Our backbone app uses gon. When we try to run our tests, we are getting a gon is undefined error in the console of the browser. Our layout file includes a call to include_gon, but that file is not being loaded by jasmine, so jasmine is failing in our first javascript file that contains gon. We tried creating a helper to assign the gon variable to an empty hash (like a fixture), but the helper was called after the first call to gon and therefore didn't fix our issue.
The secret was to use the asset pipeline to define the load order of my files. I commented out these lines from jasmine.yml
# path to parent directory of src_files
# relative path from Rails.root
# defaults to app/assets/javascripts
#src_dir: "app/assets/javascripts"
# list of file expressions to include as source files
# relative path from src_dir
#src_files:
# - "application.{js.coffee,js,coffee}"
And created spec.js.coffee with these lines:
#= require application
#= require jasmine-jquery
Now my js files get loaded in order and I am good to go.
What worked for me was to define window.gon = {} inside the test like so:
describe("A suite is just a function", function() {
beforeEach(function() {
window.gon = {}
gon.test = "this"
})
it ("should find gon", function() {
expect(gon.test).toBe("this")
})
})
Related
I am trying to serve a static HTML file using the res.sendFile() method from my express.js server that is hosted on AWS Lambda using the Serverless framework. Assuming that I am trying to serve an HTML file from the directory src/views/users/index.html.
In deployment, this is the file path that I have tried to serve my HTML file from /var/task/src/views/users/vindex.html, but I keep getting the error Error: ENOENT: no such file or directory, stat '/var/task/src/views/users/index.html' when viewing the AWS Cloudwatch log.
app.use(express.static(__dirname));
path.resolve(__dirname, "./src/views/users/index.html");
This is the results I get when I run tree src locally:
Have anyone experienced this issue before, and have solved it? Thank you so much!
Well, after many grueling hours and trying many solutions, I have found a workaround to render the html content without actually needing to render a .html file.
I ended up making a helper method that returns a string of the html content and send the html content string using the res.send() method instead.
html helper function
export const htmlHelper = () => {
return `<html content goes here>`;
}
route method
app.get('/html', (_, res) => {
const htmlString = htmlHelper();
return res.send(htmlString);
})
I have defined wdio.conf.js file (main file) and environment specific dev-chrome.conf.js file.
I can't get get cucumber to recognize my step definitions folder.
This is my structure:
And this is what I have in dev-chrome.config.js file:
const wdioConfig = require('../../../../../wdio.conf.js');
const commands = require('../../../../../src/commands/commands');
wdioConfig.config.cucumberOpts = [{
// other stuff here
require:
[
'./src/step_definitions/**/*.js',
// Or search a (sub)folder for JS files with a wildcard
// works since version 1.1 of the wdio-cucumber-framework
//'./src/**/*.js',
],
// other stuff here
}];
exports.config = wdioConfig.config;
I am getting an error:
"Step "When I add the product to a cart" is not defined. You can ignore this error by setting cucumberOpts.ignoreUndefinedDefinitions as true."
When I have same path for step definitions defined on main wdio.conf.js file then it works.
My main wdio.conf.js file is located in the root folder of the project.
Do you know how could I make it work in the environment specific conf.js file?
I am using #wdio/cucumber-framework": "^5.13.2"
As per the below example config, the cucumberopts should be an object and I think you are trying to set it as an array.
https://github.com/amiya-pattnaik/webdriverIO-with-cucumberBDD/blob/master/test/config/suite.cucumber.conf.js#L156
Maybe you should follow this example which will help to understand config setup.
Cheers!
Let's say I have modules similar to this one:
"use strict";
var moduleCss = require("module1.css");
var template = require("module1.hmtl");
module.exports = function(){
//my module code here
};
And another one with a different css file.
"use strict";
var moduleCss = require("module2.css");
var template = require("module2.hmtl");
module.exports = function(){
//my module code here
};
The main file would look like this:
"use strict";
var module1 = require("module1");
var module2 = require("module2");
var normalize = require("normalize.css");
var bootstrap = require("bootstrap.css");
module.exports = function(){
//my module code here
};
Instead of appending each css file content as a style tag, i'd like to get a main.css when building the main bundle by looking at each required css file in each module and building a single CSS file containing all the styles in all required modules so that I can append this single css file where I want.
I'm basically trying to have the css requirements in the same spot as the js and html requirements. The solutions that I found so far require a different CSS file that keeps track of each module css dependency, so adding or removing a module in my applications requires work in 2 files, the main js and the main css.
Is there a way to achieve this? To have ALL dependencies into one file? Or could a "package.json" thingy be used for each module where the CSS dependencies could be declared?
After playing around a little bit with browserify it turned out the transform function hook reveals the file name of each required file. I basically looked for css and less file, return a empty module for them, and kept each file in a reference array.
After the bundle was done, by listening to the bundle event, I created a string where I basically "included" each file that I captured in the reference array.
With the import file created, I simply passed it to the less compiler.
Here's a link to the gist for my current task : https://gist.github.com/vladnicula/fd1ff7b30ef20789e1dc
I got my less files compiled in css perfectly by grunt and I see result in .tmp/public/styles
So now livereload with grunt-contrib-watch should be made naturally in sails generated project ?
Or do I have to make a special configuration ?
I found that in tasks/pipeline.js file but not sure of what to do.
// CSS files to inject in order
//
// (if you're using LESS with the built-in default config, you'll want
// to change `assets/styles/importer.less` instead.)
var cssFilesToInject = [
'styles/**/*.css'
];
I saw in the file tasks/README.md :
###### `sails lift`
Runs the `default` task (`tasks/register/default.js`).
And in the file default.js we got :
module.exports = function (grunt) {
grunt.registerTask('default', ['compileAssets', 'linkAssets', 'watch']);
};
But watch.js file is missing in the folder...
What should it be ?
Watch does only looking for files that have changed and execute less, sass, injection and so on - but it doesn't make a reload.
You can add this in task/config/watch.js
Using rails 3 asset pipeline, I've structured the javascript (by using coffeescript) to files regarding the model. For example, all comment writing related javascript is stored to /app/assets/javascripts/comments.js.coffee, and user overlay related (fetching a:href's and triggering ajax on them) are stored in /app/assets/javascripts/users.js.coffee.
However, now I'm using more and more AJAX calls, where HTML content is pulled dynamically to the site. The problem is that I need to execute the javascript in various files, but as coffeescript is scoped inside a function, I can not access them.
Let's say that I've got a general.js.coffee file with following code
$(document).ready ->
# Parse all images with action
$("img.clickableaction").click ->
# Fetch some content
$.ajax
url: "something.php"
dataType: "html"
complete: (xhr, status) ->
# We got the content, set it up to a container
$("#somecontainer").html(receiveddata)
# The tricky part:
# run code in comments.js.coffee for #somecontainer (or for the whole dom)
# run code in users.js.coffee for #somecontainer (or for the whole dom)
And comments.js.coffee contains for example the following:
$(document).ready ->
commentDiv = $('div#commentsContainer')
commentsFetch = $('a.commentsFetch')
# Set commentid for comments fetch
commentsFetch.bind 'ajax:beforeSend', (event, xhr, settings) ->
# do stuff
The comments.js.coffee code works for the initial page view, e.g. the HTML code that was received when user loaded the page. But now, I need to parse the comments.js.coffee code for the content returned from the ajax call.
I can not make a function inside comments.js because it is scoped away, and can not be accessed from the general.js. This is what coffeescript produces:
(function() {
$(document).ready(function() {
var commentDiv, commentsFetch;
commentDiv = $('div#commentsContainer');
commentsFetch = $('a.commentsFetch');
}
})
I could make a global function for each separate file, e.g. window.comments && window.users, but then I'd need to call window.comments from the many places where I need to have ajax oncomplete call. On the long term, that will produce duplicate and hard to maintain -code.
How could something like this be made:
// comments.js
window.ajaxExecuteBlocks.push(function() { // comments related stuff });
// user.js
window.ajaxExecuteBlocks.push(function() { // user related stuff });
// general.js ajax on complete:
window.runExecuteBlocks()
Then, runExecuteBlocks would somehow run through all the functions that have been initialized in various controller-specific javascript files.
Anyone implemented similar system?