prerender - npm Package Compare versions

Comparing version 5.19.0 to 5.20.0

executionContext.patch

+diff --git a/lib/browsers/chrome.js b/lib/browsers/chrome.js
+index d3bb913..080da16 100644
+--- a/lib/browsers/chrome.js
++++ b/lib/browsers/chrome.js
+@@ -17,6 +17,7 @@ const CaptureScreenshotTimedOut = 'CaptureScreenshotTimedOut';
+ const UnableToCaptureScreenshot = 'UnableToCaptureScreenshot';
+ const PrintPdfTimedOut = 'PrintPdfTimedOut';
+ const UnableToCapturePdf = 'UnableToCapturePdf';
++const NoExecutionContext = 'NoExecutionContext';
+ 
+ chrome.name = 'Chrome';
+ 
+@@ -33,6 +34,9 @@ chrome.spawn = function (options) {
+ 		}
+ 
+ 		this.chromeChild = spawn(location, this.options.chromeFlags || [
++		// this.chromeChild = spawn('/usr/bin/taskset',  [
++		// 	'-c', '1',
++		// 	location,
+ 			'--headless',
+ 			'--disable-gpu',
+ 			'--remote-debugging-port=' + this.options.browserDebuggingPort,
+@@ -162,6 +166,7 @@ chrome.openTab = function (options) {
+ 				tab.prerender = options;
+ 				tab.prerender.errors = [];
+ 				tab.prerender.requests = {};
++				console.log("Cleanup 1");
+ 				tab.prerender.numRequestsInFlight = 0;
+ 
+ 				return this.setUpEvents(tab);
+@@ -203,18 +208,37 @@ chrome.setUpEvents = async function (tab) {
+ 		Network,
+ 		Emulation,
+ 		Log,
+-		Console
++		Console,
++		Runtime,
+ 	} = tab;
+ 
++
++
+ 	await Promise.all([
+ 		DOM.enable(),
+ 		Page.enable(),
+ 		Security.enable(),
+ 		Network.enable(),
+ 		Log.enable(),
+-		Console.enable()
++		Console.enable(),
++		Runtime.enable(),
+ 	]);
+ 
++	Runtime.executionContextCreated((event) => {
++		if ( event.context
++			&& event.context.auxData
++			&& (!tab.defaultExecutionContext
++				|| (tab.prerender.mainFrame && event.context.auxData.frameId == tab.prerender.mainFrame.id))) {
++			tab.defaultExecutionContext = event.context;
++		}
++	});
++
++	Page.frameNavigated((event) => {
++		if (!tab.prerender.mainFrame) {
++			tab.prerender.mainFrame = event.frame;
++		}
++	})
++
+ 	//hold onto info that could be used later if saving a HAR file
+ 	tab.prerender.pageLoadInfo = {
+ 		url: tab.prerender.url,
+@@ -260,6 +284,8 @@ chrome.setUpEvents = async function (tab) {
+ 		}, 1000);
+ 	});
+ 
++ Network.requestServedFromCache((e) => console.log("served from cache", e));
++
+ 	Security.certificateError(({ eventId }) => {
+ 		Security.handleCertificateError({
+ 			eventId,
+@@ -274,6 +300,7 @@ chrome.setUpEvents = async function (tab) {
+ 		tab.prerender.requests[params.requestId] = params.request.url;
+ 		if (tab.prerender.logRequests || this.options.logRequests) util.log('+', tab.prerender.numRequestsInFlight, params.request.url);
+ 
++		util.log(`Request to ${params.request.url}, ${tab.prerender.numRequestsInFlight}`);
+ 		if (!tab.prerender.initialRequestId) {
+ 			util.log(`Initial request to ${params.request.url}`);
+ 			tab.prerender.initialRequestId = params.requestId;
+@@ -297,6 +324,7 @@ chrome.setUpEvents = async function (tab) {
+ 			//during a redirect, we don't get the responseReceived event for the original request,
+ 			//so lets decrement the number of requests in flight here.
+ 			//the original requestId is also reused for the redirected request
++			console.log('R1');
+ 			tab.prerender.numRequestsInFlight--;
+ 
+ 			let redirectEntry = tab.prerender.pageLoadInfo.entries[params.requestId];
+@@ -343,6 +371,7 @@ chrome.setUpEvents = async function (tab) {
+ 		}
+ 
+ 		if (params.type === "EventSource") {
++			console.log('Response 1');
+ 			tab.prerender.numRequestsInFlight--;
+ 			tab.prerender.lastRequestReceivedAt = new Date().getTime();
+ 			if (tab.prerender.logRequests || this.options.logRequests) util.log('-', tab.prerender.numRequestsInFlight, tab.prerender.requests[params.requestId]);
+@@ -365,6 +394,7 @@ chrome.setUpEvents = async function (tab) {
+ 				util.log(`Initial request finished ${request}`);
+ 			}
+ 
++			console.log('Loading finished 1', tab.prerender.numRequestsInFlight);
+ 			tab.prerender.numRequestsInFlight--;
+ 			tab.prerender.lastRequestReceivedAt = new Date().getTime();
+ 
+@@ -384,6 +414,7 @@ chrome.setUpEvents = async function (tab) {
+ 	//all outstanding requests will fire this event
+ 	Network.loadingFailed((params) => {
+ 		if (tab.prerender.requests[params.requestId]) {
++			console.log("Loading failed 1");
+ 			tab.prerender.numRequestsInFlight--;
+ 			if (tab.prerender.logRequests || this.options.logRequests) util.log('-', tab.prerender.numRequestsInFlight, tab.prerender.requests[params.requestId]);
+ 			delete tab.prerender.requests[params.requestId];
+@@ -481,6 +512,11 @@ chrome.loadUrlThenWaitForPageLoadEvent = function (tab, url, onNavigated) {
+ 					if (!finished) {
+ 						finished = true;
+ 						util.log('page timed out', tab.prerender.url);
++					  // console.log(
++						// 	JSON.stringify(
++						// 		tab.prerender.pageLoadInfo.entries, null, 2
++						// 	)
++						// );
+ 
+ 						const timeoutStatusCode = tab.prerender.timeoutStatusCode || this.options.timeoutStatusCode;
+ 						if (timeoutStatusCode) {
+@@ -501,6 +537,7 @@ chrome.loadUrlThenWaitForPageLoadEvent = function (tab, url, onNavigated) {
+ 				let width = parseInt(tab.prerender.width, 10) || 1440;
+ 				let height = parseInt(tab.prerender.height, 10) || 718;
+ 
++				console.log(` == width: ${width}, height: ${height}`);
+ 				Emulation.setDeviceMetricsOverride({
+ 					width: width,
+ 					screenWidth: width,
+@@ -561,6 +598,7 @@ chrome.checkIfPageIsDoneLoading = function (tab) {
+ 			expression: 'window.prerenderReady'
+ 		}).then((result) => {
+ 			let prerenderReady = result && result.result && result.result.value;
++			console.log("prerenderReady", prerenderReady);
+ 			let shouldWaitForPrerenderReady = typeof prerenderReady == 'boolean';
+ 			let waitAfterLastRequest = tab.prerender.waitAfterLastRequest || this.options.waitAfterLastRequest;
+ 
+@@ -574,6 +612,14 @@ chrome.checkIfPageIsDoneLoading = function (tab) {
+ 				tab.prerender.lastRequestReceivedAt < ((new Date()).getTime() - waitAfterLastRequest)
+ 
+ 			const timeSpentAfterFirstPrerenderReady = (tab.prerender.firstPrerenderReadyTime && (new Date().getTime() - tab.prerender.firstPrerenderReadyTime)) || 0;
++			try {
++			console.log(` == CHECK IF PAGE DONE LOADING doneLoading: ${doneLoading}`);
++			console.log(` == CHECK IF PAGE DONE LOADING numRequestsInFlight: ${tab.prerender.numRequestsInFlight}`);
++			console.log(` == CHECK IF PAGE DONE LOADING waitAfterLastRequest: ${waitAfterLastRequest}`);
++			console.log(` == SHOULD WAIT FOR PRERENDER READY: ${shouldWaitForPrerenderReady}`);
++			} catch (e) {
++				console.log("c error", e);
++			}
+ 
+ 			resolve(
+ 				(!shouldWaitForPrerenderReady && doneLoading) ||
+@@ -636,7 +682,10 @@ const getHtmlWithShadowDomFunction = () => {
+   const attributeNames = htmlNode.getAttributeNames();
+   const attrStringList = attributeNames.map((attributeName) => (`${attributeName}="${htmlNode.getAttribute(attributeName)}"`))
+ 
+-  return '<!DOCTYPE html>'+'<html '+ attrStringList.join(' ') +'>'+innerText+'</html>';
++  return `<!DOCTYPE html>
++  <html ${attrStringList.join(' ')}>
++    ${innerText}
++  </html>`;
+ }
+ 
+ chrome.parseHtmlFromPage = function (tab) {
+@@ -654,10 +703,19 @@ chrome.parseHtmlFromPage = function (tab) {
+ 			? getHtmlWithShadowDomFunction.toString()
+ 			: getHtmlFunction.toString();
+ 
+-		tab.Runtime.evaluate({
+-			expression: "document.firstElementChild.outerHTML"
+-		}).then((resp) => {
++		if (!tab.defaultExecutionContext) {
++			util.log('No default exectution context found. Can\'t evaluate page. ', tab.prerender.url);
++			tab.prerender.statusCode = 504;
++			tab.prerender.errors.push(NoExecutionContext);
++			reject();
++		}
+ 
++		tab.Runtime.callFunctionOn({
++			expression: getHtmlFunctionText,
++			functionDeclaration: getHtmlFunctionText,
++			executionContextId: tab.defaultExecutionContext.id,
++			userGesture: true
++		}).then((resp) => {
+ 			tab.prerender.content = resp.result.value;
+ 			if (tab.prerender.content === undefined) {
+ 				tab.prerender.statusCode = 504;

patch-one-processor.patch

+diff --git a/lib/browsers/chrome.js b/lib/browsers/chrome.js
+index 2568daa..7afc61d 100644
+--- a/lib/browsers/chrome.js
++++ b/lib/browsers/chrome.js
+@@ -32,7 +32,9 @@ chrome.spawn = function (options) {
+ 			return reject();
+ 		}
+ 
+-		this.chromeChild = spawn(location, this.options.chromeFlags || [
++		this.chromeChild = spawn('/usr/bin/taskset',  [
++			'-c', '1',
++			location,
+ 			'--headless',
+ 			'--disable-gpu',
+ 			'--remote-debugging-port=' + this.options.browserDebuggingPort,

CHANGELOG.md

@@ -5,2 +5,5 @@ # Change Log
 
+## 5.20.0 - 2022-05-06
+- If a request made during rendering gets a 5XX response the render process will be marked as dirty
+
 ## 5.19.0 - 2022-03-23
@@ -7,0 +10,0 @@ - Ability to parse content from the shadow DOM

lib/browsers/chrome.js

@@ -351,2 +351,6 @@ const CDP = require('chrome-remote-interface');
 		}
+
+		if (params.response && params.response.status >= 500 && params.response.status < 600) { // 5XX
+			tab.prerender.dirtyRender = true;
+		}
 	});
@@ -353,0 +357,0 @@

package.json

@@ -5,3 +5,3 @@ {
   "description": "Service to prerender Javascript rendered pages for SEO",
-  "version": "5.19.0",
+  "version": "5.20.0",
   "license": "MIT",
@@ -8,0 +8,0 @@ "repository": {

New alerts

Network access

Supply chain risk

This module accesses the network.

Found 2 instances in 1 package

Shell access

Supply chain risk

This module accesses the system shell. Accessing the system shell increases the risk of executing arbitrary code.

Found 1 instance in 1 package

Environment variable access

Supply chain risk

Package accesses environment variables, which may be a sign of credential stuffing or data theft.

Found 19 instances in 1 package

Filesystem access

Supply chain risk

Accesses the file system, and could potentially read sensitive data.

Found 1 instance in 1 package

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

Found 1 instance in 1 package

Fixed alerts

Network access

Supply chain risk

This module accesses the network.

Found 2 instances in 1 package

Shell access

Supply chain risk

This module accesses the system shell. Accessing the system shell increases the risk of executing arbitrary code.

Found 1 instance in 1 package

Environment variable access

Supply chain risk

Package accesses environment variables, which may be a sign of credential stuffing or data theft.

Found 19 instances in 1 package

Filesystem access

Supply chain risk

Accesses the file system, and could potentially read sensitive data.

Found 1 instance in 1 package

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

98755

9.45%

Number of package files

23

9.52%

Lines of code

1616

0.19%

No dependency changes