llm-text-splitter/dist/index.d.ts

A super simple text splitter for RAG applications

/**
 * Enum representing the available splitter types.
 */
export type SplitterType = 'sentence' | 'paragraph' | 'markdown';
/**
 * Interface defining the options for the `splitter` function.
 */
export interface SplitOptions {
    /**
     * Minimum length of a chunk.
     * @default 0
     */
    minLength?: number;
    /**
     * Maximum length of a chunk.
     * @default 5000
     */
    maxLength?: number;
    /**
     * Number of characters to overlap between chunks.
     * @default 0
     */
    overlap?: number;
    /**
     * The type of splitter to use. Can be 'sentence', 'paragraph', or 'markdown'.
     * @default 'sentence'
     */
    splitter?: SplitterType;
    /**
     * Custom regular expression to use for splitting. If provided, `splitter` will be ignored.
     */
    regex?: RegExp | string;
    /**
     * Whether to remove extra spaces from the chunks.
     * @default false
     */
    removeExtraSpaces?: boolean;
}
export declare class Splitter {
    private options;
    private static readonly REGEX;
    constructor(options?: SplitOptions);
    private findBreakPoint;
    private getOverlapText;
    private splitChunk;
    private handleChunkSize;
    private getRegExp;
    /**
     * Splits a given text into chunks based on the options provided in the constructor.
     * @param text The text to split.
     * @returns An array of strings, where each string is a chunk of the original text.
     */
    split(text: string, options?: SplitOptions): string[];
}