An npm module and which generates your Protractor test reports in HTML (angular) with screenshots
npm install protractor-beautiful-reporter
Cross-browser Testing <3 Provided by 
Error: TypeError: Cannot set property 'searchSettings' of undefined use at least version 1.2.7, where this bug has been fixed
protractor-retry or protractor-flake. The collection of results currently assumes only one continuous run.
protractor-beautiful-reporter still generates a HTML report, but it is Angular-based and improves on the original formatting.
protractor-beautiful-reporter module is available via npm:
bash
$ npm install protractor-beautiful-reporter --save-dev
`
In your Protractor configuration file, register protractor-beautiful-reporter in Jasmine.
#### Jasmine 1.x:
`javascript
No longer supported
`
#### Jasmine 2.x:
Jasmine 2.x introduced changes to reporting that are not backwards compatible. To use protractor-beautiful-reporter with Jasmine 2, please make sure to use the getJasmine2Reporter() compatibility method introduced in protractor-beautiful-reporter@0.1.0.
`javascript
var HtmlReporter = require('protractor-beautiful-reporter');
exports.config = {
// your config here ...
onPrepare: function() {
// Add a screenshot reporter and store screenshots to /tmp/screenshots:
jasmine.getEnv().addReporter(new HtmlReporter({
baseDirectory: 'tmp/screenshots'
}).getJasmine2Reporter());
}
}
`
Configuration
$3
You have to pass a directory path as parameter when creating a new instance of
the screenshot reporter:
`javascript
var reporter = new HtmlReporter({
baseDirectory: 'tmp/screenshots'
});
`
If the given directory does not exists, it is created automatically as soon as a screenshot needs to be stored.
$3
The function passed as second argument to the constructor is used to build up paths for screenshot files:
`javascript
var path = require('path');
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, pathBuilder: function pathBuilder(spec, descriptions, results, capabilities) {
// Return '/' as path for screenshots:
// Example: 'firefox/list-should work'.
return path.join(capabilities.caps_.browser, descriptions.join('-'));
}
});
`
If you omit the path builder, a GUID is used by default instead.
Caution: The format/structure of these parameters (spec, descriptions, results, capabilities) differs between Jasmine 2.x and Jasmine 1.x.
$3
(removed because was only used in jasmine 1.x)
$3
You can modify the contents of the JSON meta data file by passing a function jasmine2MetaDataBuilder as part of the options parameter.
Note: We have to store and call the original jasmine2MetaDataBuilder also, else we break the "whole" reporter.
The example is a workaround for the jasmine quirk https://github.com/angular/jasminewd/issues/32 which reports pending() as failed.
`javascript
var originalJasmine2MetaDataBuilder = new HtmlReporter({'baseDirectory': './'})["jasmine2MetaDataBuilder"];
jasmine.getEnv().addReporter(new HtmlReporter({
baseDirectory: 'tmp/screenshots'
jasmine2MetaDataBuilder: function (spec, descriptions, results, capabilities) {
//filter for pendings with pending() function and "unfail" them
if (results && results.failedExpectations && results.failedExpectations.length>0 && "Failed: => marked Pending" === results.failedExpectations[0].message) {
results.pendingReason = "Marked Pending with pending()";
results.status = "pending";
results.failedExpectations = [];
}
//call the original method after my own mods
return originalJasmine2MetaDataBuilder(spec, descriptions, results, capabilities);
},
preserveDirectory: false
}).getJasmine2Reporter());
`
$3
You can store all images in subfolder by using screenshotsSubfolder option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, screenshotsSubfolder: 'images'
});
`
If you omit this, all images will be stored in main folder.
$3
You can store all JSONs in subfolder by using jsonsSubfolder option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, jsonsSubfolder: 'jsons'
});
`
If you omit this, all images will be stored in main folder.
$3
You can change default sortFunction option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, sortFunction: function sortFunction(a, b) {
if (a.cachedBase === undefined) {
var aTemp = a.description.split('|').reverse();
a.cachedBase = aTemp.slice(0).slice(0,-1);
a.cachedName = aTemp.slice(0).join('');
};
if (b.cachedBase === undefined) {
var bTemp = b.description.split('|').reverse();
b.cachedBase = bTemp.slice(0).slice(0,-1);
b.cachedName = bTemp.slice(0).join('');
};
var firstBase = a.cachedBase;
var secondBase = b.cachedBase;
for (var i = 0; i < firstBase.length || i < secondBase.length; i++) {
if (firstBase[i] === undefined) { return -1; }
if (secondBase[i] === undefined) { return 1; }
if (firstBase[i].localeCompare(secondBase[i]) === 0) { continue; }
return firstBase[i].localeCompare(secondBase[i]);
}
var firstTimestamp = a.timestamp;
var secondTimestamp = b.timestamp;
if(firstTimestamp < secondTimestamp) return -1;
else return 1;
}
});
`
If you omit this, all specs will be sorted by timestamp (please be aware that sharded runs look ugly when sorted by default sort).
Alternatively if the result is not good enough in sharded test you can try and sort by instanceId (for now it's process.pid) first:
`javascript
function sortFunction(a, b) {
if (a.instanceId < b.instanceId) return -1;
else if (a.instanceId > b.instanceId) return 1;
if (a.timestamp < b.timestamp) return -1;
else if (a.timestamp > b.timestamp) return 1;
return 0;
}
`
$3
You can set excludeSkippedSpecs to true to exclude reporting skipped test cases entirely.
`javascript
new HtmlReporter({
baseDirectory: tmp/screenshots
, excludeSkippedSpecs: true
});
`
Default is false.
$3
You can define if you want report screenshots from skipped test cases using the takeScreenShotsForSkippedSpecs option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, takeScreenShotsForSkippedSpecs: true
});
`
Default is false.
$3
Also you can define if you want capture screenshots only from failed test cases using the takeScreenShotsOnlyForFailedSpecs: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, takeScreenShotsOnlyForFailedSpecs: true
});
`
If you set the value to true, the reporter for the passed test will still be generated, but, there will be no screenshot.
Default is false.
$3
If you want no screenshots at all, set the disableScreenshots option to true.
`javascript
new HtmlReporter({
baseDirectory: 'tmp/reports'
, disableScreenshots: true
});
`
Default is false.
$3
Also you can define a document title for the html report generated using the docTitle: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, docTitle: 'my reporter'
});
`
Default is Test results.
$3
Also you can change document name for the html report generated using the docName: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, docName: 'index.html'
});
`
Default is report.html.
$3
You can change stylesheet used for the html report generated using the cssOverrideFile: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, cssOverrideFile: 'css/style.css'
});
`
$3
If you want to add small customizations without replaceing the whole css file:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
customCssInline:
});
`
This example will enable line-wrapping if the tests spec contains newline characters
$3
You can preserve (or clear) the base directory using preserveDirectory: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, preserveDirectory: false
});
`
Default is true.
$3
You can gather browser logs using gatherBrowserLogs: option:
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, gatherBrowserLogs: false
});
`
Default is true.
$3
If you do not want all buttons in the search filter pressed by default you can modify the default state via searchSettings: option:
For example: We filter out all passed tests when report page is opened
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, clientDefaults:{
searchSettings:{
allselected: false,
passed: false,
failed: true,
pending: true,
withLog: true
}
}
});
`
Default is every option is set to true
$3
If you do not want to show all columns by default you can modify the default choice via columnSettings: option:
For example: We only want the time column by default
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, clientDefaults:{
columnSettings:{
displayTime:true,
displayBrowser:false,
displaySessionId:false,
displayOS:false,
inlineScreenshots:false
}
}
});
`
Default is every option except inlineScreenshots is set to true
Additionally you can customize the time values for coloring the time column.
For example if you want to mark the time orange when the test took longer than 1 second and red when the test took longer than 1.5 seconds
add the following to columnSettings (values are in milliseconds scale):
`javascript
new HtmlReporter({
baseDirectory: 'tmp/screenshots'
, clientDefaults:{
columnSettings:{
warningTime: 1000,
dangerTime: 1500
}
}
});
`
$3
If you want to show the total duration in the header or footer area...
`javascript
new HtmlReporter({
baseDirectory: 'reports'
, clientDefaults:{
showTotalDurationIn: "header",
totalDurationFormat: "hms"
}
});
`
For all possible values for showTotalDurationIn and totalDurationFormat refer to the wiki entry Options for showing total duration of e2e test
$3
By default the raw data of all tests results from the e2e session are embedded in the main javascript file (array results in app.js).
If you add useAjax:true to clientDefaults the data is not embedded in the app.js but is loaded from the file combined.json which contains all the raw test results.
Currently the reason for this feature is a better testability in unit tests. But you could benefit from ajax loading, if you want to polish/postprocess tests results (e.g. filter duplicated test results). This would not be possible if the data is embedded in the app.js file.
HTML Reporter
Upon running Protractor tests with the above config, the screenshot reporter will generate JSON and PNG files for each test.
In addition, a small HTML/Angular app is copied to the output directory, which cleanly lists the test results, any errors (with stacktraces), and screenshots.
!HTML / Angular Test Report
Click More Details to see more information about the test runs.
!Click More Details
Use Search Input Field to narrow down test list.
!View Search
Click View Stacktrace to see details of the error (if the test failed). Suspected line is highlighted.
Click View Browser Log to see Browser Log (collects browser logs also from passed tests)
!View Stacktrace
Click View Screenshot to see an image of the webpage at the end of the test.
!View Inline Screenshot
Click Inline Screenshots to see an inline screenshots in HTML report.
!View Inline Screenshot
Please see the examples folder for sample usage.
To run the sample, execute the following commands in the examples folder
`bash
$ npm install
$ protractor protractor.conf.js
``