American National Standards Institute. Reference Manual for the Ada Programming Language, February 1983. ANSI/MIL-STD 1815A. Also published by Springer-Verlag as LNCS 155.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....started to notice the escalating expense of updating and maintaining code, structured programming became increasingly popular. And as programs got bigger still, a modular style of programming was generally accepted as essential for programming in the large. Languages such as Modula 2[42] Ada[3], Standard ML[27] later variants of Pascal and C[19] implemented features that enabled programmers to program using modules, strucutures and units. In the past decade, a style of programming referred to as the objectoriented (henceforth abbreviated as OO ) paradigm has proven to be ....

American National Standards Institute. Reference Manual for the Ada Programming Language, February 1983. ANSI/MIL-STD 1815A. Also published by Springer-Verlag as LNCS 155.

....Group. INTRODUCTION The International Standards Organization (ISO) published the revised Ada standard [1] and an accompanying Rationale [2] in February 1995. The new standard replaces the 1987 ISO standard [3] which was identical to the 1983 American National Standards Institute (ANSI) standard [4]. To distinguish the language defined by the new standard from that defined by the older standards, the names Ada 95 and Ada 83 are frequently used; this naming convention is adopted here 1 . The primary purpose of this paper is to discuss the provisions made in the 1 Whether this particular ....

American National Standards Institute, ANSI/MIL-Std-1815a. Reference Manual for the Ada Programming Language, 1983.

.... Figure 2: Two tuple classes holding methods for testing equality of points. 6 ColorPoint) tuple class is only invoked if both arguments are ColorPoint instances at runtime. The use of dynamic types distinguishes multiple dispatch from static overloading (as found, for example, in Ada 83 [Ada 83] The semantics of sending messages to tuples, multiple dispatch, is similar to that in Cecil [Chambers 95] Consider the expression (E 1 , E n ) I(E n 1 , E m ) where each E i has value v i , and where T d,i is the minimal dynamic type of v i . A method in a tuple class ....

.... Figure 2: Two tuple classes holding methods for testing equality of points. 6 ColorPoint) tuple class is only invoked if both arguments are ColorPoint instances at runtime. The use of dynamic types distinguishes multiple dispatch from static overloading (as found, for example, in Ada 83 [Ada 83] The semantics of sending messages to tuples, multiple dispatch, is similar to that in Cecil [Chambers 95] Consider the expression (E 1 , E n ) I(E n 1 , E m ) where each E i has value v i , and where T d,i is the minimal dynamic type of v i . A method in a tuple class (T 1 ....

American National Standards Institute. Reference Manual for the Ada Programming Language. ANSI/MILSTD 1815A, February, 1983.

....by integers between 0 and 255, we hide the storage table for instances of this class. Notice how the code, given below, resembles, say, a C class declaration, with a private part that comes before the public part. Comments start with two hyphens, and continue to the end of that line, as in Ada [Ada 83] and Cecil. 8 BeCecil Chambers Leavens We assume that declarations for generic functions intensity and paint, several generic functions to manipulate integers and floating point numbers, and a class Region rep are given elsewhere. object Grayscale rep Grayscale rep inherits Color rep hide ....

.... in singly dispatched OO and procedural styles (respectively) single oo method has method(first myClass, snd any, thd any) proc method has method(first any, second any, thd any) 21 BeCecil Chambers Leavens As Cook points out [Cook 90] one characteristic of ADT languages, such as Ada 83 [Ada 83] CLU [Liskov et al. 81] or Standard ML, is that they completely control the implementation of a particular type of object. That is, they are able to guarantee behavioral properties of objects of a given type, because they control the ability to create and modify objects of that type. ....

[Article contains additional citation context not shown here]

American National Standards Institute. Reference Manual for the Ada Programming Language. ANSI/MILSTD 1815A, February, 1983.

....synchronization paradigms. In particular, they hold for all paradigms in which a program P in the basic Ada subset can be emulated by a program P 0 , such that the number of statements in P 0 is polynomially related to the number of statements in P . 69 4.3 Accept. do construct. Ada [Ame83] has features that allow the rendezvous to act as a type of remote procedure call . The task entry (or accept) acts as the body of the called procedure, and the signaling task acts as the caller. The syntax of a remotely callable procedure is as follows: accept signal [ parameter list ) do ....

....of nodes in D, as well as D itself: 8y such that UngSis (D; y) 0 Scope(D[t y ] y) 0 SigN set (D; y) 4.11) 4.4.3 Conditional entries with else clauses or delay alternatives. The else clause will typically not include any coordinations with other tasks. However, the ANSI Ada specification [Ame83] does not rule out this possibility, so it must be 86 considered for the sake of completeness. We first examine the impact of else clauses without rendezvous on possible deadlocks, and then look at the effects of including rendezvous statements in the else clause. Else clauses and delay ....

American National Standards Institute. ANSI/MIL-STD 1815A (1983) reference manual for the Ada programming language. United States Government Printing Office, 1983.

....based on the dynamic classes of p1 and p2. In particular, the method from the (ColorPoint, ColorPoint) tuple class is only invoked if both arguments are ColorPoint instances at run time. The use of dynamic classes distinguishes multiple dispatch from static overloading (as found, for example, in Ada 83 [Ada 83] The semantics of sending messages to tuples, multiple dispatch, is similar to that in Cecil [Chambers 95] Consider the expression (E 1 , E n ) I(E n 1 , E m ) where each E i has value v i , and where C d,i is the minimal dynamic class of v i . A method in a tuple class (C 1 ....

....the dynamic classes of p1 and p2. In particular, the method from the (ColorPoint, ColorPoint) tuple class is only invoked if both arguments are ColorPoint instances at run time. The use of dynamic classes distinguishes multiple dispatch from static overloading (as found, for example, in Ada 83 [Ada 83] The semantics of sending messages to tuples, multiple dispatch, is similar to that in Cecil [Chambers 95] Consider the expression (E 1 , E n ) I(E n 1 , E m ) where each E i has value v i , and where C d,i is the minimal dynamic class of v i . A method in a tuple class (C 1 , C n ) ....

American National Standards Institute. Reference Manual for the Ada Programming Language. ANSI/MILSTD 1815A, February, 1983.

....by integers between 0 and 255, we hide the storage table for instances of this class. Notice how the code, given below, resembles, say, a C class declaration, with a private part that comes before the public part. Comments start with two hyphens, and continue to the end of that line, as in Ada [Ada 83] and Cecil. 8 BeCecil Chambers Leavens We assume that declarations for generic functions intensity and paint, several generic functions to manipulate integers and floating point numbers, and a prototype Region obj are given elsewhere. object Grayscale obj Grayscale obj inherits Color obj ....

.... in singly dispatched OO and procedural styles (respectively) single oo method has method(first myClass, snd any, thd any) proc method has method(first any, second any, thd any) 21 BeCecil Chambers Leavens As Cook points out [Cook 90] one characteristic of ADT languages, such as Ada 83 [Ada 83] CLU [Liskov et al. 81] or Standard ML, is that they completely control the implementation of a particular type of object. That is, they are able to guarantee behavioral properties of objects of a given type, because they control the ability to create and modify objects of that type. ....

[Article contains additional citation context not shown here]

American National Standards Institute. Reference Manual for the Ada Programming Language. ANSI/MILSTD 1815A, February, 1983.

....is passed as an actual parameter by reference, then the global and formal become aliases. For example, in Figure 1 the formal parameter names a and b in are aliases within size s body, as the type int indicates that the parameter b is passed by reference. var size: int = 10; array x: int[10] [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]; proc sum(a: int[ b: int) imports (size) var i: int = 1; b : 0; while (i size 1) b : b a[i] i : i 1 call sum(x, size) Fig. 1. Aliasing of a global variable. As illustrated by the above examples, a major problem with parameter aliasing is that programmers often forget that ....

....of objects of the same type. Collections are explicit program variables that act like the implicit arrays indexed by pointers ( 27] p. 14) Thus the key ideas all relate to how to eliminate aliasing for arrays. Arrays in Euclid use the direct model [10, Chapter 6] like Pascal and Ada [2, 5, 17], and thus distinct declarations of arrays do not overlap. However, Euclid does have call by reference. The approach taken to eliminating aliases resulting from reference parameters is to prohibit procedure calls when the actual parameters overlap. This includes structured data passed along with a ....

American National Standards Institute. Reference Manual for the Ada Programming Language, February 1983. ANSI/MIL-STD 1815A. Also published by Springer-Verlag as LNCS 155.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC