worker.js 21 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638
  1. /**
  2. * @licstart The following is the entire license notice for the
  3. * Javascript code in this page
  4. *
  5. * Copyright 2019 Mozilla Foundation
  6. *
  7. * Licensed under the Apache License, Version 2.0 (the "License");
  8. * you may not use this file except in compliance with the License.
  9. * You may obtain a copy of the License at
  10. *
  11. * http://www.apache.org/licenses/LICENSE-2.0
  12. *
  13. * Unless required by applicable law or agreed to in writing, software
  14. * distributed under the License is distributed on an "AS IS" BASIS,
  15. * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  16. * See the License for the specific language governing permissions and
  17. * limitations under the License.
  18. *
  19. * @licend The above is the entire license notice for the
  20. * Javascript code in this page
  21. */
  22. "use strict";
  23. Object.defineProperty(exports, "__esModule", {
  24. value: true
  25. });
  26. exports.WorkerMessageHandler = exports.WorkerTask = void 0;
  27. var _regenerator = _interopRequireDefault(require("@babel/runtime/regenerator"));
  28. var _util = require("../shared/util");
  29. var _primitives = require("./primitives");
  30. var _pdf_manager = require("./pdf_manager");
  31. var _is_node = _interopRequireDefault(require("../shared/is_node"));
  32. var _message_handler = require("../shared/message_handler");
  33. var _worker_stream = require("./worker_stream");
  34. var _core_utils = require("./core_utils");
  35. function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { "default": obj }; }
  36. function _typeof(obj) { if (typeof Symbol === "function" && typeof Symbol.iterator === "symbol") { _typeof = function _typeof(obj) { return typeof obj; }; } else { _typeof = function _typeof(obj) { return obj && typeof Symbol === "function" && obj.constructor === Symbol && obj !== Symbol.prototype ? "symbol" : typeof obj; }; } return _typeof(obj); }
  37. function _slicedToArray(arr, i) { return _arrayWithHoles(arr) || _iterableToArrayLimit(arr, i) || _nonIterableRest(); }
  38. function _nonIterableRest() { throw new TypeError("Invalid attempt to destructure non-iterable instance"); }
  39. function _iterableToArrayLimit(arr, i) { var _arr = []; var _n = true; var _d = false; var _e = undefined; try { for (var _i = arr[Symbol.iterator](), _s; !(_n = (_s = _i.next()).done); _n = true) { _arr.push(_s.value); if (i && _arr.length === i) break; } } catch (err) { _d = true; _e = err; } finally { try { if (!_n && _i["return"] != null) _i["return"](); } finally { if (_d) throw _e; } } return _arr; }
  40. function _arrayWithHoles(arr) { if (Array.isArray(arr)) return arr; }
  41. function asyncGeneratorStep(gen, resolve, reject, _next, _throw, key, arg) { try { var info = gen[key](arg); var value = info.value; } catch (error) { reject(error); return; } if (info.done) { resolve(value); } else { Promise.resolve(value).then(_next, _throw); } }
  42. function _asyncToGenerator(fn) { return function () { var self = this, args = arguments; return new Promise(function (resolve, reject) { var gen = fn.apply(self, args); function _next(value) { asyncGeneratorStep(gen, resolve, reject, _next, _throw, "next", value); } function _throw(err) { asyncGeneratorStep(gen, resolve, reject, _next, _throw, "throw", err); } _next(undefined); }); }; }
  43. var WorkerTask = function WorkerTaskClosure() {
  44. function WorkerTask(name) {
  45. this.name = name;
  46. this.terminated = false;
  47. this._capability = (0, _util.createPromiseCapability)();
  48. }
  49. WorkerTask.prototype = {
  50. get finished() {
  51. return this._capability.promise;
  52. },
  53. finish: function finish() {
  54. this._capability.resolve();
  55. },
  56. terminate: function terminate() {
  57. this.terminated = true;
  58. },
  59. ensureNotTerminated: function ensureNotTerminated() {
  60. if (this.terminated) {
  61. throw new Error('Worker task was terminated');
  62. }
  63. }
  64. };
  65. return WorkerTask;
  66. }();
  67. exports.WorkerTask = WorkerTask;
  68. var WorkerMessageHandler = {
  69. setup: function setup(handler, port) {
  70. var testMessageProcessed = false;
  71. handler.on('test', function wphSetupTest(data) {
  72. if (testMessageProcessed) {
  73. return;
  74. }
  75. testMessageProcessed = true;
  76. if (!(data instanceof Uint8Array)) {
  77. handler.send('test', false);
  78. return;
  79. }
  80. var supportTransfers = data[0] === 255;
  81. handler.postMessageTransfers = supportTransfers;
  82. var xhr = new XMLHttpRequest();
  83. var responseExists = 'response' in xhr;
  84. try {
  85. xhr.responseType;
  86. } catch (e) {
  87. responseExists = false;
  88. }
  89. if (!responseExists) {
  90. handler.send('test', false);
  91. return;
  92. }
  93. handler.send('test', {
  94. supportTypedArray: true,
  95. supportTransfers: supportTransfers
  96. });
  97. });
  98. handler.on('configure', function wphConfigure(data) {
  99. (0, _util.setVerbosityLevel)(data.verbosity);
  100. });
  101. handler.on('GetDocRequest', function wphSetupDoc(data) {
  102. return WorkerMessageHandler.createDocumentHandler(data, port);
  103. });
  104. },
  105. createDocumentHandler: function createDocumentHandler(docParams, port) {
  106. var pdfManager;
  107. var terminated = false;
  108. var cancelXHRs = null;
  109. var WorkerTasks = [];
  110. var verbosity = (0, _util.getVerbosityLevel)();
  111. var apiVersion = docParams.apiVersion;
  112. var workerVersion = '2.2.228';
  113. if (apiVersion !== workerVersion) {
  114. throw new Error("The API version \"".concat(apiVersion, "\" does not match ") + "the Worker version \"".concat(workerVersion, "\"."));
  115. }
  116. var docId = docParams.docId;
  117. var docBaseUrl = docParams.docBaseUrl;
  118. var workerHandlerName = docParams.docId + '_worker';
  119. var handler = new _message_handler.MessageHandler(workerHandlerName, docId, port);
  120. handler.postMessageTransfers = docParams.postMessageTransfers;
  121. function ensureNotTerminated() {
  122. if (terminated) {
  123. throw new Error('Worker was terminated');
  124. }
  125. }
  126. function startWorkerTask(task) {
  127. WorkerTasks.push(task);
  128. }
  129. function finishWorkerTask(task) {
  130. task.finish();
  131. var i = WorkerTasks.indexOf(task);
  132. WorkerTasks.splice(i, 1);
  133. }
  134. function loadDocument(_x) {
  135. return _loadDocument.apply(this, arguments);
  136. }
  137. function _loadDocument() {
  138. _loadDocument = _asyncToGenerator(
  139. /*#__PURE__*/
  140. _regenerator["default"].mark(function _callee(recoveryMode) {
  141. var _ref4, _ref5, numPages, fingerprint;
  142. return _regenerator["default"].wrap(function _callee$(_context) {
  143. while (1) {
  144. switch (_context.prev = _context.next) {
  145. case 0:
  146. _context.next = 2;
  147. return pdfManager.ensureDoc('checkHeader');
  148. case 2:
  149. _context.next = 4;
  150. return pdfManager.ensureDoc('parseStartXRef');
  151. case 4:
  152. _context.next = 6;
  153. return pdfManager.ensureDoc('parse', [recoveryMode]);
  154. case 6:
  155. if (recoveryMode) {
  156. _context.next = 9;
  157. break;
  158. }
  159. _context.next = 9;
  160. return pdfManager.ensureDoc('checkFirstPage');
  161. case 9:
  162. _context.next = 11;
  163. return Promise.all([pdfManager.ensureDoc('numPages'), pdfManager.ensureDoc('fingerprint')]);
  164. case 11:
  165. _ref4 = _context.sent;
  166. _ref5 = _slicedToArray(_ref4, 2);
  167. numPages = _ref5[0];
  168. fingerprint = _ref5[1];
  169. return _context.abrupt("return", {
  170. numPages: numPages,
  171. fingerprint: fingerprint
  172. });
  173. case 16:
  174. case "end":
  175. return _context.stop();
  176. }
  177. }
  178. }, _callee);
  179. }));
  180. return _loadDocument.apply(this, arguments);
  181. }
  182. function getPdfManager(data, evaluatorOptions) {
  183. var pdfManagerCapability = (0, _util.createPromiseCapability)();
  184. var pdfManager;
  185. var source = data.source;
  186. if (source.data) {
  187. try {
  188. pdfManager = new _pdf_manager.LocalPdfManager(docId, source.data, source.password, evaluatorOptions, docBaseUrl);
  189. pdfManagerCapability.resolve(pdfManager);
  190. } catch (ex) {
  191. pdfManagerCapability.reject(ex);
  192. }
  193. return pdfManagerCapability.promise;
  194. }
  195. var pdfStream,
  196. cachedChunks = [];
  197. try {
  198. pdfStream = new _worker_stream.PDFWorkerStream(handler);
  199. } catch (ex) {
  200. pdfManagerCapability.reject(ex);
  201. return pdfManagerCapability.promise;
  202. }
  203. var fullRequest = pdfStream.getFullReader();
  204. fullRequest.headersReady.then(function () {
  205. if (!fullRequest.isRangeSupported) {
  206. return;
  207. }
  208. var disableAutoFetch = source.disableAutoFetch || fullRequest.isStreamingSupported;
  209. pdfManager = new _pdf_manager.NetworkPdfManager(docId, pdfStream, {
  210. msgHandler: handler,
  211. password: source.password,
  212. length: fullRequest.contentLength,
  213. disableAutoFetch: disableAutoFetch,
  214. rangeChunkSize: source.rangeChunkSize
  215. }, evaluatorOptions, docBaseUrl);
  216. for (var i = 0; i < cachedChunks.length; i++) {
  217. pdfManager.sendProgressiveData(cachedChunks[i]);
  218. }
  219. cachedChunks = [];
  220. pdfManagerCapability.resolve(pdfManager);
  221. cancelXHRs = null;
  222. })["catch"](function (reason) {
  223. pdfManagerCapability.reject(reason);
  224. cancelXHRs = null;
  225. });
  226. var loaded = 0;
  227. var flushChunks = function flushChunks() {
  228. var pdfFile = (0, _util.arraysToBytes)(cachedChunks);
  229. if (source.length && pdfFile.length !== source.length) {
  230. (0, _util.warn)('reported HTTP length is different from actual');
  231. }
  232. try {
  233. pdfManager = new _pdf_manager.LocalPdfManager(docId, pdfFile, source.password, evaluatorOptions, docBaseUrl);
  234. pdfManagerCapability.resolve(pdfManager);
  235. } catch (ex) {
  236. pdfManagerCapability.reject(ex);
  237. }
  238. cachedChunks = [];
  239. };
  240. var readPromise = new Promise(function (resolve, reject) {
  241. var readChunk = function readChunk(chunk) {
  242. try {
  243. ensureNotTerminated();
  244. if (chunk.done) {
  245. if (!pdfManager) {
  246. flushChunks();
  247. }
  248. cancelXHRs = null;
  249. return;
  250. }
  251. var data = chunk.value;
  252. loaded += (0, _util.arrayByteLength)(data);
  253. if (!fullRequest.isStreamingSupported) {
  254. handler.send('DocProgress', {
  255. loaded: loaded,
  256. total: Math.max(loaded, fullRequest.contentLength || 0)
  257. });
  258. }
  259. if (pdfManager) {
  260. pdfManager.sendProgressiveData(data);
  261. } else {
  262. cachedChunks.push(data);
  263. }
  264. fullRequest.read().then(readChunk, reject);
  265. } catch (e) {
  266. reject(e);
  267. }
  268. };
  269. fullRequest.read().then(readChunk, reject);
  270. });
  271. readPromise["catch"](function (e) {
  272. pdfManagerCapability.reject(e);
  273. cancelXHRs = null;
  274. });
  275. cancelXHRs = function cancelXHRs() {
  276. pdfStream.cancelAllRequests('abort');
  277. };
  278. return pdfManagerCapability.promise;
  279. }
  280. function setupDoc(data) {
  281. function onSuccess(doc) {
  282. ensureNotTerminated();
  283. handler.send('GetDoc', {
  284. pdfInfo: doc
  285. });
  286. }
  287. function onFailure(e) {
  288. ensureNotTerminated();
  289. if (e instanceof _util.PasswordException) {
  290. var task = new WorkerTask('PasswordException: response ' + e.code);
  291. startWorkerTask(task);
  292. handler.sendWithPromise('PasswordRequest', e).then(function (data) {
  293. finishWorkerTask(task);
  294. pdfManager.updatePassword(data.password);
  295. pdfManagerReady();
  296. })["catch"](function (boundException) {
  297. finishWorkerTask(task);
  298. handler.send('PasswordException', boundException);
  299. }.bind(null, e));
  300. } else if (e instanceof _util.InvalidPDFException) {
  301. handler.send('InvalidPDF', e);
  302. } else if (e instanceof _util.MissingPDFException) {
  303. handler.send('MissingPDF', e);
  304. } else if (e instanceof _util.UnexpectedResponseException) {
  305. handler.send('UnexpectedResponse', e);
  306. } else {
  307. handler.send('UnknownError', new _util.UnknownErrorException(e.message, e.toString()));
  308. }
  309. }
  310. function pdfManagerReady() {
  311. ensureNotTerminated();
  312. loadDocument(false).then(onSuccess, function loadFailure(ex) {
  313. ensureNotTerminated();
  314. if (!(ex instanceof _core_utils.XRefParseException)) {
  315. onFailure(ex);
  316. return;
  317. }
  318. pdfManager.requestLoadedStream();
  319. pdfManager.onLoadedStream().then(function () {
  320. ensureNotTerminated();
  321. loadDocument(true).then(onSuccess, onFailure);
  322. });
  323. }, onFailure);
  324. }
  325. ensureNotTerminated();
  326. var evaluatorOptions = {
  327. forceDataSchema: data.disableCreateObjectURL,
  328. maxImageSize: data.maxImageSize,
  329. disableFontFace: data.disableFontFace,
  330. nativeImageDecoderSupport: data.nativeImageDecoderSupport,
  331. ignoreErrors: data.ignoreErrors,
  332. isEvalSupported: data.isEvalSupported
  333. };
  334. getPdfManager(data, evaluatorOptions).then(function (newPdfManager) {
  335. if (terminated) {
  336. newPdfManager.terminate();
  337. throw new Error('Worker was terminated');
  338. }
  339. pdfManager = newPdfManager;
  340. pdfManager.onLoadedStream().then(function (stream) {
  341. handler.send('DataLoaded', {
  342. length: stream.bytes.byteLength
  343. });
  344. });
  345. }).then(pdfManagerReady, onFailure);
  346. }
  347. handler.on('GetPage', function wphSetupGetPage(data) {
  348. return pdfManager.getPage(data.pageIndex).then(function (page) {
  349. return Promise.all([pdfManager.ensure(page, 'rotate'), pdfManager.ensure(page, 'ref'), pdfManager.ensure(page, 'userUnit'), pdfManager.ensure(page, 'view')]).then(function (_ref) {
  350. var _ref2 = _slicedToArray(_ref, 4),
  351. rotate = _ref2[0],
  352. ref = _ref2[1],
  353. userUnit = _ref2[2],
  354. view = _ref2[3];
  355. return {
  356. rotate: rotate,
  357. ref: ref,
  358. userUnit: userUnit,
  359. view: view
  360. };
  361. });
  362. });
  363. });
  364. handler.on('GetPageIndex', function wphSetupGetPageIndex(data) {
  365. var ref = _primitives.Ref.get(data.ref.num, data.ref.gen);
  366. var catalog = pdfManager.pdfDocument.catalog;
  367. return catalog.getPageIndex(ref);
  368. });
  369. handler.on('GetDestinations', function wphSetupGetDestinations(data) {
  370. return pdfManager.ensureCatalog('destinations');
  371. });
  372. handler.on('GetDestination', function wphSetupGetDestination(data) {
  373. return pdfManager.ensureCatalog('getDestination', [data.id]);
  374. });
  375. handler.on('GetPageLabels', function wphSetupGetPageLabels(data) {
  376. return pdfManager.ensureCatalog('pageLabels');
  377. });
  378. handler.on('GetPageLayout', function wphSetupGetPageLayout(data) {
  379. return pdfManager.ensureCatalog('pageLayout');
  380. });
  381. handler.on('GetPageMode', function wphSetupGetPageMode(data) {
  382. return pdfManager.ensureCatalog('pageMode');
  383. });
  384. handler.on('GetViewerPreferences', function (data) {
  385. return pdfManager.ensureCatalog('viewerPreferences');
  386. });
  387. handler.on('GetOpenActionDestination', function (data) {
  388. return pdfManager.ensureCatalog('openActionDestination');
  389. });
  390. handler.on('GetAttachments', function wphSetupGetAttachments(data) {
  391. return pdfManager.ensureCatalog('attachments');
  392. });
  393. handler.on('GetJavaScript', function wphSetupGetJavaScript(data) {
  394. return pdfManager.ensureCatalog('javaScript');
  395. });
  396. handler.on('GetOutline', function wphSetupGetOutline(data) {
  397. return pdfManager.ensureCatalog('documentOutline');
  398. });
  399. handler.on('GetPermissions', function (data) {
  400. return pdfManager.ensureCatalog('permissions');
  401. });
  402. handler.on('GetMetadata', function wphSetupGetMetadata(data) {
  403. return Promise.all([pdfManager.ensureDoc('documentInfo'), pdfManager.ensureCatalog('metadata')]);
  404. });
  405. handler.on('GetData', function wphSetupGetData(data) {
  406. pdfManager.requestLoadedStream();
  407. return pdfManager.onLoadedStream().then(function (stream) {
  408. return stream.bytes;
  409. });
  410. });
  411. handler.on('GetStats', function wphSetupGetStats(data) {
  412. return pdfManager.pdfDocument.xref.stats;
  413. });
  414. handler.on('GetAnnotations', function (_ref3) {
  415. var pageIndex = _ref3.pageIndex,
  416. intent = _ref3.intent;
  417. return pdfManager.getPage(pageIndex).then(function (page) {
  418. return page.getAnnotationsData(intent);
  419. });
  420. });
  421. handler.on('RenderPageRequest', function wphSetupRenderPage(data) {
  422. var pageIndex = data.pageIndex;
  423. pdfManager.getPage(pageIndex).then(function (page) {
  424. var task = new WorkerTask('RenderPageRequest: page ' + pageIndex);
  425. startWorkerTask(task);
  426. var start = verbosity >= _util.VerbosityLevel.INFOS ? Date.now() : 0;
  427. page.getOperatorList({
  428. handler: handler,
  429. task: task,
  430. intent: data.intent,
  431. renderInteractiveForms: data.renderInteractiveForms
  432. }).then(function (operatorList) {
  433. finishWorkerTask(task);
  434. if (start) {
  435. (0, _util.info)("page=".concat(pageIndex + 1, " - getOperatorList: time=") + "".concat(Date.now() - start, "ms, len=").concat(operatorList.totalLength));
  436. }
  437. }, function (e) {
  438. finishWorkerTask(task);
  439. if (task.terminated) {
  440. return;
  441. }
  442. handler.send('UnsupportedFeature', {
  443. featureId: _util.UNSUPPORTED_FEATURES.unknown
  444. });
  445. var minimumStackMessage = 'worker.js: while trying to getPage() and getOperatorList()';
  446. var wrappedException;
  447. if (typeof e === 'string') {
  448. wrappedException = {
  449. message: e,
  450. stack: minimumStackMessage
  451. };
  452. } else if (_typeof(e) === 'object') {
  453. wrappedException = {
  454. message: e.message || e.toString(),
  455. stack: e.stack || minimumStackMessage
  456. };
  457. } else {
  458. wrappedException = {
  459. message: 'Unknown exception type: ' + _typeof(e),
  460. stack: minimumStackMessage
  461. };
  462. }
  463. handler.send('PageError', {
  464. pageIndex: pageIndex,
  465. error: wrappedException,
  466. intent: data.intent
  467. });
  468. });
  469. });
  470. }, this);
  471. handler.on('GetTextContent', function wphExtractText(data, sink) {
  472. var pageIndex = data.pageIndex;
  473. sink.onPull = function (desiredSize) {};
  474. sink.onCancel = function (reason) {};
  475. pdfManager.getPage(pageIndex).then(function (page) {
  476. var task = new WorkerTask('GetTextContent: page ' + pageIndex);
  477. startWorkerTask(task);
  478. var start = verbosity >= _util.VerbosityLevel.INFOS ? Date.now() : 0;
  479. page.extractTextContent({
  480. handler: handler,
  481. task: task,
  482. sink: sink,
  483. normalizeWhitespace: data.normalizeWhitespace,
  484. combineTextItems: data.combineTextItems
  485. }).then(function () {
  486. finishWorkerTask(task);
  487. if (start) {
  488. (0, _util.info)("page=".concat(pageIndex + 1, " - getTextContent: time=") + "".concat(Date.now() - start, "ms"));
  489. }
  490. sink.close();
  491. }, function (reason) {
  492. finishWorkerTask(task);
  493. if (task.terminated) {
  494. return;
  495. }
  496. sink.error(reason);
  497. throw reason;
  498. });
  499. });
  500. });
  501. handler.on('FontFallback', function (data) {
  502. return pdfManager.fontFallback(data.id, handler);
  503. });
  504. handler.on('Cleanup', function wphCleanup(data) {
  505. return pdfManager.cleanup();
  506. });
  507. handler.on('Terminate', function wphTerminate(data) {
  508. terminated = true;
  509. if (pdfManager) {
  510. pdfManager.terminate();
  511. pdfManager = null;
  512. }
  513. if (cancelXHRs) {
  514. cancelXHRs();
  515. }
  516. (0, _primitives.clearPrimitiveCaches)();
  517. var waitOn = [];
  518. WorkerTasks.forEach(function (task) {
  519. waitOn.push(task.finished);
  520. task.terminate();
  521. });
  522. return Promise.all(waitOn).then(function () {
  523. handler.destroy();
  524. handler = null;
  525. });
  526. });
  527. handler.on('Ready', function wphReady(data) {
  528. setupDoc(docParams);
  529. docParams = null;
  530. });
  531. return workerHandlerName;
  532. },
  533. initializeFromPort: function initializeFromPort(port) {
  534. var handler = new _message_handler.MessageHandler('worker', 'main', port);
  535. WorkerMessageHandler.setup(handler, port);
  536. handler.send('ready', null);
  537. }
  538. };
  539. exports.WorkerMessageHandler = WorkerMessageHandler;
  540. function isMessagePort(maybePort) {
  541. return typeof maybePort.postMessage === 'function' && 'onmessage' in maybePort;
  542. }
  543. if (typeof window === 'undefined' && !(0, _is_node["default"])() && typeof self !== 'undefined' && isMessagePort(self)) {
  544. WorkerMessageHandler.initializeFromPort(self);
  545. }