Welcome to Overhave’s documentation!
Ready web-interface for easy BDD features management with Ace editor
Traditional Gherkin format for scenarios provided by pytest-bdd
Auto-collection of pytest-bdd steps and display on the web-interface
Simple business-alike scenarios structure, easy horizontal scaling
Ability to create and use several BDD keywords dictionary with different languages
Synchronization between git repository and database with features
Built-in configurable management of users and groups permissions
Configurable strategy for user authorization, LDAP also provided
Database schema based on SQLAlchemy models and works with PostgreSQL
Still configurable as Flask Admin, supports plug-ins and extensions
Distributed producer-consumer architecture based on Redis streams through Walrus
Web-browser emulation ability with custom toolkit (GoTTY, for example)
Simple command-line interface, provided with Click
Integrated interaction for files storage with s3-cloud based on boto3
You can install Overhave via pip from PyPI:
pip install overhave
The web-interface is a basic tool for BDD features management. It consists of:
Info - index page with optional information about your tool or project;
- Scenarios - section for features management, contains subsections
Features, Test runs and Versions:
gives an interface for features records management and provides info about id, name author, time, editor and publishing status; it is possible to search, edit or delete items through Script panel.
- Test runs
gives an interface for test runs management and provides info about.
contains feature versions in corresponding to test runs; versions contains PR-links to the remote Git repository (only Stash is supported now).
contains tags values, which are used for feature’s tagging.
- Access - section for access management, contains Users and
- Emulation - experimental section for alternative tools implementation
Overhave features could be created and/or edited through special
script panel in feature edit mode. Feature should have type registered by the
application, unique name, specified tasks list with the traditional format
`PRJ-NUMBER` and scenario text.
Script panel has pytest-bdd steps table on the right side of interface. These steps should be defined in appropriate fixture modules and registered at the application on start-up to be displayed.
Overhave generates Allure report after tests execution in web-interface. If you execute tests manually through PyTest, these results are could be converted into the Allure report also with the Allure CLI tool. This report contains scenarios descriptions as they are described in features.
Overhave has special demo-mode (in development), which could be possibly used for framework demonstration and manual debugging / testing. The framework provides a CLI entrypoints for easy server run in debug mode:
make up # start PostgreSQL database and Redis overhave db create-all # create Overhave database schema overhave-demo admin # start Overhave admin on port 8076 in debug mode overhave-demo consumer -s TEST # start Overhave test execution consumer
Note: you could run admin in special mode, which does not require consumers. This mode uses threadpool for running testing and publication tasks asynchronously:
overhave-demo admin --threadpool --language=ru
But this threadpool mode is unscalable in kubernetes paradigm. So, it’s highly recommended to use corresponding consumers exactly.
Overhave has a CLI that provides a simple way to start service web-interface, run consumer and execute basic database operations. Examples are below:
overhave db create-all overhave admin --port 8080 overhave consumer -s PUBLICATION
Note: service start-up takes a set of settings, so you can set them through
virtual environment with prefix
`OVERHAVE_`, for example
If you want to configure settings in more explicit way through context injection,
please see next part of docs.
Service could be configured via application context injection with prepared
instance of OverhaveContext object. This context could be set using
`set_context` function of initialized
`my_custom_context` prepared. So, application start-up could
be realised with follow code:
from overhave import overhave_app, overhave_admin_factory factory = overhave_admin_factory() factory.set_context(my_custom_context) overhave_app(factory).run(host='localhost', port=8080, debug=True)
`overhave_app`is the prepared Flask application with already enabled
Flask Admin and Login Manager plug-ins;
`overhave_factory`is a function for LRU cached instance of the Overhave
`ProxyFactory`; the instance has an access to application components, directly used in
`my_custom_context`is an example of context configuration, see an
example code in context_example.rst.
Enabling of injection¶
Overhave has it’s own built-in PyTest plugin, which is used to enable and configure injection of prepared context into application core instance. The plugin provides one option:
–enable-injection - flag to enable context injection.
The PyTest usage should be similar to:
Overhave has producer-consumer architecture, based on Redis streams, and supported 3 consumer’s types:
- TEST - consumer for test execution with it’s own factory
- PUBLICATION - consumer for features publication with it’s own factory
- EMULATION - consumer for specific emulation with it’s own factory
`overhave_test_execution_factory` has ability for context injection
and could be enriched with the custom context as the
Overhave supports it’s own special project structure:
The right approach is to create a root directory (like “demo” inside the current repository) that contains features, fixtures and steps directories.
The Features directory contains different feature types as separate directories, each of them corresponds to predefined pytest-bdd set of steps.
The Fixtures directory contains typical PyTest modules splitted by different feature types. These modules are used for pytest-bdd isolated test runs. It is necessary because of special mechanism of pytest-bdd steps collection.
The Steps directory contains pytest-bdd steps packages splitted by differrent feature types also. Each steps subdirectory has it’s own declared steps in according to supported feature type.
So, it is possible to create your own horizontal structure of different product directions with unique steps and PyTest fixtures.
Note: this structure is used in Overhave application. The formed data gives a possibility to specify registered feature type in the web-interface script panel. Also, this structure defines which steps will be displayed in the right side of script panel.
Overhave has it’s own special feature’s text format, which inherits Gherkin from pytest-bdd with small updates:
- required tag that is related to existing feature type directory, where
current feature is located;
info about feature - who is creator, last editor and publisher;
task tracker’s tickets with traditional format
An example of filled feature content is located in feature_example.rst.
The web-interface language is ENG by default and could not be switched
(if it’s necessary - please, create a
`feature request` or contribute
The feature text as well as pytest-bdd BDD keywords are configurable with Overhave extra models, for example RUS keywords are already defined in framework and available for usage:
from overhave.extra import RUSSIAN_PREFIXES language_settings = OverhaveLanguageSettings(step_prefixes=RUSSIAN_PREFIXES)
Note: you could create your own prefix-value mapping for your language:
from overhave import StepPrefixesModel GERMAN_PREFIXES = StepPrefixesModel( FEATURE="Merkmal:", SCENARIO_OUTLINE="Szenarioübersicht:", SCENARIO="Szenario:", BACKGROUND="Hintergrund:", EXAMPLES="Beispiele:", EXAMPLES_VERTICAL="Beispiele: Vertikal", GIVEN="Gegeben ", WHEN="Wann ", THEN="Dann ", AND="Und ", BUT="Aber ", )
Overhave gives an ability to sent your new features or changes to remote git repository, which is hosted by Bitbucket or GitLab. Integration with bitbucket is native, while integration with GitLab uses python-gitlab library.
You are able to set necessary settings for your project:
publisher_settings = OverhaveGitlabPublisherSettings( repository_id='123', default_target_branch_name='master', ) client_settings=OverhaveGitlabClientSettings( url="https://gitlab.mycompany.com", auth_token=os.environ.get("MY_GITLAB_AUTH_TOKEN"), )
The pull-request (for Bitbucket) or merge-request (for GitLab) created when you click the button Create pull request on test run result’s page. This button is available only for success test run’s result.
Note: one of the most popular cases of GitLab API authentication is the OAUTH2 schema with service account. In according to this schema, you should have OAUTH2 token, which is might have a short life-time and could not be specified through environment. For this situation, Overhave has special TokenizerClient with it’s own TokenizerClientSettings - this simple client could take the token from a remote custom GitLab tokenizer service.
Overhave gives an ability to synchronize your current git repository’s state with database. It means that your features, which are located on the database, could be updated - and the source of updates is your repository.
For example: you had to do bulk data replacement in git repository, and now you want to deliver changes to remote database. This not so easy matter could be solved with Overhave special tooling:
You are able to set necessary settings for your project:
overhave synchronize # only update existing features overhave synchronize --create-db-features # update + create new features
You are able to test this tool with Overhave demo mode. By default, 3 features are created in demo database. Just try to change them or create new features and run synchronization command - you will get the result.
overhave-demo synchronize # or with '--create-db-features'
Overhave gives an ability to set custom index.html file for rendering. Path to file could be set through environment as well as set with context:
admin_settings = OverhaveAdminSettings( index_template_path="/path/to/index.html" )
Overhave implements functionality for s3 cloud interactions, such as bucket creation and deletion, files uploading, downloading and deletion. The framework provides an ability to store reports and other files in the remote s3 cloud storage. You could enrich your environment with following settings:
OVERHAVE_S3_ENABLED=true OVERHAVE_S3_URL=https://s3.example.com OVERHAVE_S3_ACCESS_KEY=<MY_ACCESS_KEY> OVERHAVE_S3_SECRET_KEY=<MY_SECRET_KEY>
Optionally, you could change default settings also:
The framework with enabled
`OVERHAVE_S3_AUTOCREATE_BUCKETS` flag will create
application buckets in remote storage if buckets don’t exist.
Contributions are very welcome.
Project installation is very easy and takes just few prepared commands (make pre-init works only for Ubuntu; so you can install same packages for your OS manually):
make pre-init make init
Packages management is provided by Poetry.
make up make test make lint
Please, see make file and discover useful shortcuts. You could run tests in docker container also:
Project documentation could be built via Sphinx and simple make command:
By default, the documentation will be built using html builder into _build directory.
Distributed under the terms of the GNU GPLv2 license.
If you encounter any problems, please report them here in section Issues with a detailed description.