wordmap/dist/index/StaticIndex.d.ts

Multi-Lingual Word Alignment Prediction

import { Token } from "wordmap-lexer";
import Ngram from "../structures/Ngram";
import NgramIndex from "./NgramIndex";
/**
 * A collection of indexes on the static content.
 * TODO: maybe I should split this into sentences as well.
 * e.g. a source SentenceIndex and a target SentenceIndex
 * Then we could reuse it in other places such as word-mt.
 */
export default class StaticIndex {
    private srcNgramFreqIndex;
    private srcTokenLength;
    private srcCharLength;
    private tgtNgramFreqIndex;
    private tgtTokenLength;
    private tgtCharLength;
    /**
     * Returns an index of source n-gram frequencies in the corpus
     * @return {NgramIndex}
     */
    readonly sourceNgramFrequency: NgramIndex;
    /**
     * Returns an index of target n-gram frequencies in the corpus
     * @return {NgramIndex}
     */
    readonly targetNgramFrequency: NgramIndex;
    /**
     * Returns the {@link Token} length of the entire source
     * @return {number}
     */
    readonly sourceTokenLength: number;
    /**
     * Returns the {@link Token} length of the entire target
     * @return {number}
     */
    readonly targetTokenLength: number;
    /**
     * Returns the character length of the entire source
     * @return {number}
     */
    readonly sourceCharacterLength: number;
    /**
     * Returns the character length of the entire target
     * @return {number}
     */
    readonly targetCharLength: number;
    constructor();
    /**
     * Adds a sentence to the index.
     * The tokens in these n-grams must be measured for accurate positional metrics.
     * The n-grams are passed as arguments instead of being generated internally to reduce
     * duplicating work.
     *
     * @param sourceTokens - the source sentence tokens
     * @param targetTokens - the target sentence tokens
     * @param sourceNgrams - the source sentence n-grams
     * @param targetNgrams - the target sentence n-grams
     */
    addSentence(sourceTokens: Token[], targetTokens: Token[], sourceNgrams: Ngram[], targetNgrams: Ngram[]): void;
}