saving HTML source code for debugging in nightwatch.js - automated-tests

Suppose, I have a failing test in only one of the browsers, where one of the css/html elements is "not found".
Screenshot looks okay. How can I get the html source of the current document to check?

Using Nightwatch's wrapper for selenium function source (DOC), this can be accomplished relatively easily.
For instance:
browser
.url("http://www.google.com")
.source((result) => {
// Source will be stored in result.value
console.log(result.value);
})

Related

Print existing pdf file directly to client default printer [duplicate]

A coworker and I were having a discussion about what is and isn't possible within the browser.
Then a question came up that neither of us could answer with certainty.
Can you create a webpage such that when you navigate to it, it engages the client-side printer and attempts to print a document. For instance, whenever you visit my personal website, you'll be treated to a print out of a picture of me, smiling.
Now, this is a hideous idea. I'm aware. But the discussion intrigued me as to if it could be done, and how. My friend insisted that the best you could do was pop up the print dialog for the user, they would have to click print themselves.
Would it be possible to bypass this step? Or just some fancy script to move the mouse over the print button and click on it? Or use an activeX control to interface with a Printer API directly?
You have to prompt the user to print the current page, there's no way to bypass this step (possibly in activeX for IE). That said, there's two different ways you could prompt the user to print images of you smiling when the page is loaded.
Here's how to do it in JavaScript.
window.onload = function() {
var img = window.open("me-smiling.png");
img.print();
}
And here's how to do it in css/javascript/html (assuming your picture has the id 'me-smiling'):
CSS:
#media print {
* {
display:none;
}
img#me-smiling {
display:block;
}
}
Javascript:
window.onload = function() { window.print() }
The only solution to avoid print dialog that I found was creating a variable on Mozilla Firefox to set auto-print. Maybe is not the best solution if you need to use other browser, but in my case, I only need to print a report automatically and it works:
1- Open Firefox and type "about:config" in the address bar
2- Right click on any preference and select "New" > "Boolean"
3- Add a variable called "print.always_print_silent" with "true" value
4- Restart Firefox.
Hope help you!
AttendStar created a free add-on that suppresses the dialog box and removes all headers and footers for most versions of Firefox.
https://addons.mozilla.org/en-US/firefox/addon/attendprint/
With that feature on you can use $('img').jqprint(); and jqprint for jquery will only print that image automatically called from your web application.
As far as I know, there is no way to print a document directly, without some client intervention, like setting browser flags.
In our current project we need to print directly to the default printer, but at least with Chrome you can do it easily with additional startup arguments.
To print directly to the OS default printer you can use:
"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --user-data-dir=c:\tmp --kiosk-printing http://www.contoso.com
Another option, which may also be useful, is tos use the native print dialog instead of chromes print preview.
"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --user-data-dir=c:\tmp --disable-print-preview http://www.contoso.com
Note, that window.print() and/or Ctrl-P behave accordingly the mentioned settings.
I know, that this does not exactly answers the OPs question, but I think it somewhat related, and for web based enterprise applications this is a quite common use case. Maybe someone find it useful.
For Firefox I recommend Seamless Print Addon
You can't bypass the print dialog, as far as I know. That would be a pretty obvious security flaw if the browser allowed that. But you can bring up the print dialog with "window.print()".
I think at best you would need an ActiveX component using base windows API to obtain a device context for the default printer and try and print an embedded image using assumed values for the printer settings.
To print to the default printer automatically without seeing a print dialog prompt, I've shared some code in the following question that works in IE7, IE8 and IE9:
Bypass Printdialog in IE9
From lot of search from last few days,
I've found a best possible solution.
Till date Chrome do not support direct printing from javascript.
It has launched USB and serial API which might help.
But currently I'm using a JavaApplet solution which is open source.
https://github.com/qzind/qz-print - build
While I'm getting error in building it. I preferred a Prebuilt - QZ Print Plugin 1.9.3
desktop app, which works great.
Download it from here: https://qz.io/download/
Code Example:
/***************************************************************************
* Prototype function for printing an HTML screenshot of the existing page
* Usage: (identical to appendImage(), but uses html2canvas for png rendering)
* qz.setPaperSize("8.5in", "11.0in"); // US Letter
* qz.setAutoSize(true);
* qz.appendImage($("canvas")[0].toDataURL('image/png'));
***************************************************************************/
function printHTML5Page() {
$("#qz-status").html2canvas({
canvas: hidden_screenshot,
onrendered: function() {
if (notReady()) { return; }
// Optional, set up custom page size. These only work for PostScript printing.
// setPaperSize() must be called before setAutoSize(), setOrientation(), etc.
qz.setPaperSize("8.5in", "11.0in"); // US Letter
qz.setAutoSize(true);
qz.appendImage($("canvas")[0].toDataURL('image/png'));
//qz.setCopies(3);
qz.setCopies(parseInt(document.getElementById("copies").value));
// Automatically gets called when "qz.appendFile()" is finished.
window['qzDoneAppending'] = function() {
// Tell the applet to print.
qz.printPS();
// Remove reference to this function
window['qzDoneAppending'] = null;
};
}
});
}
Complete example can be found here:
https://gist.github.com/bkrajendra/c80de17b627e59287f7c
This is the best solution that I have found for firefox:
There is this awesome add-on Seamless Print.
It works like charm.

Meteor: [Load Order] Make JavaScript file load after HTML file is loaded?

Load Order Issues
I am having trouble making Meteor load my JavaScript after my HTML file fully loads when I go to localhost:3000. The problem is that my JavaScript keeps loading before my HTML file, and makes the page look unloaded when I use stuff like alert(); or prompt();. I've tried a lot of solutions such as naming my JavaScript file as main.js and putting my HTML file in a deeper directory and using <script> tags. I have also read the documentation concerning this: http://docs.meteor.com/#/full/structuringyourapp Solutions I've tired based off the documentation such as putting files in client/lib , client/compatibility , and lib have proven to no avail. I also tired Meteor.startupand I placed the file for it in the client folder. (The code inside it):
Meteor.startup( function () {
$.get("client/lib/testproject.html")
$.getScript("client/testproject.js");
});
The above sort of solved my problem, but it loaded the JavaScript file two times. The first time was before the HTML loaded and the second time was after the HTML loaded. I don't know a way to prevent the first JS load from happening when using Meteor.startup, so any solutions for that are also appreciated.
The JavaScript file's code I am referring to is simple. (In it's entirety):
prompt("Hello World!");
myList = ["apples", "oranges", "bananas"];
myList.forEach(function(value, index) {
alert('I have ' + value + ' for dinner.');
});
Summary
To summarize my problem:
My Problem:
Go to localhost
JavaScript loads first
HTML loads second
What I Need:
Go to localhost
HTML loads first
JavaScript loads second
The Question: How can I make my JavaScript load only after when my HTML is loaded? And how can I restructure my folder, file-names, and/or code to make it behave as I want it to in this case?
Since the code posted is extremely simple to reproduce I kindly ask that you
run your own solution with a setup similar to what I have and not something that uses a million packages since that is unnecessary for my case, on Meteor, before responding to this.
I'm on Meteor 1.1.0.2
Here is a link to my folder structure with included HTML code along with filenames I used: http://i.imgur.com/24z6bXF.png
I think you missed a decisive information : you should wrap your Javascript code into a Template.yourTemplate.rendered=function () {} function.
That is the meteor way to ensure that your related html code is properly rendered first.
First of all, Meteor will always repackage your files and load them automatically in a specific order (Meteor structuring your app). First files in client/compatibility then client/lib and then the others JS files.
You should also rewrite your code so it does not get executed immediately at load time, like wrapping everything in a function. And then, you should call this code when the DOM is loaded, which does not necessarily mean in Meteor.startup but also in onRendered callbacks in your templates.

Retrieve HTML source from webpage

I am new to Rust and trying out different projects to become more familiar with the language. As the title says I would like to retrieve the html source of a webpage. I am aware of rust-http but I am not sure how to use that library for this purpose
For a more detailed description of what I am trying to do:
Given some url: www.google.com
I would like the underlying HTML source.
I have looked at the Github Documentation of rust-http but the lack of docs have been confusing.
(answer moved from an edit to the question)
After playing around/searching more I finally was able to get what I wanted. Here is some code using Hyper that retrieves the source.
let req = request::Request::get(hyper::Url::parse("www.google.com").unwrap()).unwrap();
let res = req.start().unwrap()
.send().unwrap()
.read_to_string().unwrap();
println!("Response: {}", res);
What is wrong with using the options built into the browsers to display source code? E.g Chrome and Firefox: right click and select "View Page Source".

How to output clickable links in test tables in the latest FitNesse version '20140201'

I want to display in a FitNesse test table some clickable links. So I had created some FitNesse fixture methods that return HTML links. This worked fine in FitNesse version release 20110104
Like for example:
public String getCanonical() {
return "" + mCanonical + "";
}
However after upgrading the FitNesse server to the latest release it does no longer render the HTML tags. Is it possible to enable rendering of HTML tags again in the latest FitNesse version 20140201 ?
I have a lot of data-driven tests where outputing some clickable links was really convenient while reviewing test results, because in case of doubts I could quickly jump to the tested URL.
I already tried using FitNesse Link formatting but this does not seem to work inside the test tables.
Any solutions or workarounds for this? I am tempted to rollback FitNesse to the old version if I cannot find a solution for this. (old version works all fine for me)
What I found out is that FitNesse is picky on what HTML elements it allow to go into the output HTML table directly and what it escapes (a regular expression is used, based on the HTML spec I'm told). Looking at the code it does not allow <a.. ></a> so this would explain your problem.
My fixtures output HTML in various situations and that works great, as long as I enclose the HTML elements in a <div> </div> pair. So for your method it would be:
public String getCanonical() {
return "<div>" + mCanonical + "</div>";
}

Why is my javascript function not found by the page it is embedded in?

I have a page that has a simple javascript in the header portion of the page:
<script type="text/javascript">
function doLogout() {
var conf = confirm("Really log out?");
if (conf === true) { //changed == to === for boolean comparison
$.post("logout.aspx");
}
}
</script>
It uses jQuery to do an AJAX post to my logout page. The only issue right now is that when I click on the link (logout) to fire this function, nothing happens. I checked FireBug's console, and it told me that the function is not defined. This has happened to me before, but I think I botched a bunch of code to fix it sometimes.
Does anyone know the proper way to fix this issue?
Edit
After doing a lot of googling and trying different things, I found this very concise and informative post. Apparently, as the linked article states, the way the script is referenced in the web site is important as it won't run properly otherwise! Hopefully this information will be useful for more people.
This can also occur if there is a syntax error earlier in your javascript code. Often this will just be interpreted as the function not existing (nor any function AFTER the error). Check the code above this code (if there is any) and this code for syntax errors.
A way to tell if the cache error is it is to open Firebug and view the Script source. If the page was cached, you won't see your code. If it loaded but has syntax errors, the code will show, though it won't "find" it.
Things to test:
1) Can you call this function from something else? Like add a <script> at the bottom of the page to call it?
2) Does the page validate? Sometimes I get screwy javascript errors if there is some busted HTML like a missing </b>
3) I've been starting to wrap my javascript in <![CDATA[ ]]> just incase I've got goofy chars in my javascript.
4) I assume you've tested this in other browsers and have the same behavior, right?
5) If you haven't installed it already, install the Web Developer firefox addon. It has a nifty toolbar menu that will disable the cache for you so everything reloads.
6) As weird as it sounds, I once hit a javascript issue that was because of how my text editor was saving UTF-8 files. I forget the details, but it was adding some byte-order-mark or something that upset the browser.
I've had this occur when the page had been cached and so it didn't load the new script in. So to fix it clear all private data from Firefox. Not sure if that helps but it sure happened to me a bunch.
Other ideas for you to test:
is the function defined in the DOM tab in FireBug?
if you call doLogout() from the FireBug console, what happens?
I assume this is not the only script on that page. Make sure that some later script is not modifying doLogout to something else
I had the same issue and tried all that's been suggested here without success.
The only way I fixed it was by discovering that in the <script src="jquery.js"> tag I was using in the head of the page I forgot to close it with its </script> causing the page to ignore all Javascript functions. So please check that your includes look like:
<script src="jquery.js"></script>
I hope that helps. Ross.
If you are using DevExpress controls these links may help you: How to register and execute a JavaScript downloaded to the client via a callback and How to register and execute a JavaScript downloaded to the client via a callback (standalone JS file) and Executing javascripts from user controls dynamically created through ASPxCallback panels
The issue might occur if you have NoScript. You should check and make sure it's not blocking said script.
I had this issue and discovered the problem was just a wrong case letter inside the name.
Call: filterCheckbox()
vs
function filterCheckBox() {}
problem: lowercase "box" vs uppercase "Box".
So check if the name is exactly the same.

Resources