Arkency is a company that shares its experiments with many approaches to DDD aggregate implementation. Some time ago I’ve shared my view on this here. Arkency examples are in Ruby and use an event-sourcing approach. It’s quite different from mine, so some future blog posts will be about reimplementing the Arkency examples in Python. Experiment … Read more
In one of our projects, we needed ACL (Anti-Corruption Layer) for external entity-id and internal. Generally mapping one to another but based on the complex external id that differs depending on the platform that it came from. Below you have examples of these. Table for mappings To store these mappings we decided to use the … Read more
In the previous post Persistency of DDD aggregate, I described persisting as aggregate with proper application boundaries. Now let’s describe how to persist ValueObject in a relational database using SQLAlchemy and still with boundaries in mind. The persistency of ValueObject in relational databases can be tricky. Unlike entities, there is no id for value objects. We can … Read more
Mutation testing is a concept that I heard some time ago but lately on anti-if course (I highly recommend!) done by Andrzej Krzywda I had a chance to use it in practice. I notice a big difference in the number of mutations generated by different frameworks (mutmut, cosmic-ray, and mutpy) when I was refactoring Gilded … Read more
Basic tox config Let’s start with a basic config that will run unittest and measure coverage. Important option here is usedevelop = true. With this it won’t install your package to virtualnenv and coverage can be measured. Another thing is to erase previous coverage reports. Codestyle Next, I wanted to check the codestyle. I use pycodestyle, … Read more
Sometimes I found myself in a situation where it’s hard to understand what some components suppose to achieve. In these situations, I go to test where scenarios/contracts of this component should be.But then I find some integrations tests with some entries in DB and expected mechanical results.My strategy in this situation is to refactor test … Read more
DDD Aggregate is responsible for encapsulating business logic for the component. But we need to be aware that when we are implementing business logic for a long time process that is coordinating many components, like subscription trial, we have another building block for this. It’s a process manager. Below I pointed out some differences between … Read more
In most cases, I use UUID as an id for aggregate entity. But when it comes to big amount of entities in SQL DB it can become scalability problem to handle UUID as id. Here are some pros and cons of different approaches to entity ids. Autoincrement Integer: DB 32-bit or 64-bit integer used as … Read more
So there is this legacy code. It’s a good example of code that is pretty simple code but at the same time, hard to understand the business case. These code snippets are already simplified cause it was a part of a bigger object that was processing shipping order. My goal is to refactor it a … Read more
Introducing branching in code (by feature toggle) is not the easiest practice but gives you some direct and some not direct benefits. Let’s start with direct ones. A release can be done by business and can be a partial release. So if a release has some exact date it’s better to prepare it on feature … Read more
