There are several elements of a Protractor configuration that I need to modify at run-time, most importantly some values of the multiCapabilities object:
...
multiCapabilities: [
{
'build': buildNumber,
'app': 'app' + buildNumber + '.apk',
'browserName': '',
'appiumVersion': '1.4.16',
'deviceName': 'Android Emulator',
'deviceOrientation': 'portrait',
'platformVersion': '5.1',
'platformName': 'Android',
'autoWebview': true
},
{
...
}
],
...
Specifically, I'm running these tests against a specific build of an app that has been uploaded to SauceLabs. I'd like to be able to somehow set the 'build' and 'app' values dynamically.
Unfortunately, the Grunt Protractor Runner does not support the multiCapabilities feature, otherwise it could be handled using that wrapper.
One option would be to use browser.params and parameterize your tests from the command-line arguments. In this case, use getMultiCapabilities() method:
getMultiCapabilities: function () {
var buildNumber = browser.params.build,
app = browser.params.app;
return [
{
'build': buildNumber,
'app': app,
'browserName': '',
'appiumVersion': '1.4.16',
'deviceName': 'Android Emulator',
'deviceOrientation': 'portrait',
'platformVersion': '5.1',
'platformName': 'Android',
'autoWebview': true
},
{
...
}
],
},
Usage:
protractor protractor.conf.js --params.build="build" --params.app="app"
Related
I was able prepare my OpenUI5 app for testing. I am having issue with the working directory for tests, because for my test I need to use a pre-populated sqlite3 database. For testing I use chrome with parameters --user-data-dir="C:\tmp" --profile-directory="karma", and I can place this file into chrome profile path. But I am not able to do this with karma, because karma always start with randomly generated id in path (karma-xxxxxx).
I am trying run custom browser config, with above chrome parameters, but it does not work.
module.exports = function(config) {
"use strict";
var chromeFlags = [
"--window-size=1280,1024",
"--disable-web-security",
"--allow-file-access-from-files",
'--user-data-dir="C:\\tmp"',
'--profile-directory="karma"'
];
config.set({
basePath: '',
ui5: {
type: "application",
preload: 'async',
animation: 'false',
paths: {
webapp: "www", // application
}
},
frameworks: ["ui5"],
browsers: ["CustomChrome"],
browserConsoleLogOptions: {
level: "error"
},
customLaunchers: {
CustomChrome: {
base: "Chrome",
flags: chromeFlags
},
CustomChromeHeadless: {
base: "ChromeHeadless",
flags: chromeFlags
}
},
});
};
with this config, chrome always start like this, so automatic testing does not work:
I would like to set headless option in my protractor configuration for internet explorer, but I cannot find anyting related to this. They only say how to do it in Chrome and Firefox: Adding chrome and Firefox specific options.
This is what I have so far:
multiCapabilities: [
{
'browserName': 'chrome',
chromeOptiaons: {
args: ["--headless"]
},
},
{
'browserName': 'firefox',
'moz:firefoxOptions': {
args: ["--headless"]
}
},
{
'browserName': 'internet explorer',
internetExplorerOptions: { <---------------------------
args: ["--headless"] <---------------------------
}
}
],
internetExplorerOptions is not working
So how do I add the 'Options / args' option to IE?
IE does not have support for a headless mode. But you can try to use triflejs.
I did not try this solution by myself but seems like it's the only option since IE is basically obsolete browser and it does not receive any updates.
First of all, you need IEDriverServer.
I will try to make this as simple as possible. The proper way to get and install this is, by using protractor’s built-in functionality.
Install Webdriver-manager and gulp, globally.
npm installwebdriver-manager gulp -g
Browse to node_modules/protractor/bin and run
npm webdriver-manager --standalone update
npm webdriver-manager --ie update
Then, you will need to manually start the webdriver server
For this, you will need to run
npm webdriver-manager --ie start
As simple as that.
After that, the 2nd step is to get the config file right.
The spicy and “different” stuff is: Capabilities, direct connect, and
LocalSeleniumStandaloneOpts
Necessary and Important examples for config file, in order to run the spec test suite:
capabilities: {
'browserName': 'internet explorer',
'ignoreProtectedModeSettings': true,
'platform': 'ANY',
'version': '11',
args: ['--silent', '--no-sandbox', '--test-type=browser', '--lang=US', '--start-maximized'], //,'--headless', '--disable-gpu'
prefs: {
'download': {
'prompt_for_download': false,
'directory_upgrade': true,
'extensions_to_open': '',
'default_directory': process.cwd() + '/downloads/'
},
}
},
localSeleniumStandaloneOpts: {
jvmArgs: [
'-Dwebdriver.ie.driver=node_modules/protractor/node_modules/webdriver-manager/selenium/IEDriverServer3.14.0.exe'
]
},
directConnect: false,
framework: 'jasmine',
seleniumArgs: ['-
Dwebdriver.ie.driver=node_modules/protractor/node_modules/webdriver-
manager/selenium/IEDriverServer3.14.0.exe'],
seleniumAddress: 'http://localhost:4444/wd/hub'
I am new to protractor and I want to be able to run my chrome browser painted or headless.
So I set up something like this
let chrome = {
browserName: 'chrome',
platform: 'MAC',
'max-duration': '1800',
};
let chromeHeadless = {
browserName: 'chrome',
chromeOptions: {
args: [ "--headless", "--disable-gpu", "--window-size=800,600" ]
}
};
browserDefault = browser.params.browserToUse
exports.config = {
params: {
'browserToUse': "get from user'
},
capabilities: browserDefault,
}
and i ran this code as
protractor config.js --params.browserToUse='chromeHeadless'
But this does not work. Protractor fails saying it does not understand "browser.params.browserInput". Whats the right way to make protractor dynamically choose chrome or chromeheadless based on the input
The global variable browser is only init when code run into onPrepare(). You used browser outside onPrepare() function, browser have not been inited, it is undefined, so you met the error.
Another point you need to get it's when the variable browser inited, a browser window has been opened, means protractor has know which capabilities to launch the browser. Therefore you can't use browser.params.xxx to specify which capabilities, you need to tell protractor the capabilities before it init the browser variable.
let capabilitiesMap = {
'chrome-headful' : {
browserName: 'chrome',
platform: 'MAC',
'max-duration': '1800',
},
'chrome-headless': {
browserName: 'chrome',
chromeOptions: {
args: [ "--headless", "--disable-gpu", "--window-size=800,600" ]
}
}
};
let browserToUse = 'chrome-headful'; // set default value
// extract the browserToUse value from cli
process.argv.slice(3).forEach(function(arg) {
var name = arg.split('=')[0];
var value = arg.split('=')[1];
var name = name.replace('--', '');
if (name === 'browserToUse') {
if (Object.prototype.hasOwnProperty.call(capabilitiesMap, value) ) {
browserToUse = value;
}
}
});
let config = {
seleniumAddress: '',
specs: [],
onPrepare: function() {}
};
config.capabilities = capabilitiesMap[browserToUse];
exports.config = config;
CLI example: protractor conf.js --browserToUse=chrome-headless
I also came across this issue and soleved it using the getMultiCapabilities() function in your conf.js
const _ = require('lodash');
let capabilities = {
chrome: {
browserName: 'chrome',
platform: 'MAC',
'max-duration': '1800',
},
chromeHeadless : {
browserName: 'chrome',
chromeOptions: {
args: [ "--headless", "--disable-gpu", "--window-size=800,600" ]
}
}
}
getMultiCapabilities() {
const browsers = this.params.browserToUse.split(',');//if you pass more than one browser e.g chrome,chromeHeadless
const cap = _(capabilities).pick(browsers).values().value(); //this uses the lodash npm module
return cap;
},
In a testing context working with just Chrome, I did the following. In capabilities:
chromeOptions: {
args: []
}
beforeLaunch: function() {
//at this point browser is not yet defined, so process command line directly
if (process.argv[process.argv.length-1].search('headless=true')>-1){
config.capabilities.chromeOptions.args.push("--headless");
config.capabilities.chromeOptions.args.push("--disable-gpu");
config.capabilities.chromeOptions.args.push("--window-size=1600,1000");
}
}
That way by the time the browser was launched, it had the right configuration. Where I have "headless=true", you might want "Chrome-headless."
And then on the command line I call it like you do with --params.headless=falseso that should I want to find it in the script itself later (after the browser has launched), it is readily available.
Note I had just one command line parameter and control of the command line, so it felt okay to assume this parameter was the last.
I am running my Protractor Test from command npm run e2e
I want a way so that if I pass npm run e2e firefox then my test will get executed on Firefox browser.
Or if I run npm run e2e chrome then it should run on chrome
if I pass both npm run e2e firefox chrome then my test should run on both the browser in parallel.
Is it possible to parametrize protractor config file?
Similarly if I can pass test suite name via command and it should execute only tests under that particular test suite.
Here is my config file and this is what I want to achieve:
`//var HtmlReporter = require('protractor-html-screenshot-reporter');
exports.config = {
allScriptsTimeout: 30000,
//Add parameters for browser names
params:{
pass: {
browserName : 'chrome',
testSuitName : 'e2e/TestSuites/_BVT/*.js',
}
},
suites: {
//Define here List of Sanity Test Scenarios:
BVT : testSuitName,
},
// configure multiple browsers to run tests
multiCapabilities: [
shardTestFiles: true,
maxInstances: 2
{'browserName': browserName}
],
baseUrl: 'http://mytestUrl/',
framework: 'jasmine2',
jasmineNodeOpts: {
defaultTimeoutInterval: 30000
},
onPrepare: function() {
var jasmineReporters = require('jasmine-reporters');
browser.driver.manage().window().maximize();
return browser.getProcessedConfig().then(function(config) {
var browserName = config.capabilities.browserName;
var junitReporter = new jasmineReporters.JUnitXmlReporter({
consolidateAll: true,
savePath: 'tests/test-results',
filePrefix: browserName + '-xmloutput',
modifySuiteName: function(generatedSuiteName, suite) {
return browserName + '.' + generatedSuiteName;
}
});
jasmine.getEnv().addReporter(junitReporter);
});
},
resultJsonOutputFile: 'tests/test-results/output.json'
};`
Would appreciate any help on this.
Thanks in advance.
I know this post is a bit old now, but this solution may help people having a similar problem.
Use multiple config files, one for each browser type, setup a base config file, and require that into each of the other config files, and then create an npm script for each config file. No need for parameters and everything stacks nicely.
so the base config (called something like "protractor_base.js")would look something like:
exports.config = {
seleniumAddress: 'http://localhost:4444/wd/hub',
rootElement: '[ng-app]',
allScriptsTimeout: 60000,
framework: 'jasmine',
specs: ['example-spec.js']
};
And then your other configs (ie "protractor_chrome.conf.js") would look something like:
protractor = require('./protractor_base.js');
var config = protractor.config;
config.capabilities: {
'browserName': 'chrome'
};
exports.config = config;
You can then specify a multi browser config, just a chrome one, etc.
--capabilities.chromeOptions.args=headless --capabilities.chromeOptions.args=disable-gpu --capabilities.chromeOptions.args=window-size=1200,600
The above code should work for you.
I was looking for passing parameters from command line to protractor config file and found a way to do this:
npm run e2e -- --baseUrl=http://testurl:8080 --suite=suite_name_defined_in_config --capabilities.browserName=browser_Name
where my npm package.json :
"e2e": "protractor tests/protractor-conf.js",
and config file contains :
suites: {
BVT: 'e2e/TestSuites/_BVT/*.js',
Full: 'e2e/TestSuites/Full/**/*.js',
},
capabilities: {
'browserName': 'chrome'
},
baseUrl: 'http://localhost:8080/',
I am trying to run E2E tests on multiple browsers on browser stack, I took the reference from
E2E testing on multiple/parallel browsers in Protractor?
and
Internet Explorer Selenium protractor e2e tests
but the error I get every time I try to run the tests -
ERROR - Unable to start a WebDriver session.
C:\MrWhiteMVP\whitemvp-integrationtests_develop\node_modules\gulp-protractor\node_modules\protractor\node_modules\selenium-webdriver\lib\atoms\error.js:113
var template = new Error(this.message);
^
UnknownError: Authorization required
at new bot.Error (C:\MrWhiteMVP\whitemvp-integrationtests_develop\node_modules\gulp-protractor\node_modules\protractor\node_modules\selenium-webdriver\lib\atoms\error.js:113:18)
at Object.bot.response.checkResponse (C:\MrWhiteMVP\whitemvp-integrationtests_develop\node_modules\gulp-protractor\node_modules\protractor\node_modules\selenium-webdriver\lib\atoms\response.js:103:11)
But if I run the tests on 1 browser, then it works perfectly fine. this is how my conf file looks like
'use strict';
exports.config = {
seleniumAddress: 'http://hub.browserstack.com/wd/hub',
multicapabilities: [{
'browserstack.user': 'testuser',
'browserstack.key': 'testkey',
// Needed for testing localhost
// 'browserstack.local': 'true',
// Settings for the browser you want to test
'browserName': 'chrome',
'version': '36.0',
'os': 'OS X',
'os_version': 'Mavericks',
'resolution': '1024x768'
},
{
'browserstack.user': 'testuser',
'browserstack.key': 'testkey',
// Needed for testing localhost - 'browserstack.local': 'true',
// Settings for the browser
'browserName': 'firefox',
'os': 'windows'
}],
baseUrl: 'http://origin-develop.mvp.livebranches.com',
// The require statement must be down here, since jasmine-reporters
// needs jasmine to be in the global and protractor does not guarantee
// this until inside the onPrepare function.
onPrepare: function()
{
require('jasmine-reporters');
jasmine.getEnv().addReporter(new jasmine.JUnitXmlReporter('xmloutput', true, true));
},
},
};
could anyone please tell me what am I doing wrong here, also we use gulp ti run specs but my problem is it is saying not even going beyond authentication.
I think that first of all, you have an extra comma on your config.
onPrepare: function()
{
require('jasmine-reporters');
jasmine.getEnv().addReporter(new jasmine.JUnitXmlReporter('xmloutput', true, true));
},
}, <-----
};