puppeteer.js 5.5 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182
  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. browser = await puppeteer.launch(getBrowserOptions(options));
  32. return browser;
  33. };
  34. exports.getPageData = async (options, page) => {
  35. const privatePage = !page;
  36. try {
  37. if (privatePage) {
  38. page = await browser.newPage();
  39. }
  40. await setPageOptions(page, options);
  41. return await getPageData(browser, page, options);
  42. } finally {
  43. if (privatePage) {
  44. await page.close();
  45. }
  46. }
  47. };
  48. exports.closeBrowser = () => {
  49. if (browser) {
  50. return browser.close();
  51. }
  52. };
  53. function getBrowserOptions(options = {}) {
  54. const browserOptions = {};
  55. if (options.browserHeadless !== undefined) {
  56. browserOptions.headless = options.browserHeadless && !options.browserDebug;
  57. }
  58. browserOptions.args = options.browserArgs ? JSON.parse(options.browserArgs) : [];
  59. if (options.browserDisableWebSecurity === undefined || options.browserDisableWebSecurity) {
  60. browserOptions.args.push("--disable-web-security");
  61. }
  62. browserOptions.args.push("--no-pings");
  63. if (!options.browserStartMinimized && options.browserDebug) {
  64. browserOptions.args.push("--auto-open-devtools-for-tabs");
  65. }
  66. if (options.browserWidth && options.browserHeight) {
  67. browserOptions.args.push("--window-size=" + options.browserWidth + "," + options.browserHeight);
  68. }
  69. browserOptions.executablePath = options.browserExecutablePath || "chrome";
  70. if (options.userAgent) {
  71. browserOptions.args.push("--user-agent=" + options.userAgent);
  72. }
  73. return browserOptions;
  74. }
  75. async function setPageOptions(page, options) {
  76. if (options.browserWidth && options.browserHeight) {
  77. await page.setViewport({
  78. width: options.browserWidth,
  79. height: options.browserHeight
  80. });
  81. }
  82. if (options.browserBypassCSP === undefined || options.browserBypassCSP) {
  83. await page.setBypassCSP(true);
  84. }
  85. if (options.httpHeaders) {
  86. page.setExtraHTTPHeaders(options.httpHeaders);
  87. }
  88. if (options.browserStartMinimized) {
  89. const session = await page.target().createCDPSession();
  90. const { windowId } = await session.send("Browser.getWindowForTarget");
  91. await session.send("Browser.setWindowBounds", { windowId, bounds: { windowState: "minimized" } });
  92. }
  93. if (options.browserCookies && options.browserCookies.length) {
  94. await page.setCookie(...options.browserCookies);
  95. }
  96. }
  97. async function getPageData(browser, page, options) {
  98. const injectedScript = await scripts.get(options);
  99. await page.evaluateOnNewDocument(injectedScript);
  100. if (options.browserDebug) {
  101. await page.waitForTimeout(3000);
  102. }
  103. try {
  104. await pageGoto(page, options);
  105. } catch (error) {
  106. if (options.browserWaitUntilFallback && error.name == "TimeoutError") {
  107. const browserWaitUntil = NETWORK_STATES[(NETWORK_STATES.indexOf(options.browserWaitUntil) + 1)];
  108. if (browserWaitUntil) {
  109. options.browserWaitUntil = browserWaitUntil;
  110. return getPageData(browser, page, options);
  111. } else {
  112. throw error;
  113. }
  114. } else if (error.name != "TimeoutError") {
  115. throw error;
  116. }
  117. }
  118. try {
  119. if (options.browserWaitDelay) {
  120. await page.waitForTimeout(options.browserWaitDelay);
  121. }
  122. return await page.evaluate(async options => {
  123. const pageData = await singlefile.getPageData(options);
  124. if (options.includeInfobar) {
  125. await infobar.includeScript(pageData);
  126. }
  127. return pageData;
  128. }, options);
  129. } catch (error) {
  130. if (error.message && error.message.includes(EXECUTION_CONTEXT_DESTROYED_ERROR)) {
  131. const pageData = await handleJSRedirect(browser, options);
  132. if (pageData) {
  133. return pageData;
  134. } else {
  135. throw error;
  136. }
  137. } else {
  138. throw error;
  139. }
  140. }
  141. }
  142. async function handleJSRedirect(browser, options) {
  143. const pages = await browser.pages();
  144. const page = pages[1] || pages[0];
  145. try {
  146. await pageGoto(page, options);
  147. } catch (error) {
  148. if (error.name != "TimeoutError") {
  149. throw error;
  150. }
  151. }
  152. const url = page.url();
  153. if (url != options.url) {
  154. options.url = url;
  155. await browser.close();
  156. return exports.getPageData(options);
  157. }
  158. }
  159. async function pageGoto(page, options) {
  160. const loadOptions = {
  161. timeout: options.browserLoadMaxTime || 0,
  162. waitUntil: options.browserWaitUntil || NETWORK_IDLE_STATE
  163. };
  164. if (options.content) {
  165. await page.goto(options.url, { waitUntil: "domcontentloaded" });
  166. await page.setContent(options.content, loadOptions);
  167. } else {
  168. await page.goto(options.url, loadOptions);
  169. }
  170. }