule Automatic Runs

Add to crontab for daily processing:

# Edit crontab
crontab -e

# Add line for daily run at 9 AM
0 9 * * * cd ~/workflows/email-pdf-processor && python3 process_invoices.py >> /var/log/invoice_processor.log 2>&1

Connect to n8n Workflow

Create new n8n workflow
Add "Execute Command" node
Command: cd ~/workflows/email-pdf-processor && python3 process_invoices.py
Schedule with "Cron" trigger node

📞 Need Help?

💬 Check OpenClaw documentation: openclaw --help
🔍 Review gog CLI docs: gog --help
🐛 Report issues with logs and sample PDFs

Created for OpenClaw Sumopod | 🤖 Automated with love

Last updated: March 2025

📧 Smart Email Forward with PDF Data Extraction

Automate Invoice Processing with OpenClaw Sumopod

🎯 TL;DR (Too Long; Didn't Read)

💡 Why Use This?

Real-World Scenario

🗺️ Visual Workflow Map

🛠️ Prerequisites

Install Required Tools

🚀 Step-by-Step Implementation

Step 1: Create the Workflow Directory

Step 2: Create the Main Script

Step 3: Create Configuration File

Step 4: Create Helper Scripts

Script 1: Test Mode (No Sending)

Script 2: Dry Run Mode

🧪 Testing & Verification

Test 1: Verify PDF Extraction

Test 2: Dry Run

Test 3: Full Run (with test email)

⚠️ Troubleshooting

Common Issues & Solutions

Issue 1: pdftotext: command not found

Issue 2: Gmail Authentication Failed

Issue 3: No Data Extracted from PDF

Issue 4: Permission Denied on Download Directory

Issue 5: Email Not Sending

🔧 Customization Guide

Adding Custom Extraction Fields

Changing Email Template

Filtering by Attachment Name

📊 Monitoring & Logging

Add Logging to Your Script

Create Processing Report

🎯 Quick Reference Card

✅ Checklist Before Going Live

🚀 Next Steps

Schedule Automatic Runs

Connect to n8n Workflow

📞 Need Help?

Ada Pertanyaan? Yuk Ngobrol!

Issue 1: `pdftotext: command not found`