pdf.js-extract/lib/index.js

super-simple async PDF reader that extracts text with x,y page positions based on pdf.js

const fs = require("fs");
const path = require("path");
const utils = require("./utils.js");
const LocalCMapReaderFactory = require("./cmap-reader.js");
// HACK few hacks to let PDF.js be loaded not as a module in global space.
if (typeof window === 'undefined' || typeof process === 'object') {
	require("./pdfjs/domstubs.js").setStubs(global);
}
const pdfjsLib = require("./pdfjs/pdf.js");

class PDFExtract {

	constructor() {
	}

	extract(filename, options, cb) {
		if (!cb) {
			return new Promise((resolve, reject) => {
				this.extract(filename, options, (err, data) => {
					if (err) {
						reject(err);
					} else {
						resolve(data);
					}
				})
			});
		}
		fs.readFile(filename, (err, buffer) => {
			if (err) {
				return cb(err);
			}
			return this.extractBuffer(buffer, options, (err, pdf) => {
				if (err) {
					cb(err);
				} else {
					pdf.filename = filename;
					cb(null, pdf);
				}
			});
		});
	}

	extractBuffer(buffer, options = {}, cb) {
		if (!cb) {
			return new Promise((resolve, reject) => {
				this.extractBuffer(buffer, options, (err, data) => {
					if (err) {
						reject(err);
					} else {
						resolve(data);
					}
				})
			});
		}
		// Loading file from file system into typed array
		if (options.verbosity === undefined) {
			// get rid of all warnings in nodejs usage
			options.verbosity = -1;
		}
		if (options.cMapUrl === undefined) {
			options.cMapUrl = path.join(__dirname, "./cmaps/"); // trailing path delimiter is important
		}
		if (options.cMapPacked === undefined) {
			options.cMapPacked = true;
		}
		if (options.CMapReaderFactory === undefined) {
			options.CMapReaderFactory = LocalCMapReaderFactory;
		}
		options.data = new Uint8Array(buffer);
		const pdf = {
			meta: {},
			pages: []
		};
		// Will be using promises to load document, pages and misc data instead of callback.
		pdfjsLib.getDocument(options).promise.then(doc => {
			const firstPage = (options && options.firstPage) ? options.firstPage : 1;
			const lastPage = Math.min((options && options.lastPage) ? options.lastPage : doc.numPages, doc.numPages);
			pdf.pdfInfo = doc.pdfInfo;
			const promises = [
				doc.getMetadata().then(data => {
					pdf.meta = {info: data.info, metadata: data.metadata ? data.metadata.getAll() || null : null};
				})
			];
			const loadPage = pageNum => doc.getPage(pageNum).then(page => {
				const viewport = page.getViewport({scale: 1.0});
				const pag = {
					pageInfo: {
						num: pageNum,
						scale: viewport.scale,
						rotation: viewport.rotation,
						offsetX: viewport.offsetX,
						offsetY: viewport.offsetY,
						width: viewport.width,
						height: viewport.height
					}
				};
				pdf.pages.push(pag);
				const normalizeWhitespace = !!(options && options.normalizeWhitespace === true);
				const disableCombineTextItems = !!(options && options.disableCombineTextItems === true);
				return Promise.all([
					page.getAnnotations().then((annotations) => {
						pag.links = annotations.filter((annot) => annot.subtype === "Link" && !!annot.url)
						.map((link) => link.url);
					}),
					page.getTextContent({normalizeWhitespace, disableCombineTextItems}).then((content) => {
						// Content contains lots of information about the text layout and styles, but we need only strings at the moment
						pag.content = content.items.map(item => {
							const tm = item.transform;
							let x = tm[4];
							let y = pag.pageInfo.height - tm[5];
							if (viewport.rotation === 90) {
								x = tm[5];
								y = tm[4];
							}
							// see https://github.com/mozilla/pdf.js/issues/8276
							const height = Math.sqrt(tm[2] * tm[2] + tm[3] * tm[3]);
							return {
								x: x,
								y: y,
								str: item.str,
								dir: item.dir,
								width: item.width,
								height: height,
								fontName: item.fontName
							};
						});
					})
				]).then(() => {
					// console.log("done page content parsing");
				}, (err) => {
					cb(err);
				});
			});
			for (let i = firstPage; i <= lastPage; i++) {
				promises.push(loadPage(i));
			}
			return Promise.all(promises);
		}).then(() => {
			pdf.pages.sort((a, b) => a.pageInfo.num - b.pageInfo.num);
			cb(null, pdf);
		}, (err) => {
			cb(err)
		});
	}
}

PDFExtract.utils = utils;

module.exports.PDFExtract = PDFExtract;