In the previous article we built two Web Components (tc-book-form and tc-books), learned how to change their style through properties, how to control them through exposed functions as well as how to connect them using custom events. We developed most of the code using Test-Driven Development approach. Both front-end Web Components and back-end server are packed into a single microservice and stored as a Docker container. We did not do the last part (Docker build, push and run) together and we’ll correct that by exploring, among other things, containers definition, building and running in this article.
With all this in place, we are ready to import components we did in previous articles into a separate Web Application.
In the previous article we developed our second Polymer Web Component. With two of them functionally finished, we’ll apply styles and create a way for users of our components to have partial control of the look and feel. Once that is done, we’ll improve our Demo page and create proper communication between our components.
In the previous article we finished one Polymer Web Component. It contained a form that could be used to add, update and remove books by sending HTTP requests to the back-end server. We also explored few different ways to interact with the component from outside by calling its functions.
In this article we’ll create the second component that will be used to list all the books and send custom events to open an existing or a new book. Those events will not be sent directly to the component we built earlier but to the document the component will reside on. We’ll get to that part in time…
In the previous article we created our first Polymer Web Component using Test-Driven Development approach. While all the specifications we wrote were implemented, there are still few things we should do to polish this component. The tc-book-form currently has the form with all the input fields required by the back-end. It also has buttons to add or update and delete a book. Those buttons are calling iron-ajax Polymer element that sends requests to the back-end. Both front-end (in form of Web Component) and back-end are developed as a single microservice.
The next task in front of us is to polish the component. Later on in this series we’ll create another component that will list all books. We’ll also develop the code that will enable us to establish communication between those two components. Once we’re done with both, we’ll explore ways to integrate the microservices we developed into a separate Web Application.
In this article series we’ll go through Web Components development in context of microservices. We’ll use Polymer as the library that will help us out. The objective is to create a microservice that will handle full functionality. The service will contain not only back-end API (as is the case with most microservices) but also front-end in form of Web Components. Later on, when the time comes to use the service we’re creating, we’ll simply import Web Components. That is quite a different approach than what you might be used to. We won’t create a Web Application that calls APIs handled by microservices. We’ll import parts of the front-end from microservices. Our objective is to have a microservice that contains everything; from front-end to back-end. Please refer to the Including Front-End Web Components Into Microservices article for more information that lead to this decision.
I am a huge fan of Test-Driven Development so we’ll write all the code using test first approach. We’ll write a test, run all tests and confirm that the last one fails, write implementation, run all tests and confirm that all passed. Tests will act as executable specifications. We’ll use them to define what should be implemented. If you are new to TDD and the concept of executable documentation, please read the article Tests As Documentation.
Microservices are becoming more and more popular and many are choosing to transition away from monolithic architecture. However, this approach was mostly limited to back-end services. While it made a lot of sense to split them into smaller independent pieces that can be accessed only through their APIs, same did not apply to front-end. Why is that? I think that the answer lies in technologies we’re using. The way we are developing front-end is not designed to be split into smaller pieces.