sphinx-gherkindoc

A tool to convert Gherkin into Sphinx documentation


License
MIT
Install
pip install sphinx-gherkindoc==3.6.2

Documentation

sphinx-gherkindoc

sphinx-gherkindoc brings Gherkin goodness into the Sphinx/reStructuredText (rST) world.

Why should I use it?

Share your requirements with your larger team. Gherkin makes it easy for anyone to read requirements. Sphinx makes Gherkin pretty enough to be readable. sphinx-gherkindoc handles converting flat-text feature files into easy to read documents you won't cringe to share with your larger team.

Easily see what steps you have and where you are using them. sphinx-gherkindoc can create a glossary of your steps. This helps makes it easy to:

  • find steps to reuse
  • notice similar steps that might be duplicates or unintended variations
  • see patterns that you might exploit to reduce the number of steps you have
  • find out where and which feature files would be affected if a step's wording or implementation changes

How does it do that?

sphinx-gherkindoc recursively scans a given directory to find all the feature and markup files, and converts them into files that can be included in Sphinx-based documentation. This script was inspired by sphinx-apidoc which does a similar thing for source code.

Note

This tool only creates the rST inputs for a sphinx document run, you still need to fit these files in to your larger documentation building process yourself.

For specific details on the command line options, please consult the --help output. Most command line options mirror their counterpart in sphinx-apidoc.

For the most basic usage, an input (<gherkin_path>) and output (<output_path>) path must be provided. The files put in to the <output_path> can be incorporated into any larger documentation, created by any means.

Additionally, a list of fnmatch-compatible patterns can be added to the command line, to indicate directories to be excluded from processing.

One notable addition is the step glossary(-G, --step-glossary). The step glossary command line option causes sphinx-gherkindoc to create the named file in the <output_path> directory. The step glossary content is formatted into two lists:

  • A list of all the steps found, in alphabetical order. Each item in this list is a link to its details in the second list.
  • A list of the steps showing the file names and line numbers where they are used. This list is in order by the most number of uses first.

How are my files converted?

When scanning the <gherkin_path> directory tree, sphinx-gherkindoc will do the following:

  • Feature files found are processed into rST files in the <output_path> directory.
  • Directories will be converted to Sphinx Table of Contents (TOC) files that link to any feature files in the same directory, and to any TOC files from direct subdirectories.
  • Any rST files found in a directory have their contents copied to the front of the TOC file for that directory. If more than one rST file is found in a directory, they are processed in sort order.
  • If no rST files are found in a directory, then a header for the TOC is created based on the contents of a display_name.txt file, if present, or the name of the directory.
  • Any MarkDown (md) files are referenced from the TOC file for the directory they are in.
  • Directories with no feature, rST, or md files are pruned, recursively upwards.

The meat'n'potatoes will be your feature files. Put rST files next to your feature files to present context and additional helpful material. If any rST files are in the same directory, they should also contain any appropriate headers and other such formatting. sphinx-gherkindoc will only create (minimal) headers when there are no rST files present at all.

Examples

Disclaimer: This is not a tutorial on how to use or configure Sphinx. It is some common ways you can use sphinx-gherkindoc as part of your documentation generation. Reminder if you skipped all that stuff above: --help will show you the default values when command line options aren't used.

Conventions - Sphinx-based document production usually uses two directories:

  • _docs - the working directory where we put all the rST files from the various tools as we are building documents. This directory shoud not be checked in to version control and should only contain files created by a documentation run.
  • docs - the output directory for a documentation run. For example, this is where index.html is found when building HTML docs, etc. This directory should not be checked in to version control as it contains only derived and processed files.
  1. Convert feature files to rST; process all feature and document files in/under feature/root/directory-here into the _docs directory using all the defaults:

    sphinx-gherkindoc feature/root/directory-here _docs
    
  2. Same as above, but also create a step glossary file my_step_glossary in _docs:

    sphinx-gherkindoc -G my_step_glossary feature/root/directory-here _docs
    

3. Experiment! Once you have the 2nd step working and integrated in to your document building process, you may find you want to tweak the results some. It's a lot easier to do that after you have the basic process working. Experiment with the other optional parameters to get the effect(s) you want.

Special Instructions for User Integrated Functionality

There are certain parts of sphinx-gherkindoc that can be controlled by external, user defined code. This can be done via a plugin or command-line argument.

Here are the places where this can be used:

  • Any tag or step can be converted into an anonymous link. The converter needs to be a single function that accepts a single string parameter (the tag or full step with keyword) and returns a URL if the entry should include a link or an empty string if not. Note that, in the case of steps, only the step text after the keyword will be converted into a link. In addition to examples below, there is legacy support for a url plugin key that is equivalent to the tag_url key.
  • Directory names can be converted into display names similarly to how display_name.txt works. This converter also needs to be a single function that accepts a single string parameter, the directory name, and returns a display name string for that directory.

In order to register the plugin for a poetry-based project:

[tools.poetry.plugins."parsers"]
tag_url = "my_custom_library.parse:optional_url_from_tag"
step_url = "my_custom_library.parse:optional_url_from_step"
dir_display_name = "my_custom_library.parse:optional_display_name_from_dir"

In order to register the plugin for a setup.py-based project:

setup(
    ...
    entry_points={
        'parsers': [
            'tag_url = my_custom_library.parse:optional_url_from_tag',
            'step_url = my_custom_library.parse:optional_url_from_step',
            'dir_display_name = my_custom_library.parse:optional_display_name_from_dir'
        ]
    }
)

In order to use the parser via command line, the --url_from-tag and --display-name-from-dir flags should be used. The provided string should be be formatted <library>:<method_name>

Filtering Included Features/Scenarios By Tag

If you'd like to build documentation for only a subset of features/scenarios, you can include and exclude them by tags.

Just use the --include-tags and/or --exclude-tags flags. Run sphinx-gherkindoc --help for usage info.

Formatting Options

If you connect a .css file to your generated documentation, via the _static folder that is supported by Sphinx, you can make use of certain classes to format parts of the Gherkin however you so choose!

Here are the available CSS classes:

  • gherkin-step-keyword
  • gherkin-feature-keyword
  • gherkin-feature-content
  • gherkin-feature-description
  • gherkin-background-keyword
  • gherkin-background-content
  • gherkin-scenario-keyword
  • gherkin-scenario-content
  • gherkin-scenario-description
  • gherkin-scenario-outline-keyword
  • gherkin-scenario-outline-content
  • gherkin-scenario-outline-description
  • gherkin-examples-keyword
  • gherkin-examples-content
  • gherkin-tag-keyword

Example .css File:

.gherkin-step-keyword {
    font-weight: bold;
    color: red;
}

.gherkin-feature-keyword, .gherkin-background-keyword {
    color: yellow;
}
.gherkin-feature-content {
    color: green;
}
.gherkin-feature-description {
    color: blue;
}

.gherkin-scenario-keyword, .gherkin-scenario-outline-keyword {
    color: indigo;
}
.gherkin-scenario-content, .gherkin-scenario-outline-content {
    color: violet;
}
.gherkin-scenario-description, .gherkin-scenario-outline-description {
    color: purple;
}

.gherkin-examples-keyword {
    color: gold;
}
.gherkin-examples-content {
    color: goldenrod;
}

.gherkin-tag-keyword {
    color: salmon;
}

Note

Your .css can have any name. As long as it is in the _static folder, Sphinx will integrate it into the generated HTML. Sphinx officially documents adding a CSS file via the Application API