Skip to main content
Archived

This project has been archived by its maintainers, and is no longer receiving any updates.

Human Framework: Test Automation Framework for Humans™

NOTICE: Human Framework is still in the Proof-of-Concept stage

CONTRIBUTE! I am releasing Human Framework under GPLv3 which makes it open source and free.

Contribute by:

  1. Localization of training data (translation)
  2. Adding more intents and actions
  3. Reporting bugs
  4. Suggesting new features

NOTICE: I am working on a Rasa NLU support

Human Framework is a test automation framework designed for testers without software programming background by allowing them to write test steps or instructions in their own language like English.

Human Framework uses Natural Language Understanding (NLU) for intent classification and entity extraction to be able to perform specific actions. It currently depend on Microsoft LUIS (Language Understanding).

Below is an example of test instructions that Human Framework can understand. To get an idea of what specific sentences are supported, check the the content of trials and tests folder.

Open chrome browser https://devpost.com
Page title should be "Devpost - The home for hackathons"
Page should contain link "About"
Click link "About"
Page title should be "About us · Devpost"
Close browser

LUIS (and other NLU tools) makes it possible to write test cases on any human language. Current version of Human Framework only supports English.

Human Framework was inspired by Robot Framework but leans towards Natural Language Processing (NLP) for writing test cases.

"Human Framework is the Siri or Alexa of automation testing." (minus the voice, but it is possible in the future)

How it works

Architecture

The architecture of Human Framework can be simplified into 2 steps- Intent Classification and Action Execution:

  1. Human Framework reads the test case files and then uses an Intent Classifier (e.g. LUIS.ai) to identify intents and entities. For example, the sentence "open chrome browser" will be identified with the intent "web.open_browser" and entity "chrome" as the browser type.
  2. Human Framework then executes the action specific for the classified intent. For example, the intent "web.open_browser" will use Selenium to automate opening a chrome browser.

Usage / Installation

Create an Azure account

  1. Go to Azure portal and create an account.

  2. On the Azure portal dashboard, click Create a resource, look for Cognitive Services and click Create

    Create a resource

    Search cognitive services

    Create cognitive service

  3. Fill up form and click Create button

    Create cognitive service form

Create a LUIS.ai account

  1. Download the Human Framework training data from https://raw.githubusercontent.com/roniemartinez/HumanFramework/master/train/en.json and save it with .json extension

  2. Go to LUIS.ai and create an account

  3. Click Import new app

    Import new app

  4. Click Choose app file (JSON format) ... and select the downloaded Human Framework training data

    Choose app file

  5. Optionally, type your desired app name

    Done import new app

  6. Click Done

  7. Click on the application you created.

  8. On the application dashboard, click Train to train your LUIS.ai app

Connecting Azure account to LUIS.ai and publishing

  1. Click MANAGE > Azure Resources > Add prediction resource

    Manage LUIS.ai

  2. Fill form

    Assign a resource to your app

  3. Click Publish, select Production and then click the Publish button

    Publish LUIS.ai app

  4. Go back to Azure Resources. If you published your application to Staging environment, use Change query parameters to switch between environments.

    Copy Endpoint

  5. Copy the endpoint assigned resource in #2

NOTE: After copying the endpoint, enter it to a browser's address bar to make sure that it is accessible., Otherwise, repeat the steps or use the Starter_Key which will only give you 1000 API calls

Installing Human Framework

Human Framework is written in Python 3. Start by downloading/installing Python from python.org.

To install Human Framework, enter the following command into the terminal/cmd:

Directly from Github

pip install -e git+git@github.com:roniemartinez/HumanFramework.git#egg=humanframework

From PyPI

pip install humanframework

Setting up Human Framework

Use the copied endpoint from LUIS.ai and setup Human Framework. Paste the endpoint enclosed in double quotes.

human config --luis-endpoint "<endpoint>"

Installing Drivers

Human Framework depends on Selenium for testing web applications. Download your desired webdriver and extract them to your local working directory.

Writing tests

Write a text file (test_*.txt) containing your tests and save it to local working directory where you placed the Selenium Drivers. For example, if we have a file named test_web.txt:

open chrome https://python.org
page title should be "Welcome to Python.org"
close browser

Running tests

To run the tests in test_web.txt, enter the following on the terminal/cmd:

human --test test_web.txt

Test autodiscovery

Files inside the trials folder that starts with the text test_ are automatically executed.

trials
|- test_simple.txt
|- test_simple_2.txt
|- ...

With this structure, you can simply type human to run the test cases.

The future of Human Framework

  • Integrated Editor - While testing was greatly simplified because Human Framework will just execute sentences, using the terminal to install Python and several libraries is still more developer-centric. To remove this technical requirement, another major goal of the Human Framework project is to release a one-click installer which should include new features like an integrated IDE/Editor. This will greatly increase tester's productivity.
  • Language Support - Not all testers use the English language. Supporting most languages will benefit more testers around the world.
  • NLU Options - While LUIS.ai is very simple to use and accurate compared to other Intent Classifiers, it is not cheap. To support more testers, a free option like Rasa NLU can be used as an alternative.

Limitations

I only implemented a few intents and actions for the proof-of-concept stage. To know what are the possible tests that you can do with Human Framework, check the contents of trials and tests folders.

Author

Metadata

Release files for humanframework 0.6.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for humanframework 0.6.2
File Size Uploaded
humanframework-0.6.2.tar.gz 14.2 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for humanframework 0.6.2
File Interpreter ABI Platform
humanframework-0.6.2-py2.py3-none-any.whl Python 3, Python 2 none any Details

Total release size: 40.1 kB

Release files / humanframework-0.6.2.tar.gz

Download URL humanframework-0.6.2.tar.gz
Size 14.2 kB
Tags Source
SHA-256 checksum
How to use checksums
6c5847926acb58d0f401413899b394641ca02b83025dcc44adf96f2fe77fa220
BLAKE2b-256 checksum
How to use checksums
c044dc82dc8b82c9e620c2a09c5538fbbfe5aba23aa85bb66ebc2caf6e4f8d31
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/1.14.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.35.0 CPython/3.7.1

Release files / humanframework-0.6.2-py2.py3-none-any.whl

Download URL humanframework-0.6.2-py2.py3-none-any.whl
Size 25.9 kB
Tags Python 2 Python 3
SHA-256 checksum
How to use checksums
26db4d4d1044e86762029d99d1da957141ceeccf5e1b7904ca74808c7e27508e
BLAKE2b-256 checksum
How to use checksums
2139395db646c713c4df3c1d13e2f87e94aafebb447451cd2f76cee241881a5d
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/1.14.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.35.0 CPython/3.7.1

Release history Release notifications | RSS feed

This release

0.6.2 This release

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page