robotkernel 0.10.0

A Jupyter kernel for interactive acceptance-test-driven development with the Robot Framework

Robotkernel

Robot Framework IPython kernel for Jupyter Notebook and JupyterLab.

Requires Python 3.6 or later with setuptools 40.5.0 later and Robot Framework 3.1 or later.

Log | Report -links on existing notebooks are only active on trusted notebooks.

Try Robotkernel at Binder

Jupyter Notebook: https://mybinder.org/v2/gh/datakurre/robotkernel/master?urlpath=tree/example.ipynb

JupyterLab: https://mybinder.org/v2/gh/datakurre/robotkernel/master?urlpath=lab/tree/example.ipynb

Install Robotkernel

$ pip install robotkernel

For JupyterLab you should also install the companion syntax highlighting:

$ jupyter labextension install jupyterlab_robotmode

For some environments it might be required to run the following command to manually register robotkernel as Jupyter kernel:

$ python -m robotkernel.install

Install Robotkernel from Python 3 notebook

!pip install robotkernel

After refreshing the notebook, it is possible change the kernel to Robot Framework kernel or create a new notebook with Robot Framework kernel.

For JupyterLab you should also install the companion syntax highlighting:

!jupyter labextension install jupyterlab_robotmode

Export robot files

It is possible to export test suites direclty from Jupyter Notebook or JupyterLab user interface (into traditional .robot files), but also from command line:

$ jupyter nbconvert --to script example.ipynb

Execute notebooks

Robotkernel installs script named nbrobot, which the Robot Frameworks test runner robot with support for executing Jupyter notebooks created with Robotkernel:

$ nbrobot example.ipynb

In addition, it is also possible to execute notebooks as such, resulting into a new notebook with embedded execution logs and reports:

$ jupyter nbconvert --to notebook --execute example.ipynb

This will stop the execution at first failing test case.

When execution with errors, to also get a result notebook with execution logs saved, an extra flag --ExecutePreprocessor.allow_errors=True must be set:

$ jupyter nbconvert --ExecutePreprocessor.allow_errors=True --to notebook --execute example.ipynb

This may change in future versions of nbconvert.

Note that when executing a notebook, each cell with tests cases or tasks will be executed as its own suite. It might be more efficient to export notebook into a robot script and execute that with the traditional robot runner.

Local installation and development

See also: http://jupyter.readthedocs.io/en/latest/install.html

Create and activate clean Python virtual environment:

$ venv myenv
$ source myenv/bin/activate

Install Jupyter:

$ pip install --upgrade pip setuptools
$ pip install jupyter

Clone this kernel:

$ git clone https://github.com/datakurre/robotkernel.git
$ cd robotkernel

Install the kernel into virtualenv in develop mode:

$ python setup.py develop

Launch the jupyter:

$ jupyter notebook

Reloading the kernel reloads the code.

Nix-shell (https://nixos.org/nix/)

This repository includes opinionated instructions for running and developing Robotkernel with Nix for Jupyter Notebook:

$ nix-shell -E 'import (fetchTarball https://github.com/datakurre/robotkernel/archive/master.tar.gz + "/shell.nix")' --run "jupyter notebook"

And for Jupyter Lab:

$ nix-shell -E 'import (fetchTarball https://github.com/datakurre/robotkernel/archive/master.tar.gz + "/shell.nix")'
$ jupyter labextension install jupyterlab_robotmode --app-dir=.jupyterlab
$ jupyter lab --app-dir=.jupyterlab
$ exit

Add --arg sikuli true to include SikuliLibrary.

Add --arg vim true to enable vim bindings.

Development environment with Nix:

$ git clone https://github.com/datakurre/robotkernel.git
$ cd robotkernel
$ nix-build setup.nix -A env  # to generate ./result/bin/python for IDE
$ nix-shell setup.nix -A develop

Changelog

0.10.0 (2019-01-08)

Breaking:

• Setuptools 40.5.0 later and Robot Framework 3.1 or later. [datakurre]

New featurs:

• Add pregenerated kernel.json in data_files to auto-install robotkernel, but requiring [datakurre]

0.9.0 (2019-01-04)

New featurs:

• Rewrite status updater to the status of currently run test as trobber | test name | keyword name | robot.api.logger.console message [datakurre]

• Add embedded log and report to include Download-links at top right corner [datakurre]

• Add to display the results of the last executed keyword as the notebook result for the executed code cell [datakurre]

• Add syntax highlighting for variables and assignment operators [datakurre]

Bug fixes:

• Fix issue where setup.cfg contained OS specific path separators preventing build on Windows [datakurre]

• Fix issue where PNGs were interpreted as APNG on Windows preventing them from being rendered on Windows [datakurre]

• Update example notebook to use SeleniumLibrary and SeleniumScreenshots instead of SeleniumLibrary and Selenium2Screenshots [datakurre]

• Add to always reload libraries imported from other notebooks [datakurre]

0.8.0 (2018-12-14)

• Add to auto-import nbimporter when available to make it possible to import eg. keyword libraries from Python notebooks [datakurre]

• Fix issue where nbrobot did support %%python module magic [datakurre]

0.7.1 (2018-11-20)

• Add to require robotframework >= 3.1b1 in requires

0.7.0 (2018-10-31)

Breaking:

• Requires robotframework >= 3.1b1

Other:

• Add to create nbreader and nblibdoc cli to run robot with notebook reader support [datakurre]

• Add Selenium completions to sometimes include raw Simmer results with simplfied id completion results [datakurre]

• Add proof-of-concept selector completion for Appium and AutoIT libraries [datakurre]

• Fix screenshot processor to also discover images with absolute path or within the current working directory [datakurre]

0.6.3 (2018-10-19)

This is the last release compatible with robotframework < 3.1

• Remove deprecated replace-flag from kernel installer [datakurre]

0.6.2 (2018-10-19)

• Fix compatibility issue with robotframework < 3.1 [datakurre]

0.6.1 (2018-10-19)

• Fix issue where kernel installation produced broken kernel.json on Windows [datakurre]

0.6.0 (2018-10-18)

• Revert data source path from temporary directory into current working directory to allow local libraries and resources work in the usual use cases [datakurre]

• Add experimental Simmerjs based CSS-selector builder and element picker with when auto-completion is called with empty “css:”-selector [datakurre]

• Add experimental Selenium selector auto-completion [datakurre]

• Add dummy variable completion with only variables from current suite without context knowledge [datakurre]

• Add inline documentation links to Robot Framework User Guide for structural keywords [datakurre]

0.5.4 (2018-10-09)

• Fix issue where single term keywords got no completions [datakurre]

0.5.3 (2018-10-09)

• Update README [datakurre]

0.5.1 (2018-10-08)

• Auto completion and keyword doc inspection enhancements [datakurre]

0.5.0 (2018-10-08)

• Add auto-completion, keyword doc inspection and support for replacing and deleting cell history on Jupyter lab [datakurre]

0.4.0 (2018-09-26)

• Add support for robotframework 3.1a2 [datakurre]

• Add support for reporting RPA suites with “Tasks” instead of “Tests” [datakurre]

0.3.5 (2018-09-25)

• Update README with notebook execution instructions [datakurre]

0.3.4 (2018-09-25)

• Update README [datakurre]

0.3.3 (2018-09-25)

• Note on README that Log | Report -links require trusting the notebook [datakurre]

• Fix to wrap test execution updates with ‘<pre>’ for better readability [datakurre]

0.3.2 (2018-09-25)

• Change to always send display data updates in text/html to workaround a bug that caused ‘undefined’ to be rendered in JupyterLab [datakurre]

0.3.1 (2018-09-24)

• Update README [datakurre]

0.3.0 (2018-09-23)

• First release. [datakurre]