6.001 Structure and Interpretation of Computer Programs. Copyright o 2004 by Massachusetts Institute of Technology Slide 5.1. 14 Summary of part 1 So here is the summary of what we have seen in this section Isolate details of process from its use Designer has choice of which ideas to isolate, in order to support general patterns of computati 0O1 sIcP 6.001 Notes: Section 5.2 Slide 5.2.1 Language Elements So let's take that idea of abstraction and build on it to set the stage for what we are about to do, it is useful to think about how.mitives pnm. data: numbers, strings, booleans the language elements can be group together into a hierarchy rimitive procedures At the atomic level, we have a set of primitives. In Scheme these include primitive data objects: numbers, strings and compound data(today) Means of Abstraction Booleans. And these include built-in, or primitive, procedures for numbers, things like, +,=,>, for strings, things like block structure string=?, substring, for Booleans, things like and, or, not higher order procedures (next time To put these primitive elements together into more interesting conventional interfaces-lists(today data abstraction expressions, we have a means of combination, that is, a way of 6001 SICP combining simpler pieces into expressions that can themselves be treated as elements of other expressions. The most common one, and the one we have seen in the previous lectures, is procedure application. This is the idea of creating a combination of subexpressions, nested within a pair of parentheses the value of the first subexpression is a whic me, asas wepsesseen Chatwe s uhe ete healie f thep ts for e cer es ond ingerexpmesersni th body of the procedure, and proceed with the evaluation. We know that these combinations can themselves be included within other combinations, and the same rules of evaluation will recursively govern the computation Finally, our language has a means of abstraction: a way of capturing computational elements and treating them as if they were primitives; or said another way, a method of isolating the details of a computation from the use of a computation. Our first means of abstraction was define, the ability to give a name to an element, so that we could just use the name, thereby suppressing the details from the use of the object. This ability to give a name to something is most valuable when used with our second means of abstraction, capturing a computation within a procedure. This means of abstraction dealt with the idea that a common pattern of computation can be generalized into a single procedure, which covered every possible application of that idea to an appropriate value. When coupled with the ability to give a name to that procedure, we engendered the ability to create an important cycle in our language: we can now create procedures, name them, and thus treat them as if they were themselves primitive elements of the language. The whole goal of a high-level language is to allow us to suppress unnecessary detail in this manner, while focusing on the use of a procedural abstraction to support some more complex computational Today, we are going to generalize the idea of abstractions to include those that focus on data, rather than procedures. So we are going talk about how to create compound data objects, and we are going to examine standard6.001 Structure and Interpretation of Computer Programs. Copyright © 2004 by Massachusetts Institute of Technology. Slide 5.1.14 So here is the summary of what we have seen in this section. 6.001 Notes: Section 5.2 Slide 5.2.1 So let's take that idea of abstraction and build on it. To set the stage for what we are about to do, it is useful to think about how the language elements can be group together into a hierarchy. At the atomic level, we have a set of primitives. In Scheme, these include primitive data objects: numbers, strings and Booleans. And these include built-in, or primitive, procedures: for numbers, things like *, +, =, >; for strings, things like string=?, substring; for Booleans, things like and, or, not. To put these primitive elements together into more interesting expressions, we have a means of combination, that is, a way of combining simpler pieces into expressions that can themselves be treated as elements of other expressions. The most common one, and the one we have seen in the previous lectures, is procedure application. This is the idea of creating a combination of subexpressions, nested within a pair of parentheses: the value of the first subexpression is a procedure, and the expression captures the idea of applying that procedure to the values of the other expressions, which means as we have seen that we substitute the values of the arguments for the corresponding parameters in the body of the procedure, and proceed with the evaluation. We know that these combinations can themselves be included within other combinations, and the same rules of evaluation will recursively govern the computation. Finally, our language has a means of abstraction: a way of capturing computational elements and treating them as if they were primitives; or said another way, a method of isolating the details of a computation from the use of a computation. Our first means of abstraction was define, the ability to give a name to an element, so that we could just use the name, thereby suppressing the details from the use of the object. This ability to give a name to something is most valuable when used with our second means of abstraction, capturing a computation within a procedure. This means of abstraction dealt with the idea that a common pattern of computation can be generalized into a single procedure, which covered every possible application of that idea to an appropriate value. When coupled with the ability to give a name to that procedure, we engendered the ability to create an important cycle in our language: we can now create procedures, name them, and thus treat them as if they were themselves primitive elements of the language. The whole goal of a high-level language is to allow us to suppress unnecessary detail in this manner, while focusing on the use of a procedural abstraction to support some more complex computational design. Today, we are going to generalize the idea of abstractions to include those that focus on data, rather than procedures. So we are going talk about how to create compound data objects, and we are going to examine standard