@applitools/driver
Advanced tools
Driver is a unified interface for automation frameworks. It allows you to use any automation framework with the same featureful api.
The API and abstractions of the interface provides an extra functionality which doesn't exists in some of the frameworks. All of these could be archived for any framework the only requirement is implementing a set of simple commands.
Except SpecDriver this package operates with 4 basic abstractions - Driver, Context, Element, Selector. It is important to understand those 4 abstractions for each framework before start with a SpecDriver.
Driver is the most important abstraction, since this is the first thing you would interact with, and it will create instances of other abstractions for you. It is helpful to think about Driver as controller of an automated environment (for example: browser in web automation, or device in native app automation).
So through the Driver you are gonna have a basic access to the automation environment, such as getting and setting viewport size, navigating, taking screenshots, and the most important having access to the page's/app's contexts.
Context is where all of the page's/app's content is. If you need to find an element, execute script, or click on a button - Context is what you need.
❓ FAQ
- Can I do all of this with a Driver?
No, you can't.
Because Driver doesn't contain any content in itself, instead it has access to the Contexts. Think about it as about awindowobject in a web page, it doesn't contain elements, but it hasdocumentobject in it, anddocumentis has access to the elements.
There always at leas one Context - main Context, however one web page could contain multiple Contexts, it is possible because each iframe on the page would create its own Context. So Contexts on the web page create a tree-like structure where root is a main Context and all its iframes are child Contexts, this structure repeats for each next context.
📝 NOTE
You could be familiar with a similar to Context concept from frameworks like Puppeteer and Playwright, where the same role plays
Frame.
Element is one of the simplest for understanding concepts. It is just an element that you could find on a web page, or native application (for example: <div> or View)
The access to the Elements is possible only through the Context.
Selector concept has two different types, first is always the same and the second is framework dependant. The first type called CommonSelector it is declared as part of the package and could be used with any king of framework. The second type doesn't have a one name it could be By instance for Selenium framework, Locator for Playwright, or just a string for WDIO.
SpecDriver is the interface we need to communicate with the framework api. It should consists of a number of commands each of those is a small function which performs one of the basic actions we could automate. The combination of commands is not the same for all of the frameworks, so it is important to understand which combination should be implemented in each specific case.
📝 NOTE
Decision about what to implement should be made in context of frameworks capabilities. For example frameworks like Playwright and Puppeteer have a dedicated method to set viewport size of web page, in the same time WDIO and Selenium don't have method like this, but instead they method to set browser window size.
The first step to implement a SpecDriver should be an understanding if your framework shares a Context concept, in other words, does it have a separation between something that would be a Driver and Context. If framework doesn't share a Context concept, its role will be played by the Driver.
❓ FAQ
- How to understand if my framework has a separation between Driver and Context?
It is fairly easy, in most cases it has. But there are one big exception, that are webdriver based frameworks. Webdriver protocol doesn't separate Context from a Driver, instead it maintains Driver state and automation inside iframes is performed by switching Driver to the iframe context.
In all other cases you would have some representation of a single Context (for example:Framein Playwright and Puppeteer, orDocumentin web api)
All SpecDriver commands could be divided on a two groups:
This command is a simple type guard that checks its argument on being a framework's driver.
This command is required.
API Reference | Playwright Example | WDIO Example
This command is a simple type guard that checks its argument on being a framework's context.
This command is required for web automation frameworks, that have a differentiation between driver and context.
API Reference | Playwright Example
This command is a simple type guard that checks its argument on being a framework's element.
This command is required.
API Reference | Playwright Example | WDIO Example
This command is a simple type guard that checks its argument on being a framework's selector.
This command is required.
API Reference | Playwright Example | WDIO Example
This command will be executed when a framework's driver is going to be wrapped. It should return a value of driver type.
This command is optional.
❓ FAQ
- When should I implement it?
This command should be implemented in two cases.
- Monkey patching or configuration of an object before we actually start using it.
- Supporting multiple types of drivers as input.
In the case of supporting multiple types you would still wanna work internally with only one of them, so you have to transform whatever could be accepted as an input to the type that you actually wanna work with.
This command will be executed when a framework's element is going to be wrapped. It should return a value of element type.
This command is optional.
❓ FAQ
- When should I implement it?
See answer for
spec.transformDriver.
This command will be executed when the selector is going to be used. The command should perform the transformation of common selectors to the framework's selector type. Some frameworks might also support shadow selectors, but by default, the package will not provide common shadow selectors, to tell the package that the framework supports it, shadowSelectors feature should be specified in spec.getDriverInfo(driver).
This command is required. The only exception is if framework selectors are identical to common selectors.
API Reference | Playwright Example | WDIO Example
❓ FAQ
- What is the common selector?
The common selector is a type of object to describe selectors, that is used by the package to describe different selector types. It is a simple object with a signature -
{type: string, selector: string}.- When framework's selectors could be identical to common selectors?
It could happen when a framework doesn't have real restrictions for selector type. For instance web api doesn't have a function that could find an element using different selector types, instead
documenthas different methods to find elements usingcssandxpathselectors. It means that common selector type could be used to decide what method should be called.
This command will be executed to get a context from a driver.
This command is required for web automation frameworks, that have a differentiation between driver and context.
API Reference | Playwright Example
This command will be executed to get a framework selector from framework element. Some frameworks keep metadata inside an element, this metadata might contain a selector that was used to find an element in first place. This is selector will be used in some edge case scenarios to re-find an element.
This command is optional.
This command is a simple type guard that checks if object is a framework stale element error. Not all of the frameworks share concept of stale element errors, but those which share could benefit of implementing this in some edge cases when element with known selector was stalled during an automation.
This command is optional.
API Reference | Playwright Example | WDIO Example
This command will be executed to check if two framework elements are equal. Some frameworks have shortcuts to compare elements, such as unique element ids or some specific backend functionality. Normally on web you can compare elements by simply compare them inside a browser, but native automation doesn't have this capability so custom logic has to be applied.
This command is required for native app automation frameworks, and optional for web automation frameworks but still recommended if there is a framework-specific way to compare elements.
❓ FAQ
- What if this method returns a false-negative result?
In webdriver based frameworks every element has a unique id by the specification, but in the real world some of the implementations violate this requirement, so sometimes id comparison would produce false-negative results. In order to avoid it for the web if
spec.isEqualElementsis implemented and returns a negative result, we will verify it with an in-browser element comparison.
This command will be executed to get a top-most (main) context of a given context. If the context is already a top-most (main) one it should return itself.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
❓ FAQ
- What should I do in webdriver based framework?
In webdriver based frameworks this command would rather change state of the driver (a.k.a switch to the frame). After state is changed driver object should be returned.
This command will be executed to get a parent context of a given context. If the context is already a top-most (main) one it should return itself.
This command is optional but highly recommended for implementation, because the fallback is very time-consuming for large pages.
API Reference | Playwright Example | WDIO Example
❓ FAQ
- What should I do in webdriver based framework?
See answer for
spec.mainContext(context).
This command will be executed to get a child context of a given context for a given iframe element.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
❓ FAQ
- What should I do in webdriver based framework?
See answer for
spec.mainContext(context).
This command will be executed to run a certain script with a specific argument in a given context.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
This command will be executed to get an element by the given selector in a given context if parentElement is provided it means that the seeking element should be located as a descendant of provided parent element. As a parent element also might be provided document fragment node. If an element could not be found null should be returned as a result.
This command is required.
API Reference | Playwright Example | WDIO Example
This command will be executed to get elements by the given selector in a given context if parentElement is provided it means that the seeking elements should be located as a descendants of provided parent element. As a parent element also might be provided document fragment node. If an elements could not be found empty array should be returned as a result.
This command is required.
API Reference | Playwright Example | WDIO Example
This command will be executed to get the size of the browser window or device screens in web or native automation, respectively.
This command is required for web and native automation frameworks, but for web frameworks it is required only if spec.setWindowSize(driver) is implemented.
This command will be executed to set size of the browser window.
This command is required for web automation frameworks only if spec.setViewportSize(driver) is not implemented.
This command will be executed to set the size of the browser viewport (the area where a webpage is actually rendered).
This command is optional but highly recommended if a framework provides a simple way to get this value.
API Reference | Playwright Example
This command will be executed to set the size of the browser viewport (the area where a webpage is actually rendered).
This command is required for web frameworks and should be implemented instead of spec.setWindowSize(driver) if possible.
API Reference | Playwright Example
This command will be executed to get cookies stored in the browser or a given context. The main purpose of this command is to provide cookies of the browser, but since this is not always possible the fallback for extraction cookies for a particular context exists.
This command is optional but highly recommended for web automation frameworks.
API Reference | Playwright Example | WDIO Example
❓ FAQ
- How should I know to extract cookies from the whole browser or context?
It is always preferable to get cookies of the browser because it is a superset of the context's cookies. But it is possible only in some cases, so if the package thinks that this is not possible in a current environment, it will pass context object as a first argument and
truevalue as a second in other cases driver will be passed as first argument andfalseas a second. If the framework has a possibility to extract browser's cookies in any environment thisallCookiesfeature should be overridden inspec.getDriverInfo(driver).
This command will be executed to get basic info about the driver and the current environment. Usually, the package is able to understand all important information about a current environment using the information provided by spec.getCapabilities(driver) and user agent, but sometimes additional information about the capabilities of the framework should be provided. For instance features field contains a map of supported framework features.
| Feature | Description |
|---|---|
shadowSelector | This feature means that framework is capable of finding elements under the shadow root using a specific selector. It also means that spec.transformSelector(selector) should be able to transform a recursive version of common selector - {type: string, selector: string, shadow?: CommonSelector}. |
allCookies | This feature means that framework is capable of extracting cookies from the whole browser. Usually, this feature is calculated on the package side by checking if the browser is a Chrome browser. |
This command is optional and should be implemented only to override the default behavior of the package or provide useful metadata.
API Reference | Playwright Example | WDIO Example
This command will be executed to get webdriver capabilities of the session.
This command is required for webdriver based frameworks.
This command will be executed to get title of current page.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
This command will be executed to get url of current page.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
This command will be executed to take a screenshot of a current state of the driver.
This command is required for web automation frameworks.
API Reference | Playwright Example | WDIO Example
This command will be executed to perform click automation on the area of a given element.
This command is optional.
API Reference | Playwright Example | WDIO Example
This command will be executed to automate typing a given value into a given element.
This command is optional.
API Reference | Playwright Example | WDIO Example
This command will be executed perform a navigation to a given url address.
This command is optional.
API Reference | Playwright Example | WDIO Example
TBD
TBD
TBD
TBD
TBD
TBD
FAQs
Applitools universal framework wrapper
The npm package @applitools/driver receives a total of 390,898 weekly downloads. As such, @applitools/driver popularity was classified as popular.
We found that @applitools/driver demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 57 open source maintainers collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
Research
Socket researchers have discovered malicious npm packages targeting crypto developers, stealing credentials and wallet data using spyware delivered through typosquats of popular cryptographic libraries.
Security News
Socket's package search now displays weekly downloads for npm packages, helping developers quickly assess popularity and make more informed decisions.
Security News
A Stanford study reveals 9.5% of engineers contribute almost nothing, costing tech $90B annually, with remote work fueling the rise of "ghost engineers."
