This HTTP Val converts text files to speech using OpenAI's TTS API and stores the results for podcast generation.

1. Set the OPENAI_API_KEY environment variable in your Val Town settings
2. Set the AUTH_TOKEN environment variable for API authorization
3. The SQLite database table will be created automatically on first use

1. Visit /test for a web interface to upload and convert text files
2. Or use the API endpoints directly (see below)

Web interface for testing file uploads and conversions.

Converts a text file to speech and stores it. Requires authorization.

Headers:

• Authorization: Bearer YOUR_AUTH_TOKEN

Parameters:

• file (File): Text file to convert
• name (string): Name/title for the episode

Example using curl:

curl -X POST https://your-val-url.web.val.run/convert \
  -H "Authorization: Bearer your-auth-token-here" \
  -F "file=@your-text-file.txt" \
  -F "name=My First Episode"

Response:

{
  "success": true,
  "message": "Text converted to speech successfully",
  "data": {
    "name": "My First Episode",
    "mp3_url": "https://your-val-url.web.val.run/audio/tts_1234567890.mp3",
    "text_length": 1250
  }
}

Serves MP3 audio files from storage.

Lists all converted episodes.

Response:

{
  "success": true,
  "episodes": [
    {
      "id": 1,
      "name": "My First Episode",
      "text": "This is the text content...",
      "mp3_url": "https://your-val-url.web.val.run/audio/tts_1234567890.mp3",
      "date_added": "2024-01-01 12:00:00"
    }
  ]
}

Get database statistics and cleanup information.

Response:

{
  "success": true,
  "stats": {
    "totalEpisodes": 5,
    "recentEpisodes": 3,
    "oldEpisodes": 2,
    "oldestEpisode": {
      "name": "First Episode",
      "date_added": "2024-01-01T12:00:00.000Z"
    },
    "newestEpisode": {
      "name": "Latest Episode", 
      "date_added": "2024-01-15T12:00:00.000Z"
    },
    "cleanupInfo": {
      "eligibleForCleanup": 2,
      "cleanupThreshold": "30 days",
      "nextCleanup": "Weekly (automated via cron)"
    }
  }
}

Manually trigger cleanup of old episodes. Requires authorization.

Headers:

• Authorization: Bearer YOUR_AUTH_TOKEN

Body (optional):

{
  "days": 30
}

Parameters:

• days (number, optional): Delete episodes older than this many days (1-365, default: 30)

Response:

{
  "success": true,
  "message": "Cleanup completed for episodes older than 30 days",
  "result": {
    "deletedRecords": 2,
    "deletedFiles": 2,
    "failedFiles": 0,
    "deletedEpisodes": [
      {"id": 1, "name": "Old Episode 1"},
      {"id": 2, "name": "Old Episode 2"}
    ]
  }
}

Generates an RSS podcast feed from all episodes. Requires authorization.

Query Parameters:

• key (string): Must match the AUTH_TOKEN environment variable

Example:

https://your-val-url.web.val.run/podcast.xml?key=your-auth-token-here

The tts_episodes table contains:

• id (INTEGER PRIMARY KEY)
• name (TEXT) - Episode title
• text (TEXT) - Original text content
• mp3_url (TEXT) - URL to the MP3 file
• date_added (DATETIME) - Timestamp when added

• ✅ Text file upload and conversion to MP3
• ✅ SQLite database storage
• ✅ MP3 file serving from blob storage
• ✅ RSS podcast feed generation
• ✅ Authorization protection for content creation and feed access
• ✅ Error handling and validation
• ✅ Automatic database initialization
• ✅ Web test interface

A weekly cron job (/cleanup.ts) automatically deletes:

• Database records older than 30 days
• Associated MP3 files from blob storage
• Runs every week to keep storage usage manageable

• ✅ Deletes episodes older than 30 days
• ✅ Removes both database records and MP3 files
• ✅ Detailed logging of cleanup operations
• ✅ Error handling and reporting
• ✅ Runs weekly (schedule configurable in Val Town UI)

• Use POST /cleanup endpoint for immediate cleanup
• Configurable age threshold (1-365 days)
• Same authorization as content creation
• Detailed cleanup reports

• POST /convert: Requires Authorization: Bearer YOUR_TOKEN header
• GET /podcast.xml: Requires ?key=YOUR_TOKEN query parameter
• Both use the same AUTH_TOKEN environment variable for validation
• All other endpoints (episodes list, audio serving, test page) are public

• Uses OpenAI's tts-1 model with the alloy voice
• MP3 files are stored in Val Town's blob storage
• Database records include full text content for search/reference
• Podcast feed is automatically generated from database records
• All endpoints include proper error handling