office-text-extractor-browser/build/lib.js

Fork of office-text-extractor with unreleased changes that include browser support

// source/lib.ts
// The source code for the library.
import { Buffer } from 'buffer/index.js';
import { fileTypeFromBuffer as getFileType } from 'file-type';
import { fetchUrl } from './util.js';
/**
 * The text extractor class.
 */
export class TextExtractor {
    constructor() {
        // The list of methods supported by this instance of the extractor.
        this.methods = [];
        /**
         * Registers a new method to this instance of the extractor.
         *
         * @param method The method of text extraction to add.
         * @returns The current instance, for method chaining.
         */
        this.addMethod = (method) => {
            this.methods.push(method);
            return this;
        };
        /**
         * Extracts text from the given input.
         *
         * @param payload The input and type of input to extract text from.
         * @returns The extracted text as a simple string.
         */
        this.extractText = async ({ input, type }) => {
            // Turn the input into a buffer containing the file's contents.
            let preparedInput;
            if (typeof input === 'string') {
                // if (type === 'file') preparedInput = await readFile(input)
                if (type === 'url')
                    preparedInput = await fetchUrl(input);
                else
                    preparedInput = Buffer.from(input);
            }
            else {
                preparedInput = input;
            }
            // Check the mime type of the file. If there is no mime type, it's most
            // likely a txt/csv files.
            const mimeDetails = await getFileType(preparedInput);
            if (!mimeDetails)
                return preparedInput.toString();
            // Find the extractor that can handle that mime type, and call it.
            const extractor = this.methods.find((method) => method.mimes.includes(mimeDetails.mime));
            if (!extractor?.apply) {
                const message = `text-extractor: could not find a method to handle ${mimeDetails.mime}`;
                throw new Error(message);
            }
            return extractor.apply(preparedInput);
        };
    }
}