puppeteer.js 5.7 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186
  1. /*
  2. * Copyright 2010-2020 Gildas Lormeau
  3. * contact : gildas.lormeau <at> gmail.com
  4. *
  5. * This file is part of SingleFile.
  6. *
  7. * The code in this file is free software: you can redistribute it and/or
  8. * modify it under the terms of the GNU Affero General Public License
  9. * (GNU AGPL) as published by the Free Software Foundation, either version 3
  10. * of the License, or (at your option) any later version.
  11. *
  12. * The code in this file is distributed in the hope that it will be useful,
  13. * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14. * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Affero
  15. * General Public License for more details.
  16. *
  17. * As additional permission under GNU AGPL version 3 section 7, you may
  18. * distribute UNMODIFIED VERSIONS OF THIS file without the copy of the GNU
  19. * AGPL normally required by section 4, provided you include this license
  20. * notice and a URL through which recipients can access the Corresponding
  21. * Source.
  22. */
  23. /* global singlefile, infobar, require, exports */
  24. const puppeteer = require("puppeteer-core");
  25. const scripts = require("./common/scripts.js");
  26. const EXECUTION_CONTEXT_DESTROYED_ERROR = "Execution context was destroyed";
  27. const NETWORK_IDLE_STATE = "networkidle0";
  28. const NETWORK_STATES = ["networkidle0", "networkidle2", "load", "domcontentloaded"];
  29. let browser;
  30. exports.initialize = async options => {
  31. if (options.browserServer) {
  32. browser = await puppeteer.connect({ browserWSEndpoint: options.browserServer });
  33. } else {
  34. browser = await puppeteer.launch(getBrowserOptions(options));
  35. }
  36. return browser;
  37. };
  38. exports.getPageData = async (options, page) => {
  39. const privatePage = !page;
  40. try {
  41. if (privatePage) {
  42. page = await browser.newPage();
  43. }
  44. await setPageOptions(page, options);
  45. return await getPageData(browser, page, options);
  46. } finally {
  47. if (privatePage) {
  48. await page.close();
  49. }
  50. }
  51. };
  52. exports.closeBrowser = () => {
  53. if (browser) {
  54. return browser.close();
  55. }
  56. };
  57. function getBrowserOptions(options = {}) {
  58. const browserOptions = {};
  59. if (options.browserHeadless !== undefined) {
  60. browserOptions.headless = options.browserHeadless && !options.browserDebug;
  61. }
  62. browserOptions.args = options.browserArgs ? JSON.parse(options.browserArgs) : [];
  63. if (options.browserDisableWebSecurity === undefined || options.browserDisableWebSecurity) {
  64. browserOptions.args.push("--disable-web-security");
  65. }
  66. browserOptions.args.push("--no-pings");
  67. if (!options.browserStartMinimized && options.browserDebug) {
  68. browserOptions.args.push("--auto-open-devtools-for-tabs");
  69. }
  70. if (options.browserWidth && options.browserHeight) {
  71. browserOptions.args.push("--window-size=" + options.browserWidth + "," + options.browserHeight);
  72. }
  73. browserOptions.executablePath = options.browserExecutablePath || "chrome";
  74. if (options.userAgent) {
  75. browserOptions.args.push("--user-agent=" + options.userAgent);
  76. }
  77. return browserOptions;
  78. }
  79. async function setPageOptions(page, options) {
  80. if (options.browserWidth && options.browserHeight) {
  81. await page.setViewport({
  82. width: options.browserWidth,
  83. height: options.browserHeight
  84. });
  85. }
  86. if (options.browserBypassCSP === undefined || options.browserBypassCSP) {
  87. await page.setBypassCSP(true);
  88. }
  89. if (options.httpHeaders) {
  90. page.setExtraHTTPHeaders(options.httpHeaders);
  91. }
  92. if (options.browserStartMinimized) {
  93. const session = await page.target().createCDPSession();
  94. const { windowId } = await session.send("Browser.getWindowForTarget");
  95. await session.send("Browser.setWindowBounds", { windowId, bounds: { windowState: "minimized" } });
  96. }
  97. if (options.browserCookies && options.browserCookies.length) {
  98. await page.setCookie(...options.browserCookies);
  99. }
  100. }
  101. async function getPageData(browser, page, options) {
  102. const injectedScript = await scripts.get(options);
  103. await page.evaluateOnNewDocument(injectedScript);
  104. if (options.browserDebug) {
  105. await page.waitForTimeout(3000);
  106. }
  107. try {
  108. await pageGoto(page, options);
  109. } catch (error) {
  110. if (options.browserWaitUntilFallback && error.name == "TimeoutError") {
  111. const browserWaitUntil = NETWORK_STATES[(NETWORK_STATES.indexOf(options.browserWaitUntil) + 1)];
  112. if (browserWaitUntil) {
  113. options.browserWaitUntil = browserWaitUntil;
  114. return getPageData(browser, page, options);
  115. } else {
  116. throw error;
  117. }
  118. } else if (error.name != "TimeoutError") {
  119. throw error;
  120. }
  121. }
  122. try {
  123. if (options.browserWaitDelay) {
  124. await page.waitForTimeout(options.browserWaitDelay);
  125. }
  126. return await page.evaluate(async options => {
  127. const pageData = await singlefile.getPageData(options);
  128. if (options.includeInfobar) {
  129. await infobar.includeScript(pageData);
  130. }
  131. return pageData;
  132. }, options);
  133. } catch (error) {
  134. if (error.message && error.message.includes(EXECUTION_CONTEXT_DESTROYED_ERROR)) {
  135. const pageData = await handleJSRedirect(browser, options);
  136. if (pageData) {
  137. return pageData;
  138. } else {
  139. throw error;
  140. }
  141. } else {
  142. throw error;
  143. }
  144. }
  145. }
  146. async function handleJSRedirect(browser, options) {
  147. const pages = await browser.pages();
  148. const page = pages[1] || pages[0];
  149. try {
  150. await pageGoto(page, options);
  151. } catch (error) {
  152. if (error.name != "TimeoutError") {
  153. throw error;
  154. }
  155. }
  156. const url = page.url();
  157. if (url != options.url) {
  158. options.url = url;
  159. await browser.close();
  160. return exports.getPageData(options);
  161. }
  162. }
  163. async function pageGoto(page, options) {
  164. const loadOptions = {
  165. timeout: options.browserLoadMaxTime || 0,
  166. waitUntil: options.browserWaitUntil || NETWORK_IDLE_STATE
  167. };
  168. if (options.content) {
  169. await page.goto(options.url, { waitUntil: "domcontentloaded" });
  170. await page.setContent(options.content, loadOptions);
  171. } else {
  172. await page.goto(options.url, loadOptions);
  173. }
  174. }