Does Python have anything Like Capybara/Cucumber?
Question:
Ruby has this great abstraction layer on top of Selenium called Capybara, which you can use do functional/acceptance/integration testing. It also has another library called Cucumber which takes this a step further and lets you actually write tests in English.
Both libraries are built on top of Selenium, and can be used to test against any major browser, but because of their abstraction layers it’s super easy to write tests using them (well, as easy as functional testing gets at least).
My question is: does Python have anything like that? I have found Pythonistas doing functional testing with various tools but …
A) Splinter: doesn’t use Selenium (and doesn’t have an IE driver)
-EDIT-
It appears Spliter now does use Selenium (see answers below).
B) Alfajor: hasn’t been updated in over a year; looks dead
C) Selenium (raw): a lot of people seem to be using Selenium directly, but it seems like an abstraction layer could make it a lot easier to use
So, does anyone know of anything Capybara-like, or better yet Cucumber-like, for Python (it doesn’t have to actually use Selenium, but it needs to support all major browsers)?
* EDIT *
For those that aren’t familiar with Capybara, it basically just adds an API so that instead of the normal Selenium API you can do something like this:
When /I sign in/ do
within("#session") do
fill_in 'Login', :with => '[email protected]'
fill_in 'Password', :with => 'password'
end
click_link 'Sign in'
end
It’s used by Cucumber, which let’s you further abstract (almost to English):
Scenario Outline: Add two numbers
Given I have entered <input_1> into the calculator
And I have entered <input_2> into the calculator
When I press <button>
Then the result should be <output> on the screen
Examples:
| input_1 | input_2 | button | output |
| 20 | 30 | add | 50 |
I would LOVE a Python Cucumber equivalent, but even just a Capybara equivalent would be helpful.
Answers:
You can test Python code using Cucumber – see the Cucumber wiki on github for more information.
If you want a pure Python solution, check out Lettuce. I’ve never used it, but there is a fairly useful looking blog entry about it and splinter here.
Have you checked freshen, or pea?
Pea does not use the syntax of cucumber, but the author says that is easier
https://github.com/gfxmonk/pea
And Freshen is trying to clone Cucumber’s syntax and functionalities
While the OP was happy with finding a Python Cucumber equivalent, what led me here was the question title: a Python equivalent of Capybara. While Cucumber uses Capybara, Cucumber itself is a whole different “solution” that is only incidentally related to Capybara.
If you’re looking for something Capybara-like without having to deal with Cucumber, check out splinter. I don’t know what was true when the question was posted, but Splinter is now built on Selenium, and supports other engines as well (Webkit, PhantomJS, zope.browsertest, and others), and supports both visual and headless testing.
How about Robot Framework. It’s pretty awesome. And with Selenium2Library it works really well with SE2. http://robotframework.org/
The OP asked for Python implementations of Cucumber or Capybara but as Jim Stewart pointed out in his answer, Cucumber and Capybara are very different things. Since the title of the question is about Capybara, that’s what I will answer.
I am one of the developers of a commercial Selenium wrapper called Helium. Like Capybara, it offers a very high-level API for web automation. For example, here is a script that updates your Facebook status:
from helium.api import *
start_chrome("facebook.com")
write(your_fb_email, into="Email or Phone")
write(your_fb_password, into="Password")
click("Log In")
write("Test", into="Update Status")
click("Post")
Calls to Helium can freely be mixed with calls to Selenium. Eg. we could extend the above script by:
# get_driver() returns the WebDriver created by start_chrome() above.
chrome = get_driver()
chrome.find_element_by_id('btnG').click()
A. Cucumber like: ( English like)
- Lettuce (Approach Gherkin) or
- Behave (Approach Gherkin) or
- Robotframework (Approach keyword based)
(Additional Info: RF is larger than English Like criteria. Its keyword-based and offers loads of helper method and inbuilt libraries. Great eco-sysstem for external libraries. Any python script can also be modified and used along with RF)
- Freshen (Approach Gherkin) or
- Pea (Approach Gherkin) or
-
RedwoodHQ (Approach keyword based)
(RedwoodHQ has features larger than ‘English-Like’ criteria and encapsulates the following features: Keyword based, web-based test-framework, supports python as one of languages and much more.
Additional info on RedwoodHQ: Theoretically its possible, all the existing robot-framework inbuilt libraries and all robot-framework external test-libraries or for that matter any python library, can be called or used from this web based test framework with little modification)
-
Gauge (Gherkin approach): Reference for python: (https://gauge-python.readthedocs.io/en/latest/index.html)
Underneath Cucumber, one could have Capybara like abstraction layer that hides/ groups many of selenium actions
B. Capybara like: ( Abstraction: hides / groups action)
As one e.g to click an element, its sufficient to provide command like click(locator) instead of working with raw selenium api, where one needs to find an element and then click. Many more such abstraction exist in the optional libraries below
- Option-1 (see below)
- Option-2 (see below)
- Option-3 (see below)
- Option-4 (see below)
- Option-5 (see below)
- Option-6 : Helium (from others Answer/comment)
- Option-7 : (see below)
- Option-8 : (see below)
- Option-9: (see below)
- Option-10: (see below)
- Option-11: (see below)
- Option-12: (see below)
- Option-13: (see below)
My research: There exist almost half a dozen a. active, b. mature c.developed options.
python comes with variety of batteries included !!
Option-1: Selenium2Library
Github url: https://github.com/rtomac/robotframework-selenium2library
Developement:
Active
Purpose:
one of the many libraries of robotframework, can also be used as “standalone” library for your framework (Check e.g below for usage).
Thoughts:
- It provides an abstraction over selenium
- input of arguments to methods in this library are lot simpler. Abstraction provided by library, one e.g, hides many of the unnecessary details for finding elements. For more details one needs to understand the library
- It is possible to use this library outside robot-framework context as such without any modification, though it could be using utilities of robot package. (Its your homework to do further experiments with this lib, on this note!)
- Therefore, can be used as a standalone library for your framework.
Usage:
pip install robotframework-selenium2library
import in your ipython or idle console and start playing
e.g:
>>from Selenium2Library import Selenium2Library
>>start_testing= Selenium2Library()
>>start_testing.create_webdriver("Firefox")
>>start_testing.go_to("http://www.google.com")
>>.
...so on
Option-2: Pageobjects
Github url:
https://github.com/ncbi/robotframework-pageobjects
Development:
InActive (no show stoppers with latest release)
Purpose:
One of the libraries of robotframework. Provides a page object abstraction over Selenium2Library. Can be used as Standalone for your framework (Check e.g below for usage) or can be used along with robotframework.
Thoughts:
- It provides a “pageobject abstraction” support on top of Selenium2Library
- It is possible to use this library outside robot-framework context as such without any modification, though it could be using utilities of robot package. (Its your homework to do further experiments on this note!)
- Therefore, can be used as a standalone library
Usage:
pip install robotframework-pageobjects
e.g:
in ipython or idle do:
>>from robotpageobjects import Page
>>start_testing=Page()
>>start_testing.create_webdriver("Firefox")
>>start_testing.go_to("http://google.com")
Option-3: robotframework-pageobjectlibrary
Github Url:
https://github.com/boakley/robotframework-pageobjectlibrary
Development:
Active
Hopefully Author supports LTS (Long Term Support) : )) , Fingers crossed !!
Usage:
pip install robotframework-pageobjectlibrary
Thoughts:
- It is NOT possible to use this library outside robot-framework context. A minor change to how the page context is handled would help this library to be used outside robot-framework context (Its your homework to find out how!)
Option-4: Splinter
Github url:
https://github.com/cobrateam/splinter
Developement:
Active
Usage:
splinter.readthedocs.org/en/latest/index.html
pip install splinter
On ipython or idle do:
>>from splinter import Browser
>>browser = Browser()
>>browser.visit('http://google.com')
>>browser.fill('q', 'splinter - python acceptance testing for web applications')
>>browser.find_by_name('btnG').click()
Option-5: SST library
Github url:
https://github.com/Work4Labs/selenium-simple-test
Development:
Feature complete / Active
Usage:
testutils.org/sst/
pip install -U sst
on ipython or idle do:
>>> from sst.actions import *
>>> start()
Starting Firefox
>>> go_to('http://google.com')
Going to... http://google.com
Waiting for get_element
Option-6: helium
Not open source (Commercial)
Option-7: holmium.core
Github url:
https://github.com/alisaifee/holmium.core
Option-8: wtframework
Github url:
https://github.com/wiredrive/wtframework
Option-9: webium
Github url:
https://github.com/wgnet/webium
Option-10: elementium
Github url:
https://github.com/actmd/elementium
Option-11: saunter
Github url:
https://github.com/Element-34/py.saunter
Usage:
saunter
Option-12: webdriverplus
Github url:
https://github.com/tomchristie/webdriverplus
Usage:
webdriverplus
Comments:
repository not maintained but decent reference
Option-12: Simple-Pageobject
Github url:
https://github.com/rama-bornfree/simple-pageobject/tree/master/PageObjectLibrary
Comments:
Simplest pageobject wrapper built around selenium2library. I am the owner of the repo
Test-setup:
“All” the Test libraries in Option-1-13; can be run using any of the following framework: Lettuce, Behave, Robotframework or for that matter, any unit test framework(e.g PyUnit, Nose)…so on .
Test Framework is in general used to manage test-cases e.g
- English formats like gherkin, keyword, tabular so on …
- reporting a test-run
- hooking to CI
- set-up/tear-down of test-cases and test-suites
- tagging of test-cases
- other functionalities that one could think on about any test-framework
What matters is how comfortable one gets with libraries in above options.
Option-5: As far as SST is concerned, it has features of a framework itself, e.g it can generate report and do many more things.
So the definition of library and framework in the case of SST is blurred, depending on the extent of features one would like to use from that package
Some Math for Fun:
Total number of ways one could have a good, bad, and ugly Test-setup = (Test framework AND Test library + your custom code sandwiched b/w the framework and library ):
7 * 13 = 91 Ways
Choose the best combination (of Test Framework And Test library) that suits ones need !!
I would personally go for Robot-framework with Selenium2Library or Robot-framework with some pageobject library
ofcourse, I am leaned and positively biased in my post about robot-framework and Selenium2Library
There does now exist a port of Capybara itself to Python:
https://github.com/elliterate/capybara.py
You can find its documentation here:
https://elliterate.github.io/capybara.py/
Capybara helps you test web applications by simulating how a real user would interact with your app. It is agnostic about the driver running your tests and comes with Selenium support built in.
Ruby has this great abstraction layer on top of Selenium called Capybara, which you can use do functional/acceptance/integration testing. It also has another library called Cucumber which takes this a step further and lets you actually write tests in English.
Both libraries are built on top of Selenium, and can be used to test against any major browser, but because of their abstraction layers it’s super easy to write tests using them (well, as easy as functional testing gets at least).
My question is: does Python have anything like that? I have found Pythonistas doing functional testing with various tools but …
A) Splinter: doesn’t use Selenium (and doesn’t have an IE driver)
-EDIT-
It appears Spliter now does use Selenium (see answers below).
B) Alfajor: hasn’t been updated in over a year; looks dead
C) Selenium (raw): a lot of people seem to be using Selenium directly, but it seems like an abstraction layer could make it a lot easier to use
So, does anyone know of anything Capybara-like, or better yet Cucumber-like, for Python (it doesn’t have to actually use Selenium, but it needs to support all major browsers)?
* EDIT *
For those that aren’t familiar with Capybara, it basically just adds an API so that instead of the normal Selenium API you can do something like this:
When /I sign in/ do
within("#session") do
fill_in 'Login', :with => '[email protected]'
fill_in 'Password', :with => 'password'
end
click_link 'Sign in'
end
It’s used by Cucumber, which let’s you further abstract (almost to English):
Scenario Outline: Add two numbers
Given I have entered <input_1> into the calculator
And I have entered <input_2> into the calculator
When I press <button>
Then the result should be <output> on the screen
Examples:
| input_1 | input_2 | button | output |
| 20 | 30 | add | 50 |
I would LOVE a Python Cucumber equivalent, but even just a Capybara equivalent would be helpful.
You can test Python code using Cucumber – see the Cucumber wiki on github for more information.
If you want a pure Python solution, check out Lettuce. I’ve never used it, but there is a fairly useful looking blog entry about it and splinter here.
Have you checked freshen, or pea?
Pea does not use the syntax of cucumber, but the author says that is easier
https://github.com/gfxmonk/pea
And Freshen is trying to clone Cucumber’s syntax and functionalities
While the OP was happy with finding a Python Cucumber equivalent, what led me here was the question title: a Python equivalent of Capybara. While Cucumber uses Capybara, Cucumber itself is a whole different “solution” that is only incidentally related to Capybara.
If you’re looking for something Capybara-like without having to deal with Cucumber, check out splinter. I don’t know what was true when the question was posted, but Splinter is now built on Selenium, and supports other engines as well (Webkit, PhantomJS, zope.browsertest, and others), and supports both visual and headless testing.
How about Robot Framework. It’s pretty awesome. And with Selenium2Library it works really well with SE2. http://robotframework.org/
The OP asked for Python implementations of Cucumber or Capybara but as Jim Stewart pointed out in his answer, Cucumber and Capybara are very different things. Since the title of the question is about Capybara, that’s what I will answer.
I am one of the developers of a commercial Selenium wrapper called Helium. Like Capybara, it offers a very high-level API for web automation. For example, here is a script that updates your Facebook status:
from helium.api import *
start_chrome("facebook.com")
write(your_fb_email, into="Email or Phone")
write(your_fb_password, into="Password")
click("Log In")
write("Test", into="Update Status")
click("Post")
Calls to Helium can freely be mixed with calls to Selenium. Eg. we could extend the above script by:
# get_driver() returns the WebDriver created by start_chrome() above.
chrome = get_driver()
chrome.find_element_by_id('btnG').click()
A. Cucumber like: ( English like)
- Lettuce (Approach Gherkin) or
- Behave (Approach Gherkin) or
- Robotframework (Approach keyword based)
(Additional Info: RF is larger than English Like criteria. Its keyword-based and offers loads of helper method and inbuilt libraries. Great eco-sysstem for external libraries. Any python script can also be modified and used along with RF) - Freshen (Approach Gherkin) or
- Pea (Approach Gherkin) or
-
RedwoodHQ (Approach keyword based)
(RedwoodHQ has features larger than ‘English-Like’ criteria and encapsulates the following features: Keyword based, web-based test-framework, supports python as one of languages and much more.
Additional info on RedwoodHQ: Theoretically its possible, all the existing robot-framework inbuilt libraries and all robot-framework external test-libraries or for that matter any python library, can be called or used from this web based test framework with little modification) -
Gauge (Gherkin approach): Reference for python: (https://gauge-python.readthedocs.io/en/latest/index.html)
Underneath Cucumber, one could have Capybara like abstraction layer that hides/ groups many of selenium actions
B. Capybara like: ( Abstraction: hides / groups action)
As one e.g to click an element, its sufficient to provide command like click(locator) instead of working with raw selenium api, where one needs to find an element and then click. Many more such abstraction exist in the optional libraries below
- Option-1 (see below)
- Option-2 (see below)
- Option-3 (see below)
- Option-4 (see below)
- Option-5 (see below)
- Option-6 : Helium (from others Answer/comment)
- Option-7 : (see below)
- Option-8 : (see below)
- Option-9: (see below)
- Option-10: (see below)
- Option-11: (see below)
- Option-12: (see below)
- Option-13: (see below)
My research: There exist almost half a dozen a. active, b. mature c.developed options.
python comes with variety of batteries included !!
Option-1: Selenium2Library
Github url: https://github.com/rtomac/robotframework-selenium2library
Developement:
Active
Purpose:
one of the many libraries of robotframework, can also be used as “standalone” library for your framework (Check e.g below for usage).
Thoughts:
- It provides an abstraction over selenium
- input of arguments to methods in this library are lot simpler. Abstraction provided by library, one e.g, hides many of the unnecessary details for finding elements. For more details one needs to understand the library
- It is possible to use this library outside robot-framework context as such without any modification, though it could be using utilities of robot package. (Its your homework to do further experiments with this lib, on this note!)
- Therefore, can be used as a standalone library for your framework.
Usage:
pip install robotframework-selenium2library
import in your ipython or idle console and start playing
e.g:
>>from Selenium2Library import Selenium2Library
>>start_testing= Selenium2Library()
>>start_testing.create_webdriver("Firefox")
>>start_testing.go_to("http://www.google.com")
>>.
...so on
Option-2: Pageobjects
Github url:
https://github.com/ncbi/robotframework-pageobjects
Development:
InActive (no show stoppers with latest release)
Purpose:
One of the libraries of robotframework. Provides a page object abstraction over Selenium2Library. Can be used as Standalone for your framework (Check e.g below for usage) or can be used along with robotframework.
Thoughts:
- It provides a “pageobject abstraction” support on top of Selenium2Library
- It is possible to use this library outside robot-framework context as such without any modification, though it could be using utilities of robot package. (Its your homework to do further experiments on this note!)
- Therefore, can be used as a standalone library
Usage:
pip install robotframework-pageobjects
e.g:
in ipython or idle do:
>>from robotpageobjects import Page
>>start_testing=Page()
>>start_testing.create_webdriver("Firefox")
>>start_testing.go_to("http://google.com")
Option-3: robotframework-pageobjectlibrary
Github Url:
https://github.com/boakley/robotframework-pageobjectlibrary
Development:
Active
Hopefully Author supports LTS (Long Term Support) : )) , Fingers crossed !!
Usage:
pip install robotframework-pageobjectlibrary
Thoughts:
- It is NOT possible to use this library outside robot-framework context. A minor change to how the page context is handled would help this library to be used outside robot-framework context (Its your homework to find out how!)
Option-4: Splinter
Github url:
https://github.com/cobrateam/splinter
Developement:
Active
Usage:
splinter.readthedocs.org/en/latest/index.html
pip install splinter
On ipython or idle do:
>>from splinter import Browser
>>browser = Browser()
>>browser.visit('http://google.com')
>>browser.fill('q', 'splinter - python acceptance testing for web applications')
>>browser.find_by_name('btnG').click()
Option-5: SST library
Github url:
https://github.com/Work4Labs/selenium-simple-test
Development:
Feature complete / Active
Usage:
testutils.org/sst/
pip install -U sst
on ipython or idle do:
>>> from sst.actions import *
>>> start()
Starting Firefox
>>> go_to('http://google.com')
Going to... http://google.com
Waiting for get_element
Option-6: helium
Not open source (Commercial)
Option-7: holmium.core
Github url:
https://github.com/alisaifee/holmium.core
Option-8: wtframework
Github url:
https://github.com/wiredrive/wtframework
Option-9: webium
Github url:
https://github.com/wgnet/webium
Option-10: elementium
Github url:
https://github.com/actmd/elementium
Option-11: saunter
Github url:
https://github.com/Element-34/py.saunter
Usage:
saunter
Option-12: webdriverplus
Github url:
https://github.com/tomchristie/webdriverplus
Usage:
webdriverplus
Comments:
repository not maintained but decent reference
Option-12: Simple-Pageobject
Github url:
https://github.com/rama-bornfree/simple-pageobject/tree/master/PageObjectLibrary
Comments:
Simplest pageobject wrapper built around selenium2library. I am the owner of the repo
Test-setup:
“All” the Test libraries in Option-1-13; can be run using any of the following framework: Lettuce, Behave, Robotframework or for that matter, any unit test framework(e.g PyUnit, Nose)…so on .
Test Framework is in general used to manage test-cases e.g
- English formats like gherkin, keyword, tabular so on …
- reporting a test-run
- hooking to CI
- set-up/tear-down of test-cases and test-suites
- tagging of test-cases
- other functionalities that one could think on about any test-framework
What matters is how comfortable one gets with libraries in above options.
Option-5: As far as SST is concerned, it has features of a framework itself, e.g it can generate report and do many more things.
So the definition of library and framework in the case of SST is blurred, depending on the extent of features one would like to use from that package
Some Math for Fun:
Total number of ways one could have a good, bad, and ugly Test-setup = (Test framework AND Test library + your custom code sandwiched b/w the framework and library ):
7 * 13 = 91 Ways
Choose the best combination (of Test Framework And Test library) that suits ones need !!
I would personally go for Robot-framework with Selenium2Library or Robot-framework with some pageobject library
ofcourse, I am leaned and positively biased in my post about robot-framework and Selenium2Library
There does now exist a port of Capybara itself to Python:
https://github.com/elliterate/capybara.py
You can find its documentation here:
https://elliterate.github.io/capybara.py/
Capybara helps you test web applications by simulating how a real user would interact with your app. It is agnostic about the driver running your tests and comes with Selenium support built in.