ChunkingStrategyBase

Base Classes · AiDotNet.RetrievalAugmentedGeneration.ChunkingStrategies

Provides a base implementation for text chunking strategies with common functionality.

Implements: IChunkingStrategy

Known derived types: AgenticChunker, CodeAwareTextSplitter, FixedSizeChunkingStrategy, HeaderBasedTextSplitter, MarkdownTextSplitter, MultiModalTextSplitter, RecursiveCharacterChunkingStrategy, RecursiveCharacterTextSplitter, SemanticChunkingStrategy, SentenceChunkingStrategy, SlidingWindowChunkingStrategy, TableAwareTextSplitter, TokenBasedChunkingStrategy

For Beginners

This is the foundation that all text splitting methods build upon.

Think of it like a template for dividing text:

  • It handles common tasks (checking inputs, managing overlap)
  • Specific chunking methods (fixed-size, sentence-based) just fill in how they split text
  • This ensures all chunking strategies work consistently

How It Works

This abstract class implements the IChunkingStrategy interface and provides common functionality for text splitting strategies. It handles validation and provides utility methods for chunk overlap while allowing derived classes to focus on implementing the core chunking algorithm.

Properties

Property Summary
int ChunkOverlap { get; } Gets the number of characters that should overlap between consecutive chunks.
int ChunkSize { get; } Gets the target size for each chunk in characters.

Methods

Method Summary
IEnumerable<ValueTuple<string, int, int>> ChunkWithPositions(string text) Splits a text string into chunks and returns them with position metadata.
IEnumerable<string> Chunk(string text) Splits a text string into chunks according to the strategy’s rules.