User Tools

Site Tools


at:tutorial:basic

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
at:tutorial:basic [2007/04/06 08:58]
elisag add
at:tutorial:basic [2020/02/09 22:05] (current)
elisag
Line 1: Line 1:
-<note> 
-**IN PROGRESS: FIRST DRAFT!!** 
- 
-- Add quasi-quoting? 
-</note> 
 ====== Functional and Imperative Programming ====== ====== Functional and Imperative Programming ======
    
-This part of the tutorial shows AmbientTalk as a simple expression language with a minimum syntax which resembles very on Java script. This section mainly describes the basic features of the language, namely variables, functions and tables and control flow.+This part of the tutorial explains AmbientTalk as a simple expression language with a flexible syntax which resembles languages like Ruby, Python and Javascript. This section mainly describes the basic features of the language, namely variables, functionstables (i.e. arrays) and control flow primitives.
  
 ===== Variables ===== ===== Variables =====
    
-As usual, one can define, assign and refer to a variable. Variable definitions are made with the keyword **def**. Note that AmbientTalk is a dynamically typed language sovariables do not have a type but, they just contain values.+As usual, one can define, assign and refer to a variable. Variable definitions are made with the keyword **def**. Note that AmbientTalk is a dynamically typed languageso variables do not have a type but can contain any value.
  
-In the examples we use the interactive AmbientTalk shell (iat) where the input and output prompt are represented by > and >> respectively. +In the examples we use the interactive AmbientTalk shell (iat) where the input and output prompt are represented by > and %%>>%% respectively. 
  
 <code> <code>
Line 21: Line 16:
 </code> </code>
  
-Variable definitions can be combined with assignments as shown aboveAs in Pico, assignments uses the ":=operator. Note that there must be an space between the variable and the ":=" operator in order for the parse to resolve the ambiguity between a keyword message and a assignment, e.g. ":= 1" is understood as an assignment while "a:" as a keyword. We will further elaborate on keywords in the following sections.+Variable definitions can include an initialization expression that immediately initializes the variableVariable assignment is performed by means of the well-known '':='' operator (''='' is used for mathematical comparison)AmbientTalk supports assignment to multiple variables as a single assignment expression. For this to work, the number of variable names on the left hand side of ":=" must match the number of expressions on the right hand side of ":=". A typical application of this is to swap the values of two variables more easily:
  
-An assignment consists of one or more expressions, providing that the number of expressions on the right hand side match the number of variables on the left hand side. This allows a permutation of variables such as: 
 <code> <code>
 >[x, y] := [ y, x ] >[x, y] := [ y, x ]
 >>[7,5] >>[7,5]
 </code> </code>
-Reference is just done by evaluating the variable. + 
 +As we will explain later, the ''[y,x]'' syntax simply denotes a literal table (a.k.a. an array). 
 + 
 +The variable name is used to refer a variable. The variable is evaluated when referenced.  
 +<code> 
 +>x 
 +>>7 
 +</code> 
 + 
 +<note important> 
 +When using the '':='' assignment operator, beware of the following syntactic annoyance: the expression ''a := 1'' denotes an assignment to the variable ''a'', while ''a:= 1'' is misunderstood by the parser as ''a: = 1'', which is the invocation of a keyworded message named ''a:''. Keyworded message sends will be explained later on in this chapter. Hence, as a general rule, don't forget to always put a space between the variable name and the '':='' operator. 
 +</note>
  
 ===== Tables ===== ===== Tables =====
  
-Indexed tables represent what other languages call arrays or listsTables are unidimensional and their indexes range from 1 to the size of the tableAs variables, one can define, assign and refer to a table. Table definition is also made  with the keyword **def** in the following form:+The //table// is AmbientTalk's native compound data type. It is akin to what other languages call //arrays//The main difference is that tables are indexed from ''1'' up to their ''length'', while arrays are indexed from ''0'' up to ''length-1''Like with variables, one can define, assign and refer to a table. Table definitions are also formed with the keyword **def** in the following format:
 <code> <code>
-def t[ <size> ] { <expression> }+def t[ <sizeexpression> ] { <initexpression> }
 </code> </code>
-This means that the <expressionwill be evaluated <sizetimes, i.e., one for each slot of the table. This allows expressions such as initializing a table of ascending numbers as shown below:+This constructs a table, the size of which is determined by ''<sizeexpression>''. The content of each slot is the result of evaluating ''<initexpression>''This means that ''<initexpression>'' is evaluated for each slot in the table! Tables of e.g. ascending numbers are easily formed: 
 <code> <code>
 >def z := 0 >def z := 0
Line 44: Line 50:
 </code> </code>
  
-Although there is no special constructor for definition of multidimensional tables, a table entry can contain another table. This is internally stored as a unidimensional table whose entries are other tables.+Although there is no special constructor for definition of multidimensional tables, a table entry can contain another table. This is internally stored as a one-dimensional table whose entries are other tables.
 <code> <code>
 >def vowels := ["a", "e", "i", "o", "u"] >def vowels := ["a", "e", "i", "o", "u"]
Line 54: Line 60:
 </code> </code>
  
-As shown in the definition of the varible //vocals//evaluating a series of comma-separated abstract grammar values between square brackets (aka a tabulation) results in table.+As shown in the definition of the variable ''vowels''AmbientTalk provides literal syntax to encode in-line tables. Table assignment and indexation work as usual, but recall that table indices range from ''1'' up to ''table.length''. Some more examples of literal tables:
  
 <code> <code>
Line 60: Line 66:
 >>[1, [1, 2, ["a", "e", "i", "o", "u"], 4, 5], "ambientTalk"] >>[1, [1, 2, ["a", "e", "i", "o", "u"], 4, 5], "ambientTalk"]
 </code> </code>
 +
  
 ==== Table Splicing ==== ==== Table Splicing ====
  
-TODO!+AmbientTalk provides the //splice operator// ''@'' to splice tables into surrounding table expressions: 
 +<code> 
 +>[1,@[2,3],4] 
 +>>[1, 2, 3, 4] 
 +>[1, @[2,[3]], [4], @[5], @[], 6] 
 +>>[1, 2, [3], [4], 5, 6] 
 +</code> 
 + 
 +The splicing operator can be also used in the left-hand side of an assignment or definition to separate the head of a table with its rest elements, as shown below. 
 +<code> 
 +>def [first, @rest] := [1,2,3,4] 
 +>>[1, 2, 3, 4] 
 +>rest 
 +>>[2, 3, 4] 
 +</code> 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 + 
 +==== Multidimensional Tables ==== 
 + 
 +As mentioned before, there is no special constructor for definition of multidimensional tables, a table entry can contain another table. In what follows we have a closer look to manipulations with multidimensional tables.  Consider a multidimensional table which is extensionally defined as follows: 
 + 
 +<code> 
 +def a := [[1,0,0], [0,1,0], [0,0,1]]; 
 +>>[[1, 0, 0], [0, 1, 0], [0, 0, 1]] 
 +>a[1][2] 
 +>>0 
 +>a[1] 
 +>>[1, 0, 0] 
 +>(a[1])[2] := 3; 
 +>>3 
 +>a 
 +>>[[1, 3, 0], [0, 1, 0], [0, 0, 1]] 
 +</code> 
 + 
 +An implicit definition of the same table can be expressed as follows: 
 + 
 +<code> 
 +def i := 0; 
 +def aux[3] {0};  
 +def b[3] { i := i + 1; aux := [0,0,0]; aux[i] :=1; aux}; 
 +>>[[1, 0, 0], [0, 1, 0], [0, 0, 1]] 
 +</code> 
 + 
 +You can find later in this chapter a helper function for creating matrices  [[:at:tutorial:basic#optional_parameters|here]]. 
 + 
 + 
  
 ===== Functions ===== ===== Functions =====
    
-As variables and tables, functions are defined with the keyword **def** in the form of: +Analogous to variables and tables, functions are defined with the keyword **def** in the form of: 
 <code> <code>
 def functionname( <arglist> ) { <body> } def functionname( <arglist> ) { <body> }
 </code> </code>
-The argument list is just a list of local variables which are always evaluated one by one from left to right. A basic function looks like this:+The argument list is just a list of local variables which are always evaluated one by one from left to right. Hence, AmbientTalk employs //applicative-order// function calls, like Scheme. A basic ''square'' function looks like this:
 <code> <code>
 >def square (x) { x*x } >def square (x) { x*x }
Line 78: Line 141:
 >>25 >>25
 </code> </code>
-This example also illustrates how functions are called. Calls to functions without parameters must also include the parenthesis as shown below.+This example also illustrates the //canonical// function calling syntax. Calls to functions without parameters must also include the parentheses as shown below.
 <code> <code>
->def f(){nil}+>def f() { nil }
 >><closure:f> >><closure:f>
 >f() >f()
 >>nil >>nil
 </code> </code>
-The return value of a function is the result of the last statement executed. Functions must always return a value - i.e. they cannot be abstract. The example also illustrates how to create dumb function that doesn't do anything but returning the //nil// object+The return value of a function is the result of the last executed statement. Functions always return a value, but a function can always opt to return the //nil// object 
 + 
 +<note> 
 +A function definition is a statement. The body of a function can contain a list of statements, each separated by '';''. A syntax error often made in AmbientTalk is to write: 
 +<code> 
 +def funA() { 
 +  // do something useful 
 +
 +def funB() { 
 +  // do something else 
 +
 +</code> 
 +The parser will complain saying that ''def'' was an unexpected tokenThe reason is that the function definition statements should be separated by means of '';''. In languages like C and Java, the ''}'' token need not be followed by a semicolon, hence the confusion. 
 +</note>
  
-Functions have access to the enclosing environment of its definition as shown in the following example.  +Functions in AmbientTalk are //lexically scoped//, which means that free variables are looked up in the enclosing environment of the function definition. This is illustrated in the following example:
 <code> <code>
 >def counter := 0 >def counter := 0
Line 101: Line 177:
 >def fac(n) {  >def fac(n) { 
   def inner(n, result) {    def inner(n, result) { 
-    if: (n =0) then: { result } else: { inner( n-1, n * result)  }+    if: (n = 0) then: { result } else: { inner( n-1, n * result)  }
   };    }; 
   inner(n,1)   inner(n,1)
Line 110: Line 186:
 </code> </code>
  
-This example also illustrates how a function can be made private by means of lexical scope. Variables and functions defined locally to functions are only visible in the scope of the function where there were defined. Note that local //inner// function is only visible inside the //fac// function and its nested scopes. Thus, calling //fac.inner(2,3)// will return a lookup failure error.+This example also illustrates how a function can be made "privateby means of lexical scoping rules. Variables and functions defined locally to functions are only visible in the scope of the function where there were defined. Note that the local ''inner'' function is only visible inside the ''fac'' function and its nested scopes.
    
 +
 ==== Variable-Length Argument Functions ==== ==== Variable-Length Argument Functions ====
  
-You can create functions that take an arbitrary  number of arguments by means of the splicing operator **@** as shown below:+You can create functions that take an arbitrary number of arguments (also known as a variable arity or polyadic function) by means of the splicing operator ''@'' which collects the actual arguments into a table:
 <code> <code>
->def sum(@args)+>def sum(@args) {
   def total := 0;    def total := 0; 
   foreach: { |el|  total := total + el } in: args;    foreach: { |el|  total := total + el } in: args; 
-  total}+  total 
 +};
 >><closure:sum> >><closure:sum>
 >sum(1,2,3) >sum(1,2,3)
Line 125: Line 203:
 </code> </code>
  
-When the //sum// function is called, the arguments are passed to the function in a table called //args// which can also be modified inside the body of the function. An alternative definition of the //sum// function follows:+When the //sum// function is called, the //args// table is spliced and passed as the argument list to the function. Note that the //args// table can also be modified inside the body of the function.  
 + 
 +Alternatively, we could define the //sum// function to take at least two numbers as shown below:
 <code> <code>
 >def sum(a, b, @rest){ {  >def sum(a, b, @rest){ { 
Line 136: Line 216:
 </code> </code>
  
-In this example the //sum// function accepts an arbitrary number of arguments as long as two arguments, //a// and //b//, are supplied. //a// and //b// are thus considered as mandatory arguments. A function can also declare optional arguments as shown below:+In that case, the //sum// function still accepts an arbitrary number of arguments as long as two arguments are supplied. //a// and //b// are considered as mandatory arguments of the argument list. 
 + 
 +The splice operator can also be used to transform a table into an argument list for a function, for example: 
 +<code> 
 +def args := [3,4,5]; 
 +> sum(1,2, @args); 
 +>> 15 
 +</code> 
 + 
 +One way to think about this is that the splice operator splices the ''args'' table into the table of actual arguments. The "rest" arguments do not necessarily need to be the last parameters, for example: 
 +<code> 
 +> sum(1,2,@args,6); 
 +>> 21 
 +</code> 
 + 
 + 
 + 
 + 
 + 
 + 
 +==== Optional Parameters ==== 
 + 
 +A function can also declare optional arguments as shown below. Optional arguments can be omitted in a function call. If this is the case, the default expression provided in their definition is evaluated and passed as argument to the function instead.
 <code> <code>
 >def incr( number, step := 1){ number + step} >def incr( number, step := 1){ number + step}
Line 144: Line 246:
 >incr(3,3) >incr(3,3)
 >>6 >>6
 +</code>
 +
 +As is customary in languages with the above optional arguments, AmbientTalk requires mandatory parameters to be defined //before// optional parameters, which should in turn be defined //before// a variable-argument parameter, if any.
 +
 +Let us show how to use optional arguments to define an auxilary function that creates matrices:
 +
 +<code>
 +def makeMatrix(n, m := n, init := { |i,j| 0}){
 +  def [i,j] := [0,0];
 +  def makeCol(i,j) {
 +     def col[m] { j := j + 1; init(i,j) }
 +  };
 +  def matrix[n] { i := i + 1; makeCol(i,j)}
 +};
 +>def c := makeMatrix(3);
 +>>[[0, 0, 0], [0, 0, 0], [0, 0, 0]]
 +>c[1] := [1,2,3]
 +>>[1, 2, 3]
 +>c
 +>>[[1, 2, 3], [0, 0, 0], [0, 0, 0]]
 +>def d := makeMatrix(4,4, 
 +  {|i,j| if: (i == j) then: {1} else: {0}});
 +>> [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]]
 </code> </code>
  
Line 160: Line 285:
 >def [get, set] := makeCell(42); >def [get, set] := makeCell(42);
 >>[<closure:getter>, <closure:setter>] >>[<closure:getter>, <closure:setter>]
 +>get();
 +>>42
 </code> </code>
  
-This example also illustrates how a function can make public some of its local fields or functions by returning them as its return value. The get and set could be then passed as arguments to other functions such as //trustedFunction(get,set)// and  //distrustedFunction(get)//+This example also illustrates how a function can make public some of its local fields or functions by returning them as its return value. The ''get'' and ''set'' operations can then be passed separately throughout the application, e.g. an application module that has read-only access to ''val'' only receives the ''get'' closure.
  
 ===== Blocks ===== ===== Blocks =====
  
-In AmbientTalk, blocks are merely syntactic sugar for anonymous closures (aka lambdas).  Blocks are creating using the {} braces in the form of:+In AmbientTalk, blocks are merely syntactic sugar for the creation of anonymous closures (also known as //lambdas//).  Blocks are creating using curly braces in the form of:
 <code> <code>
 { |<parlist>| <body> } { |<parlist>| <body> }
 </code> </code>
-If the block do not require any parameter, the |<parlist>| can be omitted.  Consider a basic block to sum two numbers:+ 
 +If the block does not require any parameter, the |<parlist>| can be omitted. Consider a basic block to sum two numbers:
 <code> <code>
->{| a, b| a+ b} (3,2)+>{|a, b| a+ b}(3,2)
 >>5 >>5
 </code> </code>
 Note that the argument list passed to the block can define the different types of arguments previously explained. Note that the argument list passed to the block can define the different types of arguments previously explained.
 <code> <code>
->{|a, b, @rest| +>def sum := {|a, b, @rest| 
    def total := a + b;     def total := a + b; 
    foreach: { |el| total := total + el} in: rest; total     foreach: { |el| total := total + el} in: rest; total 
- }(1,2,3)+ }
 +>><closure:lambda>  
 +>sum(1,2,3)
 >>6 >>6
 </code> </code>
Line 186: Line 316:
 This example also illustrates that blocks are also used to iterate over enumerations, such as in //foreach: {} in: table//. This example also illustrates that blocks are also used to iterate over enumerations, such as in //foreach: {} in: table//.
  
-AmbientTalk doesn’t support function assigment. Howeverone can assign blocks to variables. In order to call the block the name of the variable must be usedIf the block defined parameters, these are required to the call as argument listWhat follows is an example of such manipulation: +<note> 
-<code+AmbientTalk borrows its block syntax from languages like Smalltalk and Selfwhere the role of the curly braces is played by square brackets, i.e. ''[ args | body ]''. In AmbientTalk, the ''<parlist>'' is only used to denote parameters to the block, not also for temporary variables as in Smalltalk
->def square := { |x| x * x } +</note
->><closure:lambda> + 
->square(1,2) +
->>3 +
-</code>+
  
 ===== Keywords ===== ===== Keywords =====
  
 +AmbientTalk supports keyworded messages, as in Smalltalk and Self. We have already seen some examples of keyword messages in the previous sections such as the ''foreach:in:'' call. Here is how to define a simple ''map:onto:'' function that takes a closure as input and applies the closure to each element of a table:
  
-AmbientTalk supports keyword messagesWe have already seen some examples of keyword messages in the previous sections such as the foreach structure. In AmbientTalk keywords are transformed by the parser into functions in the form:+<code> 
 +>def map: clo onto: tbl { 
 +  def i := 0; 
 +  def mapped[tbl.length] { 
 +    i := i+1; 
 +    clo( tbl[ i ] ) 
 +  }; 
 +
 +>> <closure:map:onto:> 
 +</code> 
 + 
 +It can be invoked as follows: 
 +<code> 
 +>map: square onto: [1,2,3] 
 +>>[1,4,9] 
 +</code> 
 + 
 +In AmbientTalk keyworded functions and function calls are actually syntactic sugar. They are transformed by the parser into their canonical equivalent. Hence:
 <code> <code>
 def foo: arg1 bar: arg2 {...} def foo: arg1 bar: arg2 {...}
-def foo:bar:(arg1,arg2){..} 
 </code> </code>
 +is transformed into:
 +<code>
 +def foo:bar:(arg1,arg2){...}
 +</code>
 +
 +It is also possible to invoke keyworded functions using the canonical function application syntax:
 +
 +<code>
 +foo:bar:(1,2)
 +</code>
 +
 +<note important>
 +Be careful when nesting calls to keyworded functions: a call of the form ''foo: foo: 1 bar: 2 bar: 3'' is parsed as ''foo: (foo: 1 bar: 2 bar: 3)'', not as ''foo: (foo: 1 bar: 2) bar: 3''. It is recommended to always explicitly parenthesize nested keyworded function calls.
 +</note>
  
 ===== Native Data Types ===== ===== Native Data Types =====
 +
 +The basic data types in AmbientTalk are numbers (i.e. integers), fractions (i.e. double precision floating point numbers), text (i.e. strings), tables (i.e. arrays) and booleans. In fact, instances of these data types are nothing but objects and as such, they respond to a variety of native methods. Objects will be the subject of the next chapter of the tutorial. This section explains the basic data types and includes some examples how to manipulate them. The complete list of methods can be found in the language reference.
  
  
-The basic types in AmbientTalk are numbers, fractions, text, tables and booleans. In fact, these data types are nothing but objects and as such, they respond to a variety of native methods. This section shows some examples how to manipulate the basic types. The complete list of methods can be found in the language reference. 
  
 ==== Numerical data types ==== ==== Numerical data types ====
  
-AmbientTalk supports numbers and fractions which represent what other languages call integers and floating point numbers, respectively.  +AmbientTalk supports numbers and fractions which represent what other languages call integers and double precision floating point numbers, respectively.
  
-Note that since numerical types are objects in AmbientTalk, the traditional operators +,-,*,/, >, <, <=, >=, =, !=  are nothing but syntactic sugar for method invocations. Therefore,  //1+1// is internally translated into //1.+(1)//. Unary operators are just applications, e.g. //-5// is internally translated into //-(5)//. What follows are some basic examples of manipulations with numeric types:+Note that since numerical types are objects in AmbientTalk, the traditional operators %%+,-,*,/, >, <, <=, >=, =, !=%%  are nothing but syntactic sugar for method invocations. Therefore,  ''1+1'' is internally translated into ''1.+(1)''. Unary operators are just applications, e.g. ''-5'' is internally translated into ''-(5)''. What follows are some basic examples of manipulations with numeric types:
 <code> <code>
 >1.inc() >1.inc()
 >>2 >>2
->-1.abs() 
->>1 
 >1.cos() >1.cos()
 >>0.5403023058681398 >>0.5403023058681398
Line 231: Line 389:
 >>2 >>2
 </code> </code>
 +
 +Beware of the precedence rules for function application versus method invocation, which may lead to unexpected results, e.g.:
 +<code>
 +>-1.abs()
 +>>-1
 +</code>
 +This code is interpreted as ''-(1.abs())'', hence the result.
  
 Numbers also support some useful iterator methods such as: Numbers also support some useful iterator methods such as:
 <code> <code>
->6.to: 0 step: 2 do: { |i| system.println(i) }+>1.to: 5 do: { |i| system.println(i)} 
 +
 +
 +
 +
 +
 +>1.to: 5 step: 2 do: { |i| system.println(i)} 
 +
 +
 +
 +>6.downTo: 0 step: 2 do: { |i| system.println(i) }
 6 6
 4 4
 2 2
 +0
 >>nil  >>nil 
 >3.doTimes: { |i| system.println(i) } >3.doTimes: { |i| system.println(i) }
Line 272: Line 448:
 >>true >>true
 </code> </code>
 +
  
 ==== Tables ==== ==== Tables ====
    
  
-We have already introduce how to define tables. Let us now focus on how to manipulate them with the native methods provided by the table object.+We have already introduced how to define tables. Let us now focus on how to manipulate them with the native methods provided by the table object.
 <code> <code>
 +>[1,2,3].filter: {|e| e != 2 }
 +>>[1, 3]
 +>[1,2,3].map: { |i| i + 1 }
 +>>[2, 3, 4]
 >def vowels := ["a", "e", "i", "o", "u"] >def vowels := ["a", "e", "i", "o", "u"]
 >>["a", "e", "i", "o", "u"] >>["a", "e", "i", "o", "u"]
Line 288: Line 469:
 >vowels >vowels
 >>["z", "e", "i", "o", "u"] >>["z", "e", "i", "o", "u"]
->vowels.isEmpty() 
->>false 
->[1,2,3].filter: {|e| e != 2 } 
->>[1, 3] 
->[1,2,3].map: { |i| i + 1 } 
->>[2, 3, 4] 
 >vowels.select(2,5).implode() >vowels.select(2,5).implode()
 >>"eio" >>"eio"
 +>vowels.isEmpty()
 +>>false
 </code> </code>
  
-Tables also support some useful iterator methods such as:+Tables also support some useful iterator methods as shown below. 
 <code> <code>
 >def sum:= 0;  >def sum:= 0; 
Line 316: Line 494:
 ==== Booleans ==== ==== Booleans ====
    
- +As any native type, booleans are objects so, they respond to keyword messages such as:
-AmbientTalk supports infix operators for booleans as &, | and !. As any native type, booleans are objects so, they respond to keyword messages such as:+
 <code> <code>
 <booleanexpr>.ifTrue: { ...}  <booleanexpr>.ifTrue: { ...} 
Line 325: Line 502:
 </code> </code>
  
-**=** and **!=** are the infix operators for equality and inequality. **true** and **false** are the boolean constant objects. What follows is some basic examples of boolean manipulation:+''='' and ''!='' are the infix operators for equality and inequality. The prefix operator ''!'' represents logical negation. ''true'' and ''false'' are the prototypical boolean singleton objects. What follows is some basic examples of boolean manipulation:
 <code> <code>
 >(0 < 1).ifTrue: { 0 }  >(0 < 1).ifTrue: { 0 } 
Line 332: Line 509:
 >>1 >>1
 > def [i, j] := [1,3] > def [i, j] := [1,3]
->>>[1, 3]+>>[1, 3]
 >{i < j}.whileTrue: { system.println(i); i := i + 1 } >{i < j}.whileTrue: { system.println(i); i := i + 1 }
 1 1
Line 339: Line 516:
 </code> </code>
  
-Boolean infix operators such as & and | are not shortcut. Thus, both arguments will be evaluated. For lazy evaluation, you should use the natives methods. For example, false.and: { 1/0 } will return false without executing the second argument.+Compound boolean expressions can be created by means of a boolean's ''and:'' and ''or:'' methodswhich both take a zero-argument closure as argument. For example, ''false.and: { 1/0 }'' will return ''false''. The block is not applied because a logical //and// with ''false'' always fails.
  
-===== Control Flow Structures =====+===== Control Flow Constructs =====
  
-Control flow structures are defined in the lexical root of AmbientTalk. The lexical root is an object containing globally visible native methods. We have already seen in the previous sections examples of usage of the foreach and if/then structures. The complete list of traditional control flow structures defined in AmbientTalk is shown below:+Control flow constructs are defined in the "lexical root". The lexical root is an object containing globally visible native methods (i.e. it is the top-level environment). We have already seen in the previous sections examples of use of the foreach and ''if:then:'' control structures. list of traditional control flow structures defined in AmbientTalk is shown below:
 <code> <code>
-if: booleanCondition then: { consequent } +if: booleanCondition then: consequentClosure 
-if: booleanCondition then: { consequent } else: { alternative } +if: booleanCondition then: consequentClosure else: alternativeClosure 
-while: { condition } do: body } +while: conditionClosure do: body 
-foreach: { |v| body } in: table ] +foreach: iteratorClosure in: table 
-do: { body } if: condition +do: bodyClosure if: condition 
-do: { body } unless: condition+do: bodyClosure unless: condition
 </code> </code>
-An example of usage for some of these structures is shown below in the definition of the sort function.+ 
 +<note warn> 
 +Note that ''conditionClosure'' in the ''while:do:'' construct denotes a //closure// that should return a boolean value. It needs to be a closure because the code is evaluated repeatedly until the closure returns false. ''bodyClosure'', ''consequentClosure'', ''alternativeClosure'' all denote zero-argument closures. As a general rule, all code that needs to be delayed or executed repeatedly must be wrapped in a closure. 
 +</note> 
 + 
 +The above definitions in the lexical root of AmbientTalk are simply convenience functions for the methods defined on booleans and closures. For example, an if-statement can also be encoded as a message send, as in Smalltalk: ''boolean.ifTrue: {...} ifFalse: {...}''
 + 
 +An example of usage for some of the above structures is shown below in the definition of the sort function.
 <code> <code>
 >def sort(table, cmp := { |e1,e2| e1 < e2 }) { >def sort(table, cmp := { |e1,e2| e1 < e2 }) {
- def quickSort(table, low, high) { +  def quickSort(table, low, high) { 
-     def left := low; +    def left := low; 
-     def right := high; +    def right := high; 
-     def pivot := table[(left+right) /- 2]; +    def pivot := table[(left+right) /- 2]; 
-     def save := nil; +    def save := nil; 
-        while: { left <= right } do: { +    while: { left <= right } do: { 
-     while: { cmp(table[left], pivot) } do: {  +      while: { cmp(table[left], pivot) } do: {  
-                left := left + 1  +        left := left + 1  
-            }; +      }; 
-     while: { cmp(pivot, table[right]) } do: {  +      while: { cmp(pivot, table[right]) } do: {  
-                right := right - 1  +        right := right - 1  
-            }; +      }; 
-     if: (left <= right) then: { +      if: (left <= right) then: { 
-     // swap elements +        // swap elements 
-     save := table[left]; +        save := table[left]; 
- table[left] := table[right]; +        table[left] := table[right]; 
- table[right] := save; +        table[right] := save; 
- left := left + 1; +        left := left + 1; 
- right := right - 1; +        right := right - 1; 
-     }; +      }; 
-    }; +    }; 
-    if: (low<right) then: { quickSort(table,low,right) }; +    if: (low<right) then: { quickSort(table,low,right) }; 
-    if: (high>left) then: { quickSort(table,left,high) }; +    if: (high>left) then: { quickSort(table,left,high) }; 
-    table; +    table; 
-   }; +  }; 
-   quickSort(table, 1, table.getLength()); +  quickSort(table, 1, table.length); 
- };+};
 >><closure:sort> >><closure:sort>
 >sort([2,37,6,4,5,8]) >sort([2,37,6,4,5,8])
 >>[2, 4, 5, 6, 8, 37] >>[2, 4, 5, 6, 8, 37]
 </code> </code>
 +
 +AmbientTalk has no ''return'' statement. To achieve a similar jump in the control flow, see the section on [[:at:tutorial:modular#escaping_continuations|escaping continuations]].
at/tutorial/basic.1175842687.txt.gz · Last modified: 2007/04/06 09:00 (external edit)