Group feature files or scenario based on cucumber-tags and execute in parallel
npm install cucumber-tags-groupingA simple module which helps to group BDD scenario based on special tags defined in feature file. We call those tags as grouping tags.
Mostly we define multiple feature file and execute them in parallel. Cucumber execute feature files in parallel using the CLI option --parallel. All the feature files are executed in parallel which matches --tags, feature-folder. This module helps to gain control over the parallel run.
Grouping scenarios based on tags (special grouping tags) and execute scenarios based a group, parallelly. All the scenarios of first group are executed in parallel, once it's completed, we can continues with second group.
Lets say you need to cleanup some test data or make some database operation or any server config operation or between your parallel execution, then one can smartly group BDD scenarios based on tags.
Another great use case is Feature Toggle - by Martin Fowler. Application behavior changes when 'featureToggle'/'featureFlag'/'switch' changes. In this case easiest way to handle the scenario is to make a grouping tag which differentiate scenarios based on feature flag. Configure the feature flags for a specific group then execute test scenarios in that group in parallel. Next configure the feature flags for next group then execute the test scenarios in that specific group in parallel... and so on.
Simple workflow:
Execute a group and then perform some operation and then jump to next group, then repeat.
``sh`
npm install cucumber-tags-grouping --save
Those who use yarn
`sh`
yarn add cucumber-tags-grouping
Four simple steps:
1; First and foremost, decide a pattern for grouping tags. Below are few example.
> NOTE: Grouping tags should follow a specific pattern which can be identified by a regular expression.
- Regular expression: /GROUP/ | tags example: @GROUP_A or @GROUP_B or @GROUP_RED or @GROUP_ALL.
- Regular expression: /GRP/ | tags example: @GRP_USER_TYPE_A or @GRP_USER_TYPE_B or @GRP_USER_TYPE_DEFAULT.
- Regular expression: /ENV/ | tags example: @ENV_DB_A or @ENV_DB_B or @ENV_DB_C.
2; Assign at least one grouping tag to each scenario or feature file or example. Rules for assigning tags are exactly same as cucumber-js.
3; Import the module and use the below code to generate the groups. It will return with array of string. Each string is a cucumber-tag-expression and is uniquely groups the scenario.
`js
import cucumberTagGroups from 'cucumber-tags-grouping';
async function test() {
const scenarioGroups = await cucumberTagGroups({
featuresGlobPattern: 'features/*/.feature', // required
tagForGrouping: /GROUP/, // required
tagExpression: undefined, // optional
name: undefined, // optional
});
console.log(scenarioGroups);
}
test();
// Feature files and step definitions referred:
// https://github.com/iAbhishek91/cucumber-tags-grouping/tree/master/features
// output [ '@GROUP_A_ON and @GROUP_B_OFF and @GROUP_C_ON', '@GROUP_A_ON and @GROUP_B_ON and @GROUP_C_OFF', '@GROUP_ALL' ]
// Each element in the array are valid tag expression
`
Refer: example code in github. To execute the test.js just clone the project and execute the below commands.
- yarn build or npm run build
-
4; Utilize these tags to invoke cucumber. Very important to keep in mind that all the parameters (featuresGlobPattern, tagForGrouping and name) should remain same(as passed in cucumberTagGroups) while invoking cucumber for each group. tagExpression parameter should be concatenated with the each groups.
With context to the above example:
if tagExpression is @firstA, and output is [ '@GROUP_A_ON and @GROUP_B_OFF and @GROUP_C_ON', '@GROUP_A_ON and @GROUP_B_ON and @GROUP_C_OFF', '@GROUP_ALL' ], then tagExpression for executing the groups should be:
- (@firstA and @GROUP_A_ON and @GROUP_B_OFF and @GROUP_C_ON) - first group
- (@firstA and @GROUP_A_ON and @GROUP_B_ON and @GROUP_C_OFF) - second group
- (@firstA and @GROUP_ALL) - third group
This function takes one parameter, an object. The object keys are described below.
- featuresGlobPattern [required]
This parameter is exactly same as cucumber. Use any valid glob pattern.
- tagForGrouping [required]
This is a regular expression. The regular expression that can match multiple tags and create groups out it. Note: *while using this module we need to tag all feature file with at least one grouping tag. Its suggested to use a default tags where none of the grouping tags are applicable.
Example of group tags
One may create tags as below. And assign every scenario or feature with at least one tag.
@GROUP_A
@GROUP_B
@GROUP_C
@GROUP_D
@GROUP_DEFAULT
Note: tags can be define tags for feature, scenario, scenario-outline and example keywords. Exactly same as cucumber.
- tagExpression [optional]
If not provided all the feature files under the directory of featuresGlobPattern will be considered for evaluation.
Tag expression parameter is also same as cucumber. Use any valid cucumber-tag-expression
- name [optional]
Same as cucumber name parameter.
As said cucumber-tags-grouping seats in between cucumber processing. When we execute cucumber there are certain stages or steps. Lets take help of an example to demonstrate.
Example:
Assume you are executing the below command
`sh`
cucumber-js features//.feature --require src/step-definition//.js --tags "@red or @blue" --parallel 5
Stages(in high level):
1. Cucumber will determine all the feature files that are to be consumed in folder features/*/.feature.@red
2. Then it will evaluate the tag-expression using cucumber-tag-expression. Cucumber will determine which scenario are in scope with tag or @blue`.
3. Next, cucumber will spawn nodejs child process for executing the scenarios that are in scope in parallel.
4. Finally, cucumber will gather the result and send it to formatter to generate the report.
Cucumber-tags-grouping seats in between step-2 and step-3. It will perform step-1 and then step-2 based on the regular expression provided, then it will handover each groups to cucumber for processing. Cucumber will go through the above four step for each group separately.
Please contact author.
In case of any suggestion, change or fix please raise an issue before raising a PR.