hugs/34

Yet another Selenium WebDriver library for Node...

34.js

Yet another Selenium WebDriver library for Node...

(Warning: This is a design doc... There's no code, yet! But when it lands, it'll most likely be based on webdriver-http-sync.

Goals:

• JavaScripty - This should feel like a JavaScript API, not a port from Java
• For automating web browsers: desktop and mobile
• Synchronous API - Async makes no sense for shell scripts or user interface test automation.
• Try to reduce locator strategy proliferation. Use CSS selectors as much as possible.
• Optimize for interactive exploration at the console.
• Yeah:
  • Pithy property names FTW
  • Easy discovery of properties via tab completion in the console.
  • e.g. thing.frobulated
• Boo:
  • e.g. driver.window().getFrobulationStatusOfThing(By.whyAmIStillTyping())

Non-Goals:

• Automating non-web native apps. (Although I hope this is temporary!)

---

Install

npm install 34

Example Usage

Driver = require('34').Driver
se = new Driver()

API

// Navigation
// ----------

// Open a URL
se.url = 'http://google.com'
se.open = function(url) { return this.url = url }
se.go = se.open

se.back()
se.forward()
se.refresh()

// Close current window
se.close()

// End current session
se.quit()


// Frame Navigation

// Switch focus to frame
se('#IdOfSomeFrame').focus()

// Switch focus to parent frame
se.parent()


// Window Navigation

// Switch focus to window
se.window = 'windowHandleId'



// Interrogation
// -------------

// Driver
se.url
se.title
se.source

// Server status
se.status

// Id of current session
se.session

// List of all sesions
se.sessions

// Handle id of current window
se.window

// List of all window handle ids
se.windows

// Id of active (focused) element
se.activeElement

// Find Element
e = se('#id')
// Find All elements
es = se('*[#id]')

// By name:
se('[name="theName"]')

// By id:
se('#theId')

// By tag name:
// all links
se('a')

// Element
e.attribute('name')
e.css('propname')
e.displayed
e.enabled
e.location
e.locationInView
e.name
e.selected
e.size
e.tag
e.text

// Find child element
e.find('#id')

// Find all child elements
e.find('#id')

// Alerts
se.alert.text


// Manipulation
// ------------

// Driver

// Keyboard
se.keys()
se.keyDown()
se.keyUp()

// Mouse
se.mouse.click()
se.mouse.doubleClick()
se.mouse.down()
se.mouse.moveBy(20,100)
se.mouse.moveTo(0,0)
se.mouse.up()

// Slim Jim (https://www.youtube.com/watch?v=KbneMYYI78Q)
se.touch.down()
se.touch.move()
se.touch.tap()
se.touch.up()

// Execute JavaScript
se.execute()
se.executeAsync()

// Wait for expression to return true or time out (milliseconds)
se.waitFor('return true', 10000)

// Take a Screenshot
se.screenshot()

// Element
e.clear()
e.click()
e.focus()    
e.keys()
e.submit()
e.scrollIntoView()

// Alerts
se.alert.accept()
se.alert.dismiss()
se.alert.focus()
se.alert.keys()



// Feng Shui
// ---------

// Get size of current window
se.size

// Set size of current window
se.size = [200, 200]

// Get position of current window
se.position

// Get position of current window
se.position = [200,200]

// Maximize size of current window
se.maximize()

// (For mobile...)
se.orientation
se.orientation = LANDSCAPE
se.orientation = PORTRAIT



// Synchronisation
// ---------------
se.timeouts.implicitWait = 10000
se.timeouts.pageLoad = 10000
se.timeouts.asyncScript = 10000

34's Issues

Synchronous API

Synchronous API - Async makes no sense for shell scripts or user interface test automation.

Do you mean Synchronous-looking API?

Sometimes elements aren't available right away, like a single page app that requires some data before UI elements animate in.

Protractor has wrapped webdriver to provide this and we've found it to be helpful. Example.

Looking forward to what you are cooking up.