LottaDB.Tiki

LottaDB.Tiki

Rich blob metadata extraction for LottaDB, powered by Tiki.Net.

When you upload a blob, Tiki.Net automatically parses it and stores strongly-typed metadata (EXIF from photos, ID3 from music, page counts from PDFs, etc.) as queryable entities in LottaDB.

Installation

dotnet add package LottaDB.Tiki

Quick Start

using Lotta;
using Lotta.Tiki;

var catalog = new LottaCatalog("myapp", connectionString);
var db = await catalog.GetDatabaseAsync("media", config =>
{
    config.UseTikiExtraction(); // one line to enable
});

// Upload a photo — metadata is extracted automatically
await using var stream = File.OpenRead("vacation.jpg");
var photo = (BlobPhoto) await db.UploadBlobAsync("photos/vacation.jpg", stream);

// Rich metadata is available immediately
Console.WriteLine($"Camera: {photo.CameraModel}");
Console.WriteLine($"ISO: {photo.IsoSpeed}");
Console.WriteLine($"Size: {photo.Width}x{photo.Height}");

How It Works

UseTikiExtraction() registers an OnUpload handler that:

1. Receives the blob content as a concurrent stream (zero-copy via TeeStream)
2. Parses it with Tiki.Net's auto-detecting parser
3. Maps the result to the correct BlobFile subtype
4. Saves the metadata entity — queryable and full-text searchable

The blob upload and parsing happen concurrently — no buffering the entire file in memory.

Querying Metadata

All metadata properties are queryable via LINQ and Lucene:

// Find high-ISO photos
var nightPhotos = db.Search<BlobPhoto>(p => p.IsoSpeed > 3200).ToList();

// Find large PDFs
var bigDocs = db.Search<BlobDocument>(p => p.PageCount > 50).ToList();

// Find music by artist
var songs = db.Search<BlobMusic>(m => m.Artist == "Pink Floyd").ToList();

// Full-text search across extracted document content
var results = db.Search<BlobFile>("quarterly revenue").ToList();

// Polymorphic — returns all file types
var allFiles = db.Search<BlobFile>().ToList();

Supported File Types

Text content is automatically extracted and indexed for full-text search.

Configuration

// Default — uses auto-configured TikiEngine
config.UseTikiExtraction();

// Custom engine with specific parsers
var engine = new TikiEngine(TikiConfig.CreateBuilder()
    .AddParser(new PdfParser())
    .AddParser(new ImageParser())
    .Build());
config.UseTikiExtraction(engine);

BlobFile Convenience Methods

var file = await db.GetAsync<BlobPhoto>("photos/vacation.jpg");

// Download the blob content
Stream? stream = await file.DownloadAsync();

// Delete blob + metadata in one call
await file.DeleteAsync();

Without Tiki (Default Handler)

LottaDB includes a built-in default handler that doesn't require Tiki.Net:

config.OnUpload(); // uses file extension for type detection

This gives you:

• File extension to MIME type mapping
• Correct BlobFile subtype (BlobPhoto for .jpg, BlobDocument for .pdf, etc.)
• Text content extraction for known text formats (.txt, .md, .json, .cs, .py, etc.)
• Basic metadata (path, name, folder, media type, content length)

The Tiki handler adds deep content parsing (EXIF, ID3 tags, page counts, etc.) on top.