office-text-extractor-browser/source/parsers/doc.ts

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

// source/parsers/docx.ts
// The text extracter for DOCX files.

import { type Buffer } from 'buffer/'
import { extractRawText as parseWordFile } from 'mammoth'

import type { TextExtractionMethod } from '../lib.js'

export class DocExtractor implements TextExtractionMethod {
	/**
	 * The type(s) of input acceptable to this method.
	 */
	mimes = [
		'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
	]

	/**
	 * Extract text from a DOCX file if possible.
	 *
	 * @param payload The input and its type.
	 * @returns The text extracted from the input.
	 */
	apply = async (input: Buffer): Promise<string> => {
		// Convert the DOCX to text and return the text.
		// @ts-expect-error: see feross/buffer#353, the types are incomplete.
		const parsedDocx = await parseWordFile({ buffer: input })
		return parsedDocx.value
	}
}