When to Use Abstract Syntax Tree Tooling to Refactor at Scale

[00:00 - 00:17] Determining when to use an AST-based tool can be equally as important as understanding the tools themselves. In practice, AST-based tooling is most useful when working with large code bases, think hundreds of files or more, and when targeting code that follows a similar pattern.

[00:18 - 00:30] In smaller code bases, it's easy to manually update things across several files , or use a find-replace tool for simpler renaming. In a large code base, it could take days or longer to make manual changes depending on the exact changes.

[00:31 - 00:43] Potentially more concerning is that with so many repetitive changes, the chances of a making mistake increases. Another way you could say it is that AST-based tooling enables maintenance and refactoring at scale.

[00:44 - 00:54] This is primarily because using AST-based tooling is a one-time fixed cost. On the other hand, the cost of making manual changes grows linearly with the size of the code base.

[00:55 - 01:03] For AST-based tools, you write a custom script once and run it on all files. It doesn't matter if another file is added, removed, or changed.

[01:04 - 01:12] It takes the same amount of time to create the custom script for one file or a thousand. On the other hand, when making manual changes, the total number of files matters.

[01:13 - 01:25] As files are added, removed, or changed, it directly impacts the amount of time it takes to make those updates. With this logic, it's easier to determine if a task would be quicker to do manually, or quicker using an AST-based tool.

[01:26 - 01:34] This decision depends on how comfortable you are with AST-based tools. As you become more comfortable working with ASTs, writing custom scripts will take less time.

[01:35 - 01:42] This means over time, the break-even size of the code base will become smaller and smaller for it to be worth writing a custom script. script.