+972 52-548-6969
A time for a DSL
Most of the time, I use a DSL as a way to setup and configure some sort of a runtime engine. We will get to that later when we will talk about imperative vs. declarative DSL. But the rule of the thumb that I use is that I’ll use a Fluent Interface if I usually need this functionality at the same time I am developing.
If I need to be expressive at a different time than the development of the application, then it probably means that I should use a DSL instead. The issues of complexity vs. expressiveness also come into play, of course.
If I can have a more expressive syntax out of a DSL than I can when using a Fluent Interface, that is also a consideration.
I’ll admit that I tend to favor building DSL over Fluent Interfaces, precisely because I leant the other way in the past, and got burned badly by the complexity involved in trying to maintain something of that complexity in a rigid language.
If you need to do things externally, a DSL is the place to look for. In fact, if you feel the urge to put XML into place, as anything except strict data storage mechanism, that is a place to go with a DSL.
Fluent Interfaces relies heavily on intellisense in order to create that fluent feeling. If you need to work on it outside the IDE, that is probably going to be a factor.
Extensibility is also a concern; let us go back to the scheduling sample. We have the scheduling engine, and we have the tasks themselves. Consider that to write a task using a DSL I usually have to write a small text file, but to write a task using a Fluent Interface requires creating a project, compiling, etc.
This means that it is often easier to use a DSL than a Fluent Interface, especially if we would like to hand this DSL to non developers.
Comments
What has your experience been with error messages in the DSLs you've created? In my experience, MS's C# compiler gives excellent error messages, at least until you get into serious extension method/type inference/lambda territory. When a your DSLs get a bit on the complex side, do the error messages suffer?
That depends on the type and complexity of the DSL.
The parser errors are not really great, but anything else is under you control.
@Luke
You may get decent error messages, but debugging's a pain. Compiler treates a bunch of fluent interface calls as one line of code, so it becomes difficult to step in to individual calls.
ScottGu suggested debugger visualizers at altnetconf, but they don't work for a lot of fluent interface scenarios. I haven't really found a better solution...
Comment preview