kryten-userstats 0.5.4

User statistics tracking microservice for CyTube channels

Kryten User Statistics Tracker

A microservice for tracking comprehensive user statistics in CyTube channels via the Kryten bridge.

Features

Core Statistics

• User Tracking: Records all seen usernames with first/last seen timestamps
• Message Counts: Tracks public messages per user per channel
• PM Counts: Tracks private messages from each user
• Channel Population: Snapshots every 5 minutes (connected users and active chatters)
• Media Changes: Logs all media title changes with timestamps
• User Activity Time:
  • Total time in channel
  • Active (not-AFK) time in channel
  • Uses CyTube's native setAFK events for accurate tracking

Emote Tracking

• Tracks emote usage per user per channel
• Detects hashtags matching the master emote list
• Example: #PogChamp or #Kappa

Kudos System

Two types of kudos tracking:

1. ++ Kudos: Detects ++username or username++ patterns
2. Phrase Kudos: Detects username near trigger phrases
   • Configurable phrases (e.g., "lol", "rofl", "haha", emojis)
   • Supports both "phrase username" and "username phrase" patterns
   • Includes repeating laughter detection (haha, hehe, hoho, etc.)

Username Aliases

• Configurable N+1 aliases per username
• Example: "kevinchrist" → ["kc", "kchrist", "kevin"]
• Kudos automatically resolved to canonical username

Data Exposure

• Prometheus Metrics: HTTP endpoint on port 28282 (configurable)
  • Service health metrics
  • Application statistics (users, messages, kudos, emotes, etc.)
  • Real-time active session count
• NATS Query Endpoints: Request/reply pattern for programmatic access
  • User statistics (messages, activity, kudos, emotes)
  • Channel leaderboards and population data
  • Media history
  • Global leaderboards
  • System health and stats

See QUERY_ENDPOINTS.md for complete API documentation.

Installation

Prerequisites

• Python 3.11+
• NATS server running
• Kryten-Robot bridge connected to target channel(s)
• kryten-py library

Quick Start

1. Clone or copy the repository

   cd d:\Devel\kryten-userstats
   

2. Copy configuration template

   # Windows
   Copy-Item config.example.json config.json
   

   # Linux/macOS
   cp config.example.json config.json
   

3. Edit configuration

   • Set your NATS connection details
   • Configure channels to monitor
   • Adjust snapshot interval (default 5 minutes)

4. Configure aliases (optional) Edit aliases.json to add username aliases

5. Start the tracker

   # Windows
   .\start-userstats.ps1
   

   # Linux/macOS
   chmod +x start-userstats.sh
   ./start-userstats.sh
   

Configuration

config.json

{
  "nats": {
    "servers": ["nats://localhost:4222"],
    "user": "${NATS_USER}",
    "password": "${NATS_PASSWORD}"
  },
  "channels": [
    {
      "domain": "cytu.be",
      "channel": "420grindhouse"
    }
  ],
  "database": {
    "path": "data/userstats.db"
  },
  "snapshots": {
    "interval_seconds": 300
  },
  "kudos": {
    "default_phrases": [
      "lol",
      "rofl",
      "lmao",
      "haha",
      "😂",
      "🤣",
      "😆"
    ]
  }
}

aliases.json

{
  "aliases": {
    "kevinchrist": ["kc", "kchrist", "kevin"],
    "exampleuser": ["example", "ex"]
  },
  "kudos_phrases": [
    "lol",
    "rofl",
    "lmao",
    "haha",
    "nice",
    "great",
    "awesome"
  ]
}

Database Schema

The tracker uses SQLite3 for persistent storage with the following tables:

Core Tables

• users - All seen usernames with timestamps
• user_aliases - Username → alias mappings
• message_counts - Public message counts per user/channel
• pm_counts - PM counts per user
• population_snapshots - Channel population every N minutes
• media_changes - Media title change log
• user_activity - Total and not-AFK time per user/channel

Emote & Kudos Tables

• emote_usage - Emote usage counts per user/channel/emote
• kudos_plusplus - ++ kudos counts per user/channel
• kudos_phrases - Phrase kudos counts per user/channel/phrase
• kudos_trigger_phrases - Configured trigger phrases

Usage Examples

Running in Background

# Windows
.\start-userstats.ps1 -Background

# Stop
Stop-Process -Id (Get-Content userstats.pid)

Custom Configuration

# Windows
.\start-userstats.ps1 -ConfigFile custom-config.json

# Linux/macOS
./start-userstats.sh custom-config.json

Monitoring Logs

# Windows
Get-Content -Path userstats.log -Wait

# Linux/macOS
tail -f userstats.log

Querying Statistics

CLI Tool (Recommended)

Use the included CLI tool for easy querying:

# System statistics
python query_cli.py system stats

# System health
python query_cli.py system health

# User statistics
python query_cli.py user alice
python query_cli.py user alice --channel 420grindhouse

# Leaderboards
python query_cli.py leaderboard messages --limit 10
python query_cli.py leaderboard kudos --limit 10
python query_cli.py leaderboard emotes --limit 20

# Channel statistics
python query_cli.py channel top 420grindhouse --limit 10
python query_cli.py channel media 420grindhouse --limit 20
python query_cli.py channel population 420grindhouse --hours 24

HTTP Metrics (Prometheus)

# View metrics
curl http://localhost:28282/metrics

# Or use the example script
python examples/metrics_example.py

NATS Queries (Advanced)

See QUERY_ENDPOINTS.md for complete API documentation and examples.

# Using nats-cli
nats request cytube.query.userstats.cytu.be.system.stats '{}'

# Using Python
python examples/query_example.py

Database Queries

Example SQL queries for common statistics:

Top Message Senders

SELECT username, channel, message_count 
FROM message_counts 
ORDER BY message_count DESC 
LIMIT 10;

Top Kudos Recipients (++)

SELECT username, channel, kudos_count 
FROM kudos_plusplus 
ORDER BY kudos_count DESC 
LIMIT 10;

Most Used Emotes

SELECT emote, SUM(usage_count) as total 
FROM emote_usage 
GROUP BY emote 
ORDER BY total DESC 
LIMIT 10;

Recent Media Changes

SELECT channel, timestamp, media_title 
FROM media_changes 
ORDER BY timestamp DESC 
LIMIT 20;

User Activity Report

SELECT 
    username, 
    channel,
    total_time_seconds / 3600.0 as hours_total,
    not_afk_time_seconds / 3600.0 as hours_active
FROM user_activity 
ORDER BY not_afk_time_seconds DESC;

Architecture

The tracker follows the kryten-misc pattern:

• Database Layer (database.py): SQLite3 operations with async executor
• Activity Tracker (activity_tracker.py): Monitors user sessions and AFK status
• Kudos Detector (kudos_detector.py): Regex-based ++ and phrase detection
• Emote Detector (emote_detector.py): Hashtag matching against emote list
• Main Application (main.py): Event handlers and coordination

Requirements

• Python 3.11+
• kryten-py >= 0.2.3
• NATS server
• Kryten-Robot bridge (connected to monitored channels)

Troubleshooting

Tracker Not Receiving Events

1. Check NATS connection: nats-cli sub "cytube.events.>"
2. Verify Kryten-Robot bridge is running and connected
3. Ensure channel names match between config and bridge

Database Locked Errors

• SQLite only allows one writer at a time
• Tracker uses async executor to prevent blocking
• If issues persist, check for other processes accessing the DB

Kudos Not Detected

1. Check trigger phrases in database: SELECT * FROM kudos_trigger_phrases
2. Verify username aliases: SELECT * FROM user_aliases
3. Test regex patterns with sample messages

License

MIT License - See kryten-py for details

Credits

Built using:

• kryten-py - CyTube microservices library
• NATS - Message bus
• SQLite3 - Database "# kryten-userstats"