Dsl Migration

4 February 2009

One danger that DSL advocates need to guard against is the notion that first you design a DSL, then people use it. Like any other deice of software, a successful DSL will evolve. This means that scripts written in an earlier version of a DSL may fail when run with a later version.

Like many properties of DSL, good and bad, this is really very much the same as happens with a library. If you take a library from a someone and they upgrade the library, you may end up stuck. In essence DSLs don't really do anything to change that. Your DSL definition is essentially a PublishedInterface and you have to deal with the consequences just the same.

This problem can be more prominent with external DSLs. Many changes to an internal DSL can be handled through refactoring tools (for those languages that have them). But refactoring tools won't help with an external DSL. In practice this problem is less of an issue than it might be. An internal DSL with scripts that are outside the control of the DSL implementors won't be picked up with refactoring. So the only difference between internal and external lies with DSL scripts within the same code base.

One technique for handling evolution of DSLs is to provide tools that automatically migrate a DSL from one version to another. These can be run either during an upgrade, or automatically should you try to run an old version script against a new version.

There are two broad ways to handle migration. The first is an incremental migration strategy. This is essentially the same notion that's used by people doing evolutionary database design. For every change you do to your DSL definition, create a migration program that automatically migrates DSL scripts from the old version to the new version.

An important part of incremental migration is that you keep the changes as small as you can. Imagine you are upgrading from version 1 to 2, and have ten changes you want to make to your DSL definition. In this case, don't create just one migration script to migrate from version 1 to 2, instead create at least 10 scripts. Change the DSL definition one feature at a time, and write a migration script for each change. You may find it useful to break it down even more and add a feature with more than one step (and thus more than one migration). They way I've described it may sound like more work than a single script, but the point is that migrations are much easier to write if they are small, and it's easy to chain multiple migrations together. As a result you'll be much faster writing ten scripts than one.

The other approach is model-based migration. This is a tactic you can use if you are using a Semantic Model (which is something I almost always recommend). With this approach you support multiple parsers for your language, one for each released version. (So you only do this for version 1 and 2, not for the intermediate steps.) Each parser populates the semantic model. When you use a semantic model, the parser's behavior is pretty simple, so it's not too much trouble to have several of them around. You then run the appropriate parser for the version of script you are working with. This handles multiple versions, but doesn't migrate the scripts. To do the migration you write a generator from the semantic model that generates a DSL script representation. This way you can run the parser for a version 1 script, populate the semantic model, and then emit a version 2 script from the generator.

One problem with the model-based approach is that it's easy to lose stuff that doesn't matter to the semantics, but is something that the script writers want to keep. Comments are the obvious example. This is exacerbated if there's too much smarts in the parser, but then the need to migrate this way may help encourage the parsers to stay dumb - which is Good Thing.

If the change to the DSL is big enough, you may not be able to transform a version 1 script into a version 2 semantic model. In which case you may need to keep a version 1 model (or intermediate model) around and give it the ability to emit a version 2 script.

I don't have a strong preference between these two alternatives.

Migration scripts can be run by script programmers themselves when needed, or automatically by the DSL system. In order to run automatically it's very useful to have the script record which version of the DSL it is so the parser can detect it easily and trigger the resulting migrations.