protractor-junit-xml-plugin
v2.0.0
Published
A Protractor plugin. Report results in junit xml format including requirement ids if available.
Downloads
1,162
Readme
protractor-junit-xml-plugin
Description: This plugin is a protractor reporter plugin that reports the test results in JUnit XML file so if you are looking for a plugin to export JUnit XML results you are at the right place. This plugin also has an additional option to parse and extract Xray Id (A JIRA plugin to identify requirements) from a test name and put it in a separate attribute in the output exported XML. This can help in exporting test results with xrayId and link the test to a requirement in Xray Jira (Continuous Integration). It also has an option to just use tests with an xrayId and ignore the tests that don't have an XrayId. By default these additional options are turned off. Please see the configuration section for the option details
- Technology stack: This is a protractor plugin written in Javascript. It is a node module and can be imported using npm
- Status: This is the first functional version of this plugin. We are planning to add a change log starting from the next version
Dependencies
This plugin is dependent on the node modules "xmlbuilder currently 12.0.1" and "xml2js-parser currently v1.1.1"". Although these dependencies should be install automatically when this plugin would be installed using npm
Installation
The easiest way is to keep protractor-junit-xml-plugin
as a dependency in your package.json
pointing to its current repo
npm install protractor-junit-xml-plugin --save-dev
To update to the latest version
npm update protractor-junit-xml-plugin
Configuration
For an example of a protractor config file for this plugin, please see plugin section of protractor.conf.js located in the tests folder of this plugin
The following are the configurable options
outdir: output directory of the test results, if not specified it defaults the directory name to "_test-reports"
filename: the filename starting of output result file, if not specified it defaults the file name to "test-results.xml"
parseXrayId: If set to true it will try to extract Xray Jira id from the test name and put it in the attribute "requirement" in testcase result xml. The id should be in between two colons e.g ":XRAY-1234: test to check report graphs". The result testcase element would be like <testcase name="Failing test with xrayId" ... requirements="XRAY-1234">. This is helpful if this tet is attached to a jira requirement and you want to report back the result(Continuous Integration).
if not specified it defaults to "false"
xrayIdOnlyTests: If set to true, it will only consider the tests that have Xray Jira Id in their name. If not specified it defaults to "false"
uniqueName: If set to true, it will generate a unique name for the filename. Default is true. This can not be true if appendToFile is also true.
appendToFile: If set to true, it will append xml data to the bottom of the file instead of creating a new file or overwriting the file. This can not be true if uniqueName or uniqueFolder is set to true. Default is false
uniqueFolderPerExection: Default is false. If set to true, it will create a new directory for each time the test suite is run though it creates a one line(very small) file resultDirName.txt where it will store the directory name
captureSapphireWebAppContextVar(Default is false): If set to true, it will look for a global object `sapphireWebAppConfig` in the app and capture the properties like appName, appVersion etc from it to a metadata file that this plugin will generate for each run(in addition to testresults xml file). The idea is to capture all the runtime info related to the test results.
Usage
After setting the plugin config in the protractor config file (usually named as protractor.conf.js). This plugin would run everytime you run the tests. It will generate the junit output xml in the specified output directory. For each browser there will be a seapate junit out xml file. For e.g if we have set two browsers in our protractor config lets say firefox and chrome and we gave filename option in plugin cofig as e2e-tests
then the file names will be
- chrome-e2e-tests.xml
- firefox-e2e-tests.xml
If the result file already exist then the plugin will not overwrite it but it will append the current test results in a new testsuite
element with a timestamp
How to test the software
Unit Tests
Unit tests are in Mocha/Chai framework. After installing dependencies(npm install
), one can run the unit tests by mocha tests/unitTests.js
End to end tests:
mocha e2e test:
We have a Mocha e2e test that calls 2 protractor tests using a config file (tests/protractor.conf.js
) and check the following test case
- If we run 2 tests in parallel and
uniqueFolderPerExection
option is set to true then results should be created in a single folder
Here are the steps to run mocha test:
- Firstly, install all the dependencies if not done so far
npm install
npm i -g mocha
mocha tests/mocha-e2eTest.js
protractor e2e tests You can also run protractor e2e tests that uses this plugin to generates a XML output file and manually verify the correct entries in the XML file. The e2e tests are in protractor. To run the tests
- Firstly, install all the dependencies if not done so far
npm install
. - Install protractor as a global dependency
npm i -g protractor
- Start the node test app
node tests/testapp/server.js
(tested on node v12). - Run the tests
protractor tests/protractor.conf.js
Known issues
We are actively using this plugin for our protractor e2e tests and have not seen any issues so for. Please contact us if you run into any issues.
Contact / Getting help
You can contact any of us if you run into any issues
- Arsalan Siddiqui <[email protected]>
- Stephanie Gawroriski <[email protected]>
License
Credits and references
We needed a protractor plugin to export our test results to a JUnit XML file with XRay Ids so we wrote this plugin. We have tried to make it general with the configurable options so that it can be as useful as possible. We thank you to "Philips Health Solutions" in general to give us opportunity to write this plugin. We are also very thankful to Ryan Gatto for his positive and supportive efforts all along.