I have a simple test using the selenium-webdriver javascript package that finds a text field by name and enters text. The test runs fine in Firefox and Chrome, but fails in Phantomjs. I also tried finding the element by xpath and id, but it keeps failing in Phantomjs. Has anyone had a similar experience and could point me in the right direction for debugging?
Thanks
I was facing same issue. In my case, problem was that we have self signed certificate on our test server. I found solution here: phantomjs-and-self-signed-certificates
You have to set desired capabilities for your driver:
DesiredCapabilities dCaps = new DesiredCapabilities();
dCaps.setJavascriptEnabled(true);
dCaps.setCapability(PhantomJSDriverService.PHANTOMJS_CLI_ARGS, new String[]{"--web-security=no", "--ignore-ssl-errors=yes"});
dCaps.setJavascriptEnabled(true);
driver = new PhantomJSDriver(dCaps);
To ignore self signed certificates set "--ignore-ssl-errors=yes"
Related
We have some functionality for exporting data to an excel file.
When the 'export' button is clicked, some client-side javascript is called, firstly checking the client browser version, and based on this, deciding which way to render the excel document.
It is working in Chrome & Firefox & IE11 when tested locally.
However, when I remotely test using a windows 10 machine running Edge browser, the excel is not rendered.
I might add that my local machine is a Win7 machine and Im running VS2012 and IE11. The remote machine is Win10 with Edge, hence the need to test remotely.
I've tried the emulation in IE11 F12 dev tools but cant replicate the Edge error there.
An error of 'undefined or null reference' is thrown for 'open' when using the following code:
excelIFrame.document.open("txt/html", "replace");
excelIFrame.document.write(sHTML);
excelIFrame.document.close();
excelIFrame.focus();
excelIFrame.document.execCommand("SaveAs", true, "Spreadsheet.xls");
The iframe exists in the html and is not added dynamically.
<iframe id="excelIFrame" style="display:none"></iframe>
I have tried the following possible solutions to get this working, to no avail -
Possible Solution 1: Same 'undefined or null reference error when assigning the document to a temp var
var doc = excelIFrame.document;
doc.open("txt/html", "replace");
doc.write(sHTML);
doc.close();
doc.focus();
doc.execCommand("SaveAs", true, "Spreadsheet.xls");
Possible Solution 2: Using the contentWindow property of the iFrame. No error thrown, it just opens 'about:blank' containing no content.
excelIFrame.contentWindow.contents = sHTML;
excelIFrame.src = 'javascript:window["contents"]';
Totally at a loss with this at this stage.
The page is an angularJS web page.
From reading up on it, I'm aware the document.open is problematic in edge when using iframes. But the following link document.open fails in an iframe I felt would solve the problem.
Any thoughts or suggestions greatly appreciated.
This may be helpful to others who are searching for it.
//For Edge browser ….. U have to write separate logic for each browser
if (ua.match(/Edge/)){
var blob = new Blob([sHTML], {type: 'data:application/vnd.ms-excel'});
window.navigator.msSaveBlob(blob, "P2P_Report_"+new Date().getTime()+".xls");
}
I have a program that tests some of our sites with javascript disabled. It worked well, until we updated from WebDriver 2.4.5 to the latest. Now I get this error:
java.lang.IllegalArgumentException: Preference javascript.enabled may not be overridden: frozen value=true, requested value=false
It looks like the argument to disable JS is no longer allowed. I can't downgrade the WebDriver because the earlier versions don't work correctly with our updated Firefox.
What are my options for testing with JS disabled? I know Chrome never worked before, and now Firefox doesn't. I tried searching for "webdriver js disabled" but the results just bring back the javascript.enabled, false argument that no longer seems to work.
Best I could come up with is;
Create a new Firefox profile called NoJs by starting it with -P
Go to about:config and disable javascript there
Use this profile in selenium
ProfilesIni allProfiles = new ProfilesIni();
FirefoxProfile profile = allProfiles.getProfile("NoJs");
WebDriver driver = new FirefoxDriver(profile);
How do i set up a custom protocol handler in chrome? Something like:
myprotocol://testfile
I would need this to send a request to http://example.com?query=testfile, then send the httpresponse to my extension.
The following method registers an application to a URI Scheme. So, you can use mycustproto: in your HTML code to trigger a local application. It works on a Google Chrome Version 51.0.2704.79 m (64-bit).
I mainly used this method for printing document silently without the print dialog popping up. The result is pretty good and is a seamless solution to integrate the external application with the browser.
HTML code (simple):
Click Me
HTML code (alternative):
<input id="DealerName" />
<button id="PrintBtn"></button>
$('#PrintBtn').on('click', function(event){
event.preventDefault();
window.location.href = 'mycustproto:dealer ' + $('#DealerName').val();
});
URI Scheme will look like this:
You can create the URI Scheme manually in registry, or run the "mycustproto.reg" file (see below).
HKEY_CURRENT_USER\Software\Classes
mycustproto
(Default) = "URL:MyCustProto Protocol"
URL Protocol = ""
DefaultIcon
(Default) = "myprogram.exe,1"
shell
open
command
(Default) = "C:\Program Files\MyProgram\myprogram.exe" "%1"
mycustproto.reg example:
Windows Registry Editor Version 5.00
[HKEY_CURRENT_USER\Software\Classes\mycustproto]
"URL Protocol"="\"\""
#="\"URL:MyCustProto Protocol\""
[HKEY_CURRENT_USER\Software\Classes\mycustproto\DefaultIcon]
#="\"mycustproto.exe,1\""
[HKEY_CURRENT_USER\Software\Classes\mycustproto\shell]
[HKEY_CURRENT_USER\Software\Classes\mycustproto\shell\open]
[HKEY_CURRENT_USER\Software\Classes\mycustproto\shell\open\command]
#="\"C:\\Program Files\\MyProgram\\myprogram.exe\" \"%1\""
C# console application - myprogram.exe:
using System;
using System.Collections.Generic;
using System.Text;
namespace myprogram
{
class Program
{
static string ProcessInput(string s)
{
// TODO Verify and validate the input
// string as appropriate for your application.
return s;
}
static void Main(string[] args)
{
Console.WriteLine("Raw command-line: \n\t" + Environment.CommandLine);
Console.WriteLine("\n\nArguments:\n");
foreach (string s in args)
{
Console.WriteLine("\t" + ProcessInput(s));
}
Console.WriteLine("\nPress any key to continue...");
Console.ReadKey();
}
}
}
Try to run the program first to make sure the program has been placed in the correct path:
cmd> "C:\Program Files\MyProgram\myprogram.exe" "mycustproto:Hello World"
Click the link on your HTML page:
You will see a warning window popup for the first time.
To reset the external protocol handler setting in Chrome:
If you have ever accepted the custom protocol in Chrome and would like to reset the setting, do this (currently, there is no UI in Chrome to change the setting):
Edit "Local State" this file under this path:
C:\Users\Username\AppData\Local\Google\Chrome\User Data\
or Simply go to:
%USERPROFILE%\AppData\Local\Google\Chrome\User Data\
Then, search for this string: protocol_handler
You will see the custom protocol from there.
Note: Please close your Google Chrome before editing the file. Otherwise, the change you have made will be overwritten by Chrome.
Reference:
https://msdn.microsoft.com/en-us/library/aa767914(v=vs.85).aspx
Chrome 13 now supports the navigator.registerProtocolHandler API. For example,
navigator.registerProtocolHandler(
'web+custom', 'http://example.com/rph?q=%s', 'My App');
Note that your protocol name has to start with web+, with a few exceptions for common ones (like mailto, etc). For more details, see: http://updates.html5rocks.com/2011/06/Registering-a-custom-protocol-handler
This question is old now, but there's been a recent update to Chrome (at least where packaged apps are concerned)...
http://developer.chrome.com/apps/manifest/url_handlers
and
https://github.com/GoogleChrome/chrome-extensions-samples/blob/e716678b67fd30a5876a552b9665e9f847d6d84b/apps/samples/url-handler/README.md
It allows you to register a handler for a URL (as long as you own it). Sadly no myprotocol:// but at least you can do http://myprotocol.mysite.com and can create a webpage there that points people to the app in the app store.
This is how I did it. Your app would need to install a few reg keys on installation, then in any browser you can just link to foo:\anythingHere.txt and it will open your app and pass it that value.
This is not my code, just something I found on the web when searching the same question. Just change all "foo" in the text below to the protocol name you want and change the path to your exe as well.
(put this in to a text file as save as foo.reg on your desktop, then double click it to install the keys)
-----Below this line goes into the .reg file (NOT including this line)------
REGEDIT4
[HKEY_CLASSES_ROOT\foo]
#="URL:foo Protocol"
"URL Protocol"=""
[HKEY_CLASSES_ROOT\foo\shell]
[HKEY_CLASSES_ROOT\foo\shell\open]
[HKEY_CLASSES_ROOT\foo\shell\open\command]
#="\"C:\\Program Files (x86)\\Notepad++\\notepad++.exe\" \"%1\""
Not sure whether this is the right place for my answer, but as I found very few helpful threads and this was one of them, I am posting my solution here.
Problem: I wanted Linux Mint 19.2 Cinnamon to open Evolution when clicking on mailto links in Chromium. Gmail was registered as default handler in chrome://settings/handlers and I could not choose any other handler.
Solution:
Use the xdg-settings in the console
xdg-settings set default-url-scheme-handler mailto org.gnome.Evolution.desktop
Solution was found here https://alt.os.linux.ubuntu.narkive.com/U3Gy7inF/kubuntu-mailto-links-in-chrome-doesn-t-open-evolution and adapted for my case.
I've found the solution by Jun Hsieh and MuffinMan generally works when it comes to clicking links on pages in Chrome or pasting into the URL bar, but it doesn't seem to work in a specific case of passing the string on the command line.
For example, both of the following commands open a blank Chrome window which then does nothing.
"c:\Program Files (x86)\Google\Chrome\Application\chrome.exe" "foo://C:/test.txt"
"c:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --new-window "foo://C:/test.txt"
For comparison, feeding Chrome an http or https URL with either of these commands causes the web page to be opened.
This became apparent because one of our customers reported that clicking links for our product from a PDF being displayed within Adobe Reader fails to invoke our product when Chrome is the default browser. (It works fine with MSIE and Firefox as default, but not when either Chrome or Edge are default.)
I'm guessing that instead of just telling Windows to invoke the URL and letting Windows figure things out, the Adobe product is finding the default browser, which is Chrome in this case, and then passing the URL on the command line.
I'd be interested if anyone knows of Chrome security or other settings which might be relevant here so that Chrome will fully handle a protocol handler, even if it's provided via the command line. I've been looking but so far haven't found anything.
I've been testing this against Chrome 88.0.4324.182.
open
C:\Users\<Username>\AppData\Local\Google\Chrome\User Data\Default
open Preferences then search for excluded_schemes you will find it in 'protocol_handler' delete this excluded scheme(s) to reset chrome to open url with default application
I am trying to use Selenium WebDriverJS with the iOS browser. Unfortunately this doesn't work as I need it to do.
I've set up the "iWebDriver"-Project like described at the iPhoneDriver wiki page. I've copied the python script and it worked. Even a small nodeJS project worked, but when I tried to run it from the Browser Firefox gives me the following Error message.
Error: Unable to create a new client for this browser. The WebDriver session ID has not been defined.
The Code I'm trying to run:
var driver = new webdriver.Builder().
usingServer('http://localhost:3001/wd/hub').
withCapabilities({'browserName': 'iPhone'}).
build();
driver.get('http://www.google.com');
driver.quit();
I've compiled webdriverjs like discribed at the WebDriverJs wiki page. And yes, webdriver.js is loaded ;)
So I dug in and I think I have a solution. It seems like a hack but for firefox, and the firefox extension that I am making it works perfectly fine so far.
I edited the webdriver.js file's webdriver.Builder.prototype.build as so
webdriver.Builder.prototype.build = function() {
var a;
// if(webdriver.FirefoxDomExecutor.isAvailable()) {
// return a = new webdriver.FirefoxDomExecutor, webdriver.WebDriver.createSession(a, this.getCapabilities())
// }
a = new webdriver.http.CorsClient(this.getServerUrl());
a = new webdriver.http.Executor(a);
// if(this.getSession()) {
// return webdriver.WebDriver.attachToSession(a, this.getSession())
// }
// throw Error("Unable to create a new client for this browser. The WebDriver session ID has not been defined.");
return webdriver.WebDriver.createSession(a, this.getCapabilities());
};
Basically I am forcing it to create a session, not sure why FirefoxDomExecutor was required since it works just fine on firefox and chrome with changes that I made.
Anyway thanks for getting me on the right track, and hopefully this will help others.
It seems like WebDriverJS is not capable of creating a new Session when not running in Node.js.
You have to create a Session manually and provide WebDriverJS with the session-id as the wdsid GET-Parameter (e.g. urltowebdriver.html?wdsid=someID&wdurl=localhost:9000).
I am using library called arbor.js(beautified). It contains the following line of code,
i = new Worker(<path to arbor.js>);
At this point an error is thrown, and the following message is reported in Firebug,
"Could not get domain!"
The line which gives this error is 258. The arbor_path() function on that line returns "js/", so effectively function called is i = new Worker("js/arbor.js").
This used to work perfectly in Firefox 7. Does anybody know how to solve it, or what the problem could be?
You're running into https://bugzilla.mozilla.org/show_bug.cgi?id=683280
It'll be fixed in Firefox 9 in about a month....
I came across the same thing, however I only noticed the problem on one of my computers, and not the other (both running Firefox 8), so I started to explore a little bit. It turns out, the computer the workers were working on is using a subdomain of localhost (and modified HOSTS file). I just tested it out as simply as I could think of, just adding localhost.com to the HOSTS file, and accessing the site through that and the workers are doing ok with it. I didn't create any virtual hosts or directories, just mapped localhost.com to 127.0.0.1. If you are running on Windows (or otherwise can do the same kind of mapping) you might want to give that a try as a temporary solution.
Edit: In response to chinmayv's comment: Something I noticed I have set in the HOSTS file is the IPv6 version of the local IP address, so there is both: ::1 localhost and 127.0.0.1 localhost, as well as the one to get the worker running on FF8 127.0.0.1 localhost.com. I don't know why this would matter, however. I just did a fresh install of Apache, so I don't think there is any setting that could account for the difference in behavior. Are you still getting the 'Could not get domain!' error? Maybe there is an absolute URL in the script somewhere that is throwing it off?