shenseanchen/launch-agent-skills/whatsapp-parser

[DD/MM/YYYY, HH:MM:SS] Sender Name: Message content

with open(chat_file_path, 'r', encoding='utf-8') as f:
    content = f.read()

import re
from datetime import datetime
# Pattern for WhatsApp message format (handles multiple date formats)
patterns = [
    # EU format: [DD/MM/YYYY, HH:MM:SS]
    r'\[(\d{2}/\d{2}/\d{4}),\s*(\d{2}:\d{2}:\d{2})\]\s*([^:]+):\s*(.*)',
    # US format: [MM/DD/YY, H:MM:SS AM/PM]
    r'\[(\d{2}/\d{2}/\d{2}),\s*(\d{1,2}:\d{2}:\d{2}\s*[AP]M)\]\s*([^:]+):\s*(.*)',
]
messages = []
current_message = None
for line in content.split('\n'):
    matched = False
    for pattern in patterns:
        match = re.match(pattern, line)
        if match:
            if current_message:
                messages.append(current_message)
            
            date_str, time_str, sender, text = match.groups()
            current_message = {
                'date': date_str,
                'time': time_str,
                'sender': sender.strip(),
                'content': text.strip(),
                'is_media': '<media omitted>' in text.lower() or '<omitted>' in text.lower()
            }
            matched = True
            break
    
    # Handle multi-line messages
    if not matched and current_message and line.strip():
        current_message['content'] += '\n' + line
if current_message:
    messages.append(current_message)

participants = list(set(msg['sender'] for msg in messages))
date_range = {
    'start': messages[0]['date'] if messages else None,
    'end': messages[-1]['date'] if messages else None
}
message_count = len(messages)
media_count = sum(1 for msg in messages if msg['is_media'])

{
  "metadata": {
    "participants": ["Alice Developer", "Bob Engineer"],
    "message_count": 42,
    "media_count": 3,
    "date_range": {
      "start": "28/01/2026",
      "end": "28/01/2026"
    }
  },
  "messages": [
    {
      "timestamp": "28/01/2026 09:15:32",
      "sender": "Alice Developer",
      "content": "Hey Bob, did you push the code?",
      "is_media": false
    }
  ]
}