This document outlines development practices that we follow internally while developing Drop-in.
nvm use
npm install
npm run build
This creates the following dist
structure:
dist
├── gh-pages/
├── npm/
└── web/
└── dropin/
├── 1.0.0/
│ ├── css/
│ └── js/
└── dev -> 1.0.0/
dist/gh-pages
contains the demo app and JSDocs that are published to gh-pages.
dist/npm
contains the pre-processed source files that are published to npm.
dist/web/
mirrors the structure of Drop-in assets available at https://assets.braintreegateway.com.
dist/web/dropin/dev
assets are available only in development and are never deployed.
Note: If you are developing on a Windows machine, you will need to run your command prompt as an administrator so the symlinking step succeeds.
A demo app for the latest release of braintree-web-drop-in
and JSDocs are published on gh-pages.
To run the app and docs locally, follow the build instructions and then run npm run development
.
This will serve dist/gh-pages
at port 4567 and watch for changes in src
.
In sandbox and production environments, Drop-in injects a stylesheet onto the page retrieved from https://assets.braintreegateway.com/web/dropin//css/dropin.css.
When developing, you can include a locally built CSS file on the page that will override the hosted stylesheet by giving it the id braintree-dropin-stylesheet
:
<link rel="stylesheet" type="text/css" href="/path/to/local/dropin.css" id="braintree-dropin-stylesheet">
Adding a new payment method requires changing a number of files. For each of these sections, you will see an example of adding a fake payment method, FooPay
to the file. Note: This guide may fall out of date as things change in the repo. If you find a mistake, please open a PR to fix it.
There are a few constants that should be updated in src/constants.js.
paymentOptionIDs
- In this case, we'll addfooPay: 'fooPay'
to the listpaymentMethodTypes
- For this, we need to mapfooPay
to whatever the Braintree Gateway lists the tokenization type as. So, if the Gateway lists it asFooPayAccount
, we should addfooPay: 'FooPayAccount'
analyticsKinds
- The same is true here, if the Gateway usesFooPayAccount
, then we addfooPay: 'FooPayAccount'
Unless you need to translate error messages or other UI elements (see below), simply add your payment method name to src/translations/en_US.js. Braintree developers will have to work with PayPal to provide the translation files for them.
'FooPay': 'FooPay'
The src/dropin-model.js must be updated so that Drop-in can check if the payment option is available for the customer to use.
Primarily, the isPaymentOptionEnabled
function must be adjusted to account for the new payment method. It checks if the merchant is enabled for the particular payment method in the Braintree gateway, if the merchant has configured Drop-in to enable the payemnt method, and any other requirements the payment method may have to be used in Drop-in. For instance, if FooPay can only be used if the FooPay
global exists on the window, we would probably add code that looks like this to isPaymentOptionEnabled
.
} else if (paymentOption === paymentOptionIDs.fooPay) {
return gatewayConfiguration.fooPay && Boolean(options.merchantConfiguration.fooPay) && global.FooPay;
}
The payment option id will also need to be added to the DEFAULT_PAYMENT_OPTION_PRIORITY
to determine the default order for displaying the payment method options.
var DEFAULT_PAYMENT_OPTION_PRIORITY = [
// ...
paymentOptionIDs.applePay,
paymentOptionIDs.fooPay
];
If a customer's vaulted payment methods cannot be used on the client (i.e., ApplePay, GooglePay, Venmo), you must also add the payment method to the VAULTED_PAYMENT_METHOD_TYPES_THAT_SHOULD_BE_HIDDEN
array. For instance:
var VAULTED_PAYMENT_METHOD_TYPES_THAT_SHOULD_BE_HIDDEN = [
paymentMethodTypes.ApplePayCard,
paymentMethodTypes.FooPayAccount
];
The src/html/main.html must be updated with a div for the payment method's view. This is the UI where the payment method is initiated by the customer (filling the card form, pushing the PayPal button, etc). In our example, FooPay requires a button for the customer to press to initate the flow.
<div data-braintree-id="foo-pay" class="braintree-foo-pay braintree-sheet">
<div data-braintree-id="foo-pay-sheet-header" class="braintree-sheet__header">
<div class="braintree-sheet__header-label">
<div class="braintree-sheet__logo--header">
<svg height="24" width="40">
<use xlink:href="#logoFooPay"></use>
</svg>
</div>
<div class="braintree-sheet__label">{{FooPay}}</div>
</div>
</div>
<div class="braintree-sheet__content braintree-sheet__content--button">
<div data-braintree-id="foo-pay-button" class="braintree-sheet__button--foo-pay foo-pay-button"></div>
</div>
</div>
We will also need to update src/less/main.less to make the payment sheet view visible when selected. Add .braintree-show-{payment-method-name} .braintree-{payment-method-name}
to the list of classes in the "Dropin Visibility States" section.
// ...
.braintree-show-applePay .braintree-applePay,
.braintree-show-fooPay .braintree-fooPay {
display: block;
height: auto;
overflow: visible;
visibility: visible;
}
You will also need to add your payment method to the .braintree-show-{payment-method-name} [data-braintree-id='other-ways-to-pay']
section.
// ...
.braintree-show-paypal [data-braintree-id='other-ways-to-pay'],
.braintree-show-applePay [data-braintree-id='other-ways-to-pay'],
.braintree-show-fooPay [data-braintree-id='other-ways-to-pay'] {
display: block;
}
You will need to add your payment method to the switch statement in src/views/payment-method-view.js.
If your payment method is vaultable on the client, set the @TITLE
to be an indentifier for the account. Such as a username, email, last 4 numbers of the account, etc. The @SUBTITLE
will be the payment method name.
case paymentMethodTypes.fooPay:
html = html.replace(/@ICON/g, 'logoFooPay')
.replace(/@CLASSNAME/g, '')
.replace(/@TITLE/g, this.paymentMethod.details.accountName)
.replace(/@SUBTITLE/g, this.strings.FooPay);
break;
If the payment method is not vaultable, there is no need to identify it, since you can only ever have one of that payment method type available. In that case, leave the @SUBTITLE
as an empty string and set the @TITLE
to the payment method name.
case paymentMethodTypes.fooPay:
html = html.replace(/@ICON/g, 'logoFooPay')
.replace(/@CLASSNAME/g, '')
.replace(/@TITLE/g, this.strings.FooPay)
.replace(/@SUBTITLE/g, '');
break;
Similiarly, the switch statement in src/views/payment-options-view.js will also need to be updated.
case paymentOptionIDs.fooPay:
paymentSource = this.strings.FooPay;
html = html.replace(/@ICON/g, 'logoFooPay');
break;
Add a new file at src/views/payment-sheet-views/{payment-method-name}-view.js, this will be the JS portion of the view. This view will take care of setting up the Braintree component.
If you need to collect form details, such as with a credit card or us bank account, include a requestPaymentMethod
function.
var BaseView = require('../base-view');
var btFooPay = require('braintree-web/foo-pay');
var DropinError = require('../../lib/dropin-error');
var paymentOptionIDs = require('../../constants').paymentOptionIDs;
function FooPayView() {
BaseView.apply(this, arguments);
}
FooPayView.prototype = Object.create(BaseView.prototype);
FooPayView.prototype.constructor = FooPayView;
FooPayView.ID = FooPayView.prototype.ID = paymentOptionIDs.fooPay;
FooPayView.prototype.initialize = function () {
var self = this;
self.model.asyncDependencyStarting();
return btFooPay.create({
client: this.client
}).then(function (fooPayInstance) {
self.fooPayInstance = fooPayInstance;
self.model.asyncDependencyReady();
}).catch(function (err) {
self.model.asyncDependencyFailed({
view: self.ID,
error: new DropinError(err)
});
});
};
FooPayView.prototype.requestPaymentMethod = function () {
return this.fooPayInstance.tokenize();
};
If your payment method is self contained, like PayPal or ApplePay, simply set up whatever your payment method needs to function, such as a button click to initialize the flow and call this.model.addPaymentMethod
with the payload and this.model.reportError
with the error.
btFooPay.create({
client: this.client
}).then(function (fooPayInstance) {
var btn = self.getElementById('foo-pay-button');
btn.addEventListener('click', function (event) {
event.preventDefault();
fooPayInstance.tokenize().then(function (payload) {
self.model.addPaymentMethod(payload);
}).catch(function (tokenizeErr) {
self.model.reportError(tokenizeErr);
});
});
self.model.asyncDependencyReady();
}).catch(function (err) {
self.model.asyncDependencyFailed({
view: self.ID,
error: new DropinError(err)
});
});
Add the payment method view to the index file at src/views/payment-sheet-views/index.js.
result[paymentOptionIDs.fooPay] = require('./foo-pay-view');
Finally, we just need to add a map for the Gateway Payment Method Type to the translation string for that type in src/views/payment-methods-view.js.
var PAYMENT_METHOD_TYPE_TO_TRANSLATION_STRING = {
// ...
ApplePayCard: 'Apple Pay',
FooPayAccount: 'FooPay'
};
If your payment method has a configuration that can be updated after Drop-in has been created, but before tokenization occurs, such as updating the amount setting when authorizing a PayPal account, add an updateConfiguration
method to the sheet view. If your payment method does not require this, skip this section.
FooPayView.prototype.updateConfiguration = function (key, value) {
this.fooPayConfiguration[key] = value;
};
Next, open src/dropin.js and update the UPDATABLE_CONFIGURATION_OPTIONS
constant to include your payment method:
var UPDATABLE_CONFIGURATION_OPTIONS = [
paymentOptionIDs.paypal,
paymentOptionIDs.paypalCredit,
// others
paymentOptionsIDs.fooPay
];
Finally, if updating the configuration would invalidate the authorized payment method, add it to the UPDATABLE_CONFIGURATION_OPTIONS_THAT_REQUIRE_UNVAULTED_PAYMENT_METHODS_TO_BE_REMOVED
constant.
var UPDATABLE_CONFIGURATION_OPTIONS_THAT_REQUIRE_UNVAULTED_PAYMENT_METHODS_TO_BE_REMOVED = [
paymentOptionIDs.paypal,
paymentOptionIDs.paypalCredit,
// others
paymentOptionsIDs.fooPay
];
If you don't need to handle specific errors, you can let Drop-in populate a generic error. If you do need to handle a specific error, you can pass the key of a specific string to use in the translation file.
Alternatively, you can pass the BraintreeError
from braintree-web into this.model.reportError
and create a translation string for the error code, where the property name is the camel cased version of the code with Error
appended to it. (See hostedFieldsTokenization
errors in src/translations/en_US.js for examples).
fooPayTokenizationFailedError: 'Something went wrong when connecting to FooPay.'
Add documentation info to src/dropin.js, src/index.js, and jsdoc/home.md
Run unit tests and lint:
npm test
We use Browserstack to automate end to end testing on Google Chrome, Safari, and Firefox.
First, sign up for a free open source Browserstack account and a and PayPal Sandbox.
Copy the .env.example
file to .env
cp .env.example .env
Fill in the BROWSERSTACK_USERNAME
and BROWSERSTACK_ACCESS_KEY
environmental variables with your Browserstack credentials:
BROWSERSTACK_USERNAME=username
BROWSERSTACK_ACCESS_KEY=access_key
Fill in the PAYPAL_USERNAME
and PAYPAL_PASSWORD
environmental variables with your PayPal sandbox customer credentials:
PAYPAL_USERNAME=<PayPal sandbox username here>
PAYPAL_PASSWORD=<PayPal sandbox password>
To start the tests, start the app:
npm run development
And run the integration test command:
npm run test:integration
This will run the tests from test/integration
in the browsers specified in wdio.conf.js
.
To run only the PayPal tests, run:
npm run test:integration:paypal-only
To run all tests but the PayPal tests, run:
npm run test:integration:paypal-skipped
By default, each test will retry once if it fails. PayPal tests will retry up to 4 times. You can ignore the retry behavior by passing DISABLE_RETRIES=true
before running the test command.
DISABLE_RETRIES=true npm run test:integration
You can mark a test with an .only
tag:
it.only('asserts something', function () {
And then run the test command to run only that test:
npx wdio wdio.conf.js --spec test/integration/path-to-test.test.js
To run tests in only one browser, prefix the test command with an ONLY_BROWSERS
env variable:
# run only in chrome browser
ONLY_BROWSERS=chrome npm run test:integration
# run only in safari browser
ONLY_BROWSERS=safari npm run test:integration
# run only in firefox browser
ONLY_BROWSERS=firefox npm run test:integration
To run tests in certain browsers, prefix the test command with an ONLY_BROWSERS
env variable, with each browser comma separated:
# run only in firefox and chrome browsers
ONLY_BROWSERS=firefox,chrome npm run test:integration
If you need to update a key for a translation that you have already updated in en_US
, run the following with the existing key and then the new key as arguments:
node ./scripts/update-translation-key.js OLD_KEY_NAME NEW_KEY_NAME
If you have added strings to en_US
that you need to get translated, run the following to get the formatted output:
node ./scripts/get-translation-strings.js