Command Line Tool - behat¶
Behat on its own is just a command line utility that executes your behavior descriptions written in Gherkin language. behat comes with bunch of useful options and commands:
To show all available commands and options, run:
$ behat -h
To view the version of the behat tool used, call:
$ behat -V
To initialize the feature suite structure inside your project, run:
$ behat --init
Running this command will setup a features directory and a skeleton FeatureContext class inside your project:
You can use configuration files to configure your feature suite. By default, behat will try to load behat.yml and config/behat.yml, but if you want to name your config file differently, tell behat about it with the --config option:
$ behat --config custom-config-file.yml
Your configuration files can have multiple profiles (named configurations). You can run behat with a specific profile by calling it with the --profile option:
$ behat --config behat.yml --profile ios
The default profile is always named default.
Behat supports different ways of printing output information. Output printers in behat are called formats or formatters. You can tell behat to run with a specific formatter by providing the --format option:
$ behat --format progress
The default formatter is pretty.
behat supports 6 formatters out of the box:
pretty - prints the feature as is:
progress - prints one character per step:
html - almost the same as pretty, but prints HTML output.
junit - generates a report similar to Ant+JUnit.
failed - prints paths to failed scenarios.
snippets - prints only snippets for undefined steps.
If you don’t want to print output to the console, you can tell behat to print output to a file instead of STDOUT with the --out option:
$ behat --format html --out report.html
Some formatters, like junit, always require the --out option to be specified. The junit formatter generates *.xml files for every feature, so it needs a destination directory to put these XML files into.
Also, you can specify multiple formats to be used by Behat with comma (,):
$ behat -f pretty,progress
In this case, default output will be used as output for both formatters. But if you want them to use different ones - specify them with --out:
$ behat -f pretty,progress,junit --out ~/pretty.out,,xml
In this case, output of pretty formatter will be written to ~/pretty.out file, output of junit formatter will be written to xml folder and progress formatter will just print to console. Empty out option (as in case with progress in example) tells Behat to use stdout. So:
$ behat -f pretty,progress,junit --out ,progress.out,xml
Will print pretty output instead, but will write progress output to progress.out file.
Behat tries hard to identify if your terminal supports colors or not, but sometimes it still fails. In such cases, you can force behat to use colors (or not) with the options --ansi or --no-ansi, respectively:
$ behat --no-ansi
Behat prints suite execution time information after each run. If you don’t want this information, you can turn it off with the --no-time option:
$ behat --no-time
Also, there are a bunch of options to hide some default output information from the output:
- --no-paths - hides paths after steps and scenarios.
- --no-snippets - hides snippet proposals for undefined steps after statistics.
- --snippets-paths - prints step information with snippets.
- --no-multiline - hides multiline arguments (tables and pystrings) from pretty output.
By default, Behat prints scenario outlines the same as you define them:
The output is pretty minimal and enough for you to see some errors. But in some complex cases, it may be hard to actually find failed steps in the output. To make this easier, behat offers the --expand option:
$ behat --expand
This options will make the previous output more verbose:
$ behat --name="element of feature"
Only execute the feature elements which match part of the given name or regex.
If the element is part of feature-definition behat executes the whole feature, otherwise one or more scenarios.
$ behat --rerun="return.log"
Save list of failed scenarios into new file or use existing file to run only scenarios from it.
If the file contains no scenarios, then all scenarios will executed.
If you don’t know where to start with the Gherkin syntax, Behat can help you with some feature example:
$ behat --story-syntax
This command will print an example feature for you to understand what keywords to use and where to use them in your *.feature files:
If you write features in a language other than English, you can view a localized example feature in your language of choice by using the --lang option:
$ behat --story-syntax --lang fr
Will print the feature example in French:
Also, if you forgot what step definitions you’ve already implemented, or how to spell a particular step, behat will print all available definitions by calling it with the -dl option:
$ behat -dl
This command will print all available definition regular expressions:
If you want to get extended info about definitions, use:
$ behat -di
To search for a specific definition, use:
$ behat -d 'search string'
If you want to run only part of your suite, or some scenarios, you can do it with name or tags filters.
Tag filters supports three logical operators:
- Tags separated by commas will be combined using the OR operation
- Tags separated by && will be combined using the AND operation
- Tags preceded by ~ will be excluded using the NOT operation
The first command will run only features or scenarios which have the @orm OR @database tag (or both).
The second command will run only features or scenarios which have the @ticket OR @723 tag (or both).
The third command will run only features or scenarios with both the @orm AND @fixtures tags (must have both).
The fifth command will run only features and scenarios that contain number has in their title.