Validating architecture

This chapter describes how to validate your source code against the architecture, design patterns, and other team conventions.

Benefits

Validating code is especially important for projects developed by a large team or maintained for an extended period.

Executable rules instead of paper guidelines. Architectural guidelines can now be enforced in real-time within the code editor instead of being simply written down and put on shelves.
Immediate feedback. Developers don't need to wait for the CI build to complete. Feedback is given within seconds.
Smoother code reviews. Petty rule violations are automatically detected so that code reviews can focus on flows and concepts.
Better team alignment. Automated code validation encourages the team to respect consistent patterns and practices.
Lower complexity. The resulting codebase is simpler when everyone on the team follows consistent patterns and practices.
Reduced architecture erosion. The gap between the initial architecture and its implementation in source code stays smaller.

You can validate code using both aspects or fabrics. The approach is very similar.

Use aspects to encapsulate your validation logic as a custom attribute that can be used in the current or other projects. Custom attributes are preferable when the target declarations must be hand-picked.
Use fabrics if you do not need reusable logic or prefer to expose an object-oriented API instead of custom attributes. Fabrics are preferable when the target declarations can be selected as a general rule.

This chapter contains the following articles:

Article	Description
Validating declarations	This article explains how aspects and fabrics validate a declaration. It also shows how to validate the state of the code model after all aspects have been applied.
Validating references	This article describes how to validate the way a declaration is used, i.e., how to validate every reference instead of the declaration itself.