Modifying an existing nightwatch.json configuration to expand usage

[Matthew Kidd]

I've recently taken over the test suite for a company that already got nightwatch "working". I'm trying to get to the point that the following 3 commands will all work without conflict:

docker-compose run --rm site1 -e localchrome -t ./tests/basic_search.js
nightwatch -e chrome -t ./tests/basic_search.js
nightwatch -t ./tests/basic_search.js

My thought process is as follows:

The docker command is what is current and works but it is completely headless unless the browser window is opening inside the docker container. The two nightwatch calls I would like to run locally by opening the browser, either specified by the environment flag or the default, so I can actually watch the execution locally with the hopes that would facilitate debugging in the browser.

I would like all this to work off just a single nightwatch.json file but I'm open to suggestions.

nightwatch.json:

{

"src_folders" : ["./tests"],

"output_folder" : "./reports",

"page_objects_path" : "./pages",

"custom_assertions_path": "./assertions",

"globals_path" : "",

"live_output" : false,

"parallel_process_delay" : 10,

"disable_colors": false,

"test_workers" : false,

"test_settings" : {

"default" : {

"launch_url" : "https://www.site1.com",

"selenium_host" : "ondemand.saucelabs.com",

"selenium_port" : 80,

"username" : "${USERNAME}",

"access_key" : "${ACCESS_KEY}",

"silent" : true,

"disable_colors": false,

"screenshots" : {

"enabled" : true,

"path" : ""

},

"desiredCapabilities" : {

"browserName" : "chrome",

"platform" : "Windows 10",

"version" : "53.0",

"deviceName" : "",

"javascriptEnabled" : true,

"acceptSslCerts" : true,

"avoidProxy" : true,

"screenResolution" : "1280x1024"

}

},

"chrome" : {

"use_ssl" : false,

"silent" : true,

"output" : true,

"screenshots" : {

"enabled" : true,

"on_failure" : true,

"path" : "screens/chrome"

},

"desiredCapabilities": {

"name" : "",

"browserName": "",

"platform": ""

},

"selenium" : {

"start_process" : false

}

},

"localchrome" : {

"screenshots": {

"enabled": true,

"on_failure": true,

"path": "screens/chrome"

},

"selenium_host" : "chromedriver",

"selenium_port" : 4444

}

}

}

docker-compose.yml:

version: '2'

services:

  chromedriver:

    image: selenium/standalone-chrome

    volumes:

      - /dev/shm:/dev/shm

  site1:

    build:

        context: ./

    depends_on:

      - chromedriver

    environment:

      - WAIT_FOR_HOSTS=chromedriver:4444

    volumes:

      - ./projects/site1:/home/node

  site2:

    build:

        context: ./

    depends_on:

      - chromedriver

    environment:

      - WAIT_FOR_HOSTS=chromedriver:4444

    volumes:

      - ./projects/site2:/home/node

  site3:

    build:

        context: ./

    depends_on:

      - chromedriver

    environment:

      - WAIT_FOR_HOSTS=chromedriver:4444

    volumes:

      - ./projects/site3:/home/node

If either file is confusing, as I attempted to obfuscate the websites, usernames, logins that are hard coded (I plan on moving some to a .env), I'm happy to provide clarification but it is accurate. I've spent a few days googling already but while I'm all the better for having done that I'm still confused. Any help is appreciated to help me understand this configuration.

[Matthew Kidd]

I've made progress. 

I've got the first and third commands "working" without conflict. I think I accomplished this by moving the settings for selenium_host, selenium_port, username, and access_key from default to the localchrome environment.

I'm still working on this but I believe I'm at a point where both test runs are hitting the site under test through a selenium instance be it local on my machine or what appears to be saucelabs.

Thanks for now.

[Eric Mumford]

Your desired capabilities for 'chrome' will fill in from 'default' if you don't specify them. You may be getting incompatible combinations of capabilities for 'chrome' if you are not re-specifying each capability.

[Matthew Kidd]

Yes I discovered this while editing the configuration. For example, chrome is defined as the default browser. I'm working towards a more modular and building upon itself set up.

[bharat joshi]

Hi ,

I need your help regarding the setup of docker for nightwatch , I am trying to run the tests with the same command as you have mentioned

I am getting an exception as "No such file or directory" for the scripts I want to execute

Can you please help me out with the same

[Matthew Kidd]

Turns out, after looking a little deeper into what I took over, the simplest thing to do was start fresh with a new nightwatch.js and get that working and then iterate.

[Eric Mumford]

What's your command, and what is your file structure.

Check to make sure you are running nightwatch from the right directory.

--
You received this message because you are subscribed to a topic in the Google Groups "NightwatchJs" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/nightwatchjs/HzgELssKGkQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to nightwatchjs+unsubscribe@googlegroups.com.
To post to this group, send email to nightw...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/nightwatchjs/f9360a51-42b2-4cf5-9ada-7e888de6b5e1%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.