1.3 Powerful and familiar procedural programming
Dylan is not just an object-oriented language. It includes and extends the language features that you expect to find in a more purely procedural language. Dylan's syntax encourages clear and structured programming. It includes familiar, economical notation for infix operators and slot and array references. Dylan offers a choice of concise or expanded equivalents for many syntactic constructs to accommodate a range of programming styles, from terse to descriptive.
You do not have to write a lot of intrusive code to support Dylan's object orientation. For example, the most common language expressions for defining a method automatically define a generic function if necessary. A method-defining expression looks much like a function-defining expression in other languages.
You can define a function to take a variable number of arguments. You can also define a function to take arguments in the form of name-value pairs, thus supporting self-documenting function invocation.
Functions can return more than one value. In fact, you can use a single expression to initialize multiple variables to the values returned by a single function call. You do not have to use a potentially confusing mechanism, such as output parameters, to obtain multiple values.
Dylan has a rich set of variable-sized aggregate data types, called collections. Collection classes include strings, arrays, sets, queues, lists, stacks, and tables. Dylan has flexible iteration constructs and permits applications to extend them so that they operate on application-defined collection subclasses. In this way, a module that uses specialized collection classes can cooperate with another module that defines general collection operations.
Dylan has a built-in exception-signaling and exception-handling system that permits both error handling and recovery. Exceptions are based on a class and object model that fits smoothly with the rest of the language and can be extended by the program. You do not have to return and check error codes from functions — an error-prone process in itself — to ensure that no exception has occurred.